OceanofPDF.

com
Anatomy of Game Design
People have played games forever, but it’s only in the past few decades that
people really started thinking about what games are, how they work, and
how to make them better.
Anatomy of Game Design takes some of the most popular and beloved
games of all time and dissects them to see what makes them tick. By
breaking down the systems and content of each game, the underlying
systems of game design are laid bare.
Eight games are analyzed – including Settlers of Catan; Centipede;
Candy Crush Saga; Papers, Please; Magic: The Gathering; and more –
each representing a different genre or era of game design. Each game is
discussed in detail, using the same methods for each game. What are the
verbs of the game that give the player agency? How do those verbs fit
together to form a core loop that makes the game engaging? What are the
systems that power the gameplay? What is the larger flow that makes the
game interesting over and over again?
Each game is then used as an example to tie back to one or more larger
topics in game design, such as systems design, randomness, monetization,
game theory, and iterative approaches to game development.

Key Features

Uses well-known games to provide specific, discrete examples of

broader game design theory
Discusses eight popular games using the same methodology to allow
comparison of different types of games
 Includes both high-level theory and academic perspective and
practical, real-world guidance from a working game designer who has
created these games for commercial release
Provides clear direction for deeper inquiry into game design or related
fields such as psychology, anthropology, game development, or
systems thinking

OceanofPDF.com
 Anatomy of Game Design
Tom Smith

OceanofPDF.com
First edition published 2025
by CRC Press
2385 NW Executive Center Drive, Suite 320, Boca Raton FL 33431
and by CRC Press
4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
CRC Press is an imprint of Taylor & Francis Group, LLC
© 2025 Tom Smith
The views and opinions expressed in this work are those of the author alone and do not necessarily
reflect those of Roblox or any other previous employers or any related entities.
Candy Crush Saga is the copyright product of King.com Ltd. and is referenced for discussion
purposes only. The author of this book has no affiliation with Candy Crush Saga or King.com Ltd.
This book represents the personal views and opinions of the author.
Reasonable efforts have been made to publish reliable data and information, but the author and
publisher cannot assume responsibility for the validity of all materials or the consequences of their
use. The authors and publishers have attempted to trace the copyright holders of all material
reproduced in this publication and apologize to copyright holders if permission to publish in this
form has not been obtained. If any copyright material has not been acknowledged please write and let
us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, access
www.copyright.com or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive,
Danvers, MA 01923, 978-750-8400. For works that are not available on CCC please contact
[email protected]
Trademark notice: Product or corporate names may be trademarks or registered trademarks and are
used only for identification and explanation without intent to infringe.
ISBN: 978-1-032-38755-0 (hbk)
ISBN: 978-1-032-38738-3 (pbk)
ISBN: 978-1-003-34658-6 (ebk)
DOI: 10.1201/9781003346586
Typeset in Minion
by SPi Technologies India Pvt Ltd (Straive)
Access the Support Material: https://www.routledge.com/9781032387550

OceanofPDF.com
 I dedicate this book to my lovely wife, Wendy, who has been there for me
every day and who I love more than anything in the world. You are beautiful
and clever and wise in ways that I will never be, and I’m so blessed to have
you in my life.
And the kids – Duncan, Griffin, Sophie – who make it worthwhile. I’m so
proud of each of you.
OceanofPDF.com
Contents
Acknowledgments
About the Author
Chapter 1 Game Dissections
Chapter 2 Shovel Knight / Verbs
Chapter 3 Candy Crush Saga / Goals
Chapter 4 Centipede / Dynamics
Chapter 5 Settlers of Catan / Uncertainty
Chapter 6 Spelunky / Systems
Chapter 7 Magic: The Gathering / Content
Chapter 8 Royale High / Audience
Chapter 9 Papers, Please / Meaning
Chapter 10 Post Mortem
Bibliography – Games
Bibliography – Non-Game
Index

OceanofPDF.com
Acknowledgments
Thanks to my family for putting up with me: Wendy; Duncan, Griffin,
Sophie; Mom and Dad; Ken and Pam; Ed and Dena; Daniel and Amy and
Evan.
Thanks to my friends who took a look at early drafts.
Thanks to my mentors Eric, Michael, Cynthia, and many others who
helped me formulate the ideas you see here.
Thanks to my many co-workers who learned these lessons alongside me.
Game development is a team effort, and I’ve had the luxury of working on
many great teams with talented people who care.
Thanks to all my Moorpark students over the years who were the anvil
upon which this book was forged.
Thanks to my publisher for accepting my imperfect process.
Thanks to the game creators and publishers who generously gave
permission to discuss their games.
Thanks to everyone who has made a game. It’s hard, and I appreciate the
fruits of your labors.

OceanofPDF.com
Author
Tom Smith was born the same year as the first home video game console.
Tom Smith is not the folk musician of the same name.
Tom Smith was raised near Chicago and received early arcade training at
Galaxy World. His first home console was an Atari 2600.
Tom Smith has an Ivy League degree that only slightly applies to his
career in that they’re both about systems.
Tom Smith has been a game designer for a long time and worked on a lot
of different types of games. Here are a few of the ones he contributed the
most to:

1995-1997: Mayfair Games: Settlers of Catan (board game), Fantasy

Adventures (trading card game), Unknown Providence (RPG
sourcebook)
1997-2005: High Voltage Software: Paperboy 64 (Nintendo 64),
Ground Control: Dark Conspiracy (PC), Disney’s Stitch: Experiment
626 (PS2), Hunter the Reckoning (Xbox), Disney’s The Haunted
Mansion (PS2, Gamecube, Xbox), Duel Masters (PS2), Leisure Suit
Larry: Magna Cum Laude (PS2, Xbox, PC), Charlie and the
Chocolate Factory (PS2, Gamecube, Xbox), Codename: Kids Next
Door - Operation V.I.D.E.O.G.A.M.E. (PS2, Gamecube, Xbox)
2005-2007: THQ: The Sopranos: Road to Respect (PS2), Conan (PS3,
Xbox360), Drawn to Life: Spongebob Squarepants Edition (DS)
2009-2014: Disney Mobile: Pirates: Master of the Seas (mobile),
Where’s My Water? (mobile), Where’s My Perry? (mobile), Where’s
My Water? Featuring XYY (mobile), Where’s My Water? 2 (mobile),
 Alice in Wonderland (DS), Cars 2 (mobile), Disney Fairies: Fly!
(mobile, tablet), Tron Legacy (mobile)
2014-2015: Disney Imagicademy: Mickey’s Magical Math World
(tablet), Frozen Early Science (tablet), Mickey’s Magical Art World
(tablet)
2015-2017: Zindagi Games, acquired by Zynga: Crazy Cake Swap
(mobile), Games With Friends (messenger), Disney Dream Treats
(mobile)
2017: PierPlay: Scrabble Go (mobile)
2017-now: Imbellus, acquired by Roblox: assessment games

Tom Smith has been teaching game design at Moorpark College for many
years.
Tom Smith lives near Los Angeles with his wife, children, and cats.

OceanofPDF.com
Chapter 1
Game Dissections
DOI: 10.1201/9781003346586-1

Hi, I’m Tom.

I’ve been a game designer for a very long time. I’ve worked on all sorts
of games on all sorts of platforms – see the Author section a few page ago if
you want a list. I also teach a class on game design. I love it.
Game design is a relatively new discipline. People have played games
forever, but it’s only in the past few decades that people really started
thinking about what games are and how they work and how to make them
better.
There are already lots of good resources on the underlying concepts of
games (Rules of Play, Book of Lenses) and practical advice on how to make
games (Level Up!). I want this book to fall somewhere in between. A very
practical look at specific games and what makes them work. But looking at
those specifics is a great way to better understand the deeper underlying
ideas. When I teach, I usually start each class with a dissection and I’ve
found it’s a good way to introduce ideas.
This is written for an introductory level. It is intended for someone who
loves games but hasn’t spent a lot of time thinking critically about them, or
someone whose primary interest is in another aspect of games but is curious
about game design. If you’ve been designing games for a few years, many
of these lessons will be more of the “nod your head in agreement” type of
insight, not the “I’ve never thought of it that way before” type. But in my
experience, both types of insights are good.

What Is a Game?
One theme that will come up in these discussions is that terminology
around games is not very precise. Because games are a new field of study,
people are still struggling to find the right terms for things. A definition may
be thrown around for a while before it finally settles down somewhere.
So, there are terms like “fun” that most people know but don’t define,
terms like “mechanic”, which people generally seem to agree on but some
use slightly differently, and terms like “game” that seem like they make
sense until you think about it. You would think that if I’m going to spend a
whole book talking about how games work, I would at least tell you exactly
what a game is. But often the simplest words are the hardest to define. It’s
the same with words like “art” or “life” in other fields.
When confronted with these tough-to-define words, I generally take a
very broad approach. If it might be a game, I’ll call it a game. When you’re
trying to understand what someone else is saying, assume they mean the
broadest possible definition until they indicate otherwise. Don’t stress out
about it if their definition isn’t the same as yours.
But to understand games, we must understand what makes something a
game. What makes a game different from other forms of entertainment? I
like to read books and watch movies and have deep conversations and go on
hikes. These are all fun ways to pass the time. But a game is the only one
you can win. It may seem like you can win a conversation, but that hardly
ever actually works. A game is the only one with clearly defined rules that
participants agree to follow. “Don’t talk at the movies” is a rule, but not a
consistent one, unfortunately. A game is the only one where the things that
make it difficult are artificially and intentionally added. A hike may be hard,
but you don’t add spikes and bottomless pits along the way.
So for me, the word “game” includes lots of things:

Video games like Centipede (1981)

Board games like Settlers of Catan (1995)
Card games like poker
Solo games like Klondike
Digital worlds like World of Warcraft (2004)
Playground games like tag
Sports like basketball

What do all these things have in common?

Interactive: The player can change what happens.

Goals: The player wants something.
Rules: There are explicitly defined rules on how to do the thing.
Uncertainty: The outcome is not predetermined.
Challenge: Achieving the goal requires some effort, usually by making
the right choices and/or demonstrating a skill.
Pretend: This isn’t real.

This doesn’t really lead to a cleanly packaged definition. “Games are

interactive goals where players choose to follow rules that use uncertainty
to create a challenge”. Feel free to work on a dictionary definition of your
own if that’s important to you. What’s important to me is understanding
each of these concepts and how they relate to games.
Let’s talk about each of these words.
Interactivity
The most important quality that makes something a game is interactivity.
The player needs to be able to change the outcome of the activity based on
the player’s actions. The player’s actions can be conscious choices (“I want
to turn left, not right”) or reactions (“There’s a ball flying toward my
head!”) or tests of skill (“I want to catch the ball, but it’s moving very
fast”).
This is very different than most other forms of entertainment. When you
watch a movie, you can’t change what happens in the movie. Sure, you can
change your perception and in that respect the observer is part of the
process. But with games, the player is an active participant whose input can
change the outcome. The same cannot be said for paintings, books, statues,
or really any other type of art. The intersection of games and fine art is a
fascinating space that people have been playing with for a while (see: dada)
and continue to poke at in interesting ways. Pay attention to this – I predict
great things as the gaming generations get older.

Choices
Player input can come in many forms. In many games, interactivity is based
on choices. When you’re playing checkers, you pick a piece and choose
where to move it. When you’re playing a strategy game like Civilization
(1991) on a PC, you choose where to build things or how to spend your
resources. When you’re playing a fighting game like Street Fighter 2
(1991), you choose where to move your character and which attacks to use
at what times. These choices are what drive the game forward and
determine the outcome. Good choices lead to a win, bad choices lead to
failure.
 There’s a quote from Sid Meier, the designer of many amazing games
including Civilization, about this: “a game is a series of interesting
choices”. This isn’t a precise definition and doesn’t include all games, but
it’s a concise and insightful reminder of what the job of a game designer is.
If you’re not giving the player choices, and/or if those choices aren’t
interesting, then you’re not doing a great job.
While I consider choices the most important part of any game, there are
exceptions. Candyland (1948) is a classic example of a game where the
player makes no choices. You don’t choose which card to take, and you
don’t choose where to go when the card is flipped. Candyland is interesting
to me because I do think there are important interesting decisions in the
game, but they’re all outside of the rules. For the target audience of three- to
five-year-olds, the real challenge of Candyland is learning how to play a
game. Speaking as a parent, I can attest that the target audience is
adequately challenged by drawing the correct number of cards, moving to
the correct space, and only taking a turn when it is the correct time. Often,
the real challenge and interactivity of the game is only visible when you
understand the game’s audience and how they interact with the game.

Skill
Choices aren’t the only way success can be determined in games.
Baseball is a game with lots of decisions. What type of pitch should the
pitcher throw? Where do you throw the ball when you field it? But a large
part of a team’s success in baseball is separate from those decisions. The
individual skill of each player has a huge effect. Choosing where you want
to throw the ball doesn’t mean it’ll get there, or that it’ll get there in time. In
baseball, the outcome of a choice is determined largely by the player’s skill
at making that happen.
 This is not only true in sports. The same could be said for hand–eye
coordination in a first-person shooter or even in certain physical board
games like tiddlywinks or marbles. If I choose to shoot someone in a first-
person shooter, my ability to aim precisely and quickly is going to have
more effect on the outcome than my choice of targets. Which is why I don’t
play a lot of competitive first-person shooters. Even before I got old, I
wasn’t great at twitch gameplay.
It’s rare to find a game that is entirely skill with no decision-making.
Some games get close – a 50-yard dash, a carnival strength-tester, or certain
old arcade games like Track and Field (1983). Even when it seems like a
game is all about skill, you can find subtle decisions that have a large
impact. Talk to a runner, and they’ll tell you how they chose when to exert
themselves or their choice of breathing patterns. Skill is often the way to
determine the result of a decision more so than a replacement for a decision.
The important thing for our definition is that the player brings something
to the table (choice or skill) that has a meaningful effect on the game. The
player has the agency to change outcomes.

Goals
If being a game means that the player can “change outcomes”, that means
there should be an outcome. A goal. Something the player is trying to do or
accomplish to guide the player’s actions. A way to win the game. In most
games these are clearly defined. Most board game rulebooks start with a
quick summary of the goal, as that is the context necessary to understand all
the other rules.
This draws the line between games and other activities. In English, we
say that you “play” a game. In many languages, the word for “play” and
“game” are the same. So, play is definitely a key part of games. But if
you’re bouncing a ball against a wall, you might say that you’re playing
with the ball, but you wouldn’t say you’re playing a game. It’s only once
you set a goal for yourself – “how many times can I bounce this ball before
the teacher stops me” – that it becomes a game. Goals are a mindset.
Anything can be a game if you think of it as a game – this is the concept
behind using gamification to make tedious or unappealing tasks more
engaging for people.
Not all games explicitly provide a goal. Some games let the player make
up their own goals. Most tabletop role-playing games (like Dungeons &
Dragons (1974)) are never-ending stories told by a group. The players may
have different goals each session, but the goals will change all the time and
continuously layer on top of each other, so there’s usually not a way to
“win” the game. But the goals are still present to guide players’
interactivity.
Similarly, in the digital space, some games don’t tell the player much
about what they should do. The classic PC game SimCity (1989) is famous
for this. The player is clearly intended to build a city, but the game doesn’t
tell the player what type of city to build or ever tell the player “You Win!”.

When I started working on SimCity, I showed it to Brøderbund and

they said, “Sure, let’s do it.” But they kept wanting to change it. I’d
kind of programmed it to the point where I thought it was done, and
they didn’t think it was nearly done. They kept wanting a win/lose.
They were expecting more of a traditional game out of it. But I always
wanted it to be much more open-ended, more of a toy. So they never
published it.
(The Replay Interviews: Will Wright (gamedeveloper.com))
Will Wright wants to give players a toy, and let them decide how many
times to bounce it on the wall. The goal still exists, but the power and
responsibility to create it move to the player. The vast majority of SimCity
players know what they’re trying to do. And the small percentage who
really aren’t trying to achieve anything can enjoy SimCity as a toy. They’re
just bouncing their ball, which can be really fun and doesn’t detract from
the value of the experience – but does change the categorization.
And you can see this in other modern games. Massively Multiplayer
Online Role-Playing games are endless layers of goals similar to a tabletop
role-playing game. The game provides never-ending goals, and picking
which ones to care about creates a personal path through the game.
Minecraft (2011) lets you find your own goals, in the same way that
SimCity does. Many games on Roblox like Adopt Me (2017) or Bird (2019)
provide the structures of a game but let the player decide if they want to
build, progress, collect, or just socialize.
But that’s not to say these games don’t have goals. A goal that the player
makes up on their own has as much validity as a goal that is stated by the
game. Either way, the goal exists in the player’s mind. And that’s where it
should be. The goal of a goal is to guide the player as they make decisions.
As a game designer, trusting the player to make their own goals is
actually harder than pre-generating the goal for the player. You need to
provide all the necessary building blocks of fun goals and trust the player to
assemble them on their own. As with many things in games, player goals
don’t happen by accident and require careful thought ahead of time to get
right.
Goals are there to make your player’s choices interesting. Human minds
love the feeling of a goal. Being able to measure progress and check off
accomplishments is something that we’re built to do. Psychology has shown
that this is true in reality. In games, we simulate controlled experiences in
order to trigger those same responses without the accompanying risk that
makes risk-taking in reality more complicated. Games provide a safe space
to try out different goals and trigger those positive responses.
In order to trigger those positive responses, games need to define those
goals in a way that makes sense and makes fun. That’s where rules come in.

Rules
Rules provide the structure of a game to make the interactivity and goals
interesting.
Games are about changing outcomes. Players use abilities to strive
toward goals. Rules are the network of pieces that make this possible. Rules
determine how the player’s controls function, and how the game responds
when the player does something. Good rules make that process fun.
In checkers, players express their agency by moving pieces. But players
aren’t allowed to do whatever they want with those pieces.

You can only move pieces of your color.

You can only move one piece per turn.
You can only move one space per turn (usually).
You can only move diagonally.
You can jump over enemy pieces.

These constraints on the player action are necessary for the game to
function and for the player to have fun. If I could move all my pieces
anywhere I wanted as often as I wanted, I could win the game on my first
turn and then it’s no fun for anyone. Limits on the player’s actions are what
make them fun. Being all-powerful is boring. Games have rules to provide a
structure that makes the experience interesting.
 Another type of rule defines outcomes of player actions. In checkers,
these are pretty simple.

If your piece jumps over an enemy piece, remove it from the game.
If your piece moves onto the opponent’s back row, upgrade the piece.
If you remove the opponent’s last piece, you win.

These rules are as strict and controlling as the limits on the player’s actions.
You can’t remove pieces unless you meet the conditions set forth in the
rule. You can’t upgrade unless you’re in the back row.
In a video game, these reaction rules get much more complicated. When I
move in a first-person shooter, there can be lots of consequences, and the
code needs to instantly account for all of them:

The player’s character may bump into a wall and stop moving.
The player’s character may move over a power-up and acquire it.
The player’s character may collide with a projectile and start a hit
reaction.
Every AI character needs to adjust their behavior based on the player’s
new position.

Video game rules can also get complicated because there are often two
layers to these reactive rules. Rules are there to guide the player to their
goal and display the results of their actions. But in many modern video
games, there is also an attempt to “feel” like the real world. This
verisimilitude is only possible with a large swath of rules. Getting 3D
objects to move in a way that convincingly feels like real-world physics
takes a lot of rules. When we’re talking about rules in this book, we’re
usually not talking about those rules. But those rules are a big part of game
development, so it’s important to keep them in mind. Modern game engines
like Unity or Roblox do a great job of handling many of these basic
background rules for you, but it’s important to consider them when
designing a game that wants to feel like reality.
A large part of a video game designer’s job is thinking of all of these
reactions. And coming up with the player’s actions and how they work
before any reactions occur. Defining the rules of the game defines a big
chunk of the player’s experience and also defines the possibility space of
what can happen in the game. The job of a game designer is to find the right
balance of rules so the game creates an enjoyable set of actions and
reactions.

Uncertainty
Designing a game involves adding uncertainty to a system. Being all-
powerful is boring. Players want to have actions they can do, with clear
rules that define how those actions work, but without knowing ahead of
time how all the actions and rules are going to combine. If you know how
it’s going to end, it’s not a game. Or at least not a fun game.
For interactivity to work, there need to be multiple possible outcomes. If
the player has three choices, and all of them lead to the same result, then it’s
not really a choice.
Remember that Sid Meier quote from earlier? “Games are a series of
interesting choices”. What makes those choices interesting? A big part of
that is the uncertainty. If you tell me that there are three doors and they all
open up to a big ballroom, then my choice of which door to open isn’t very
interesting (Figure 1.1). It lacks uncertainty. If instead (Figure 1.2) you tell
me that there are three doors, and one leads to treasure (represented here by
the letter A) and one leads to a hungry tiger (here represented by a poorly
drawn face), that makes the choice much more interesting. The range of
possible outcomes has increased, and the cognitive difference between
those outcomes has also increased. Three similar outcomes are not as
interesting as three outcomes with different positive and negative effects.

Figure 1.1 Three doors.

Figure 1.2 Three doors, open, with different things behind each.

Actually, making a player feel like they’re in a game doesn’t even require
three real results. It just has to feel like there might be three results before
the player opens the door. This is how a lot of narrative games work. The
story always ends up in the same place, but the player doesn’t know that.
Each door seems different and consequential, and in the player’s mind, each
leads to a completely different room. So, the player opens one, finds a room
on the other side, and doesn’t think too hard about what was behind the
other doors. It’s only when the player goes back and replays the game again
that they realize how little the story changed as a result of their choices. But
by that point, the player has already had a positive experience and is willing
to forgive the game for some necessary limits. The important thing is that it
felt uncertain that first time.
Another way to look at this is tic-tac-toe (aka Noughts and Crosses).
There aren’t a lot of adult fans of tic-tac-toe because the game lacks
uncertainty. When playing with two relatively knowledgeable players, the
only uncertainty is whether or not one player makes a mistake. It becomes a
game about attention to detail rather than a game about strategic thinking.
And while you can build a fun game around attention to detail, such as the
hidden object genre, tic-tac-toe is not that game.
The uncertainty that makes a game fun can come from many sources.
We’ll discuss these in future chapters.
Generating uncertainty requires more than just actions, goals, and rules.
Especially because uncertainty in any of those areas is not fun. Players want
their actions to behave predictably and reliably. Players want goals that
make sense and don’t change unexpectedly. And a large part of why games
are fun comes from rules that are consistent and fair, unlike the real world.
So, uncertainty needs to come from somewhere else. That’s why games
include challenges.

Challenges
There’s a book I really like called Grasshopper. It’s a philosophy book
about games. At one point, there’s a discussion of golf. Golf is a game
where to goal is to get a ball into a hole. Given that goal, then logically the
player should pick up the ball, walk to the hole, and drop it in. That’s the
most efficient way to achieve the goal. But that’s not golf. Golf has a
structure of rules that limit the player’s ability to achieve the goal
efficiently, including “you can’t pick up the ball”. The only way you can
interact with the ball in golf is by hitting it with a stick. This is terribly
inefficient, but it makes the activity fun. Constraining the player’s agency
makes the activity challenging, which makes it fun.
This ties back to another book I (generally) like called Flow. Mihaly
Csikszentmihalyi was interested in what similarities he could find in people
who were the top of various fields. So, he interviewed concert pianists and
top scientists and business leaders. A key thing he found was they were
always pushing themselves. They would engage in activities that would
challenge their skills and abilities but were still within their skill range. And
when their activities fit these definitions, they were so engaged that the real
world melted away and they completely lost track of time while pushing
against these challenges.
Flow was written for psychology, but professional game designers found
it fascinating. This is exactly what games had been doing all along. Games
create scenarios where a player is tested and challenged and has to work
hard to achieve a victory. But the player always knows that victory is
possible. A game where the player knows they can never win is not fun
(sorry, Kobayashi Maru). The best games find that perfect balance of
difficult but possible that aligns perfectly with the psychology described in
Flow.
The challenges of a game are what make it fun. Without challenge, goals
don’t matter and the player’s decisions don’t really have meaning. This
doesn’t need to be a high bar – there’s a growing genre of “cozy games”
like Animal Crossing (2001) or Cozy Grove (2021) that strive to be more
relaxing than challenging. There are still goals to accomplish and economic
limits that prevent the player from doing anything at any time, but the
progression from goal to goal doesn’t take a lot of complex decision-
making or advanced skills. There are still challenges and uncertainty, but
they’re toned down to focus on other things. It’s more about the progression
than the challenge, which creates a relaxing mellow experience in contrast
to the high-energy mood of many other games.

Pretend
What’s the difference between a game and a presidential election? Both
have rules that have to be followed. Both have a goal. Both have challenges
to overcome. But no one would seriously consider the election to be “just a
game”. The difference is the outcome of the election has real consequences
that matter.
When something is real, it’s not a game. Games are inherently imaginary.
The book Rules of Play does a nice job of covering this idea. They apply
the concept of the “Magic Circle” from anthropology – specifically, Johan
Huizinga in Homo Ludens: A Study of the Play-Element in Culture. Inside
the circle, the rules of the game take effect and the objects of the game take
on special meaning. Outside of the circle, they’re just part of reality. So
outside of the circle, it’s a painted block of wood. Inside of the Settlers of
Catan circle, it’s a village with a thriving community life and backstory,
which is important because it is worth one Victory Point. Once you leave
the circle, that Victory Point is meaningless and the thriving village
becomes a piece of wood again, but inside the circle it’s all that matters.
The important thing about the Magic Circle is that it is imaginary and it
is voluntary. No one can make you play a game if you don’t want to. If you
don’t agree that the piece of wood is a village, it’s going to remain a piece
of wood for you no matter how hard your friend tries to convince you
otherwise. And if you insist that Victory Points aren’t a real thing, you’re
not playing a game.
Even when games have other goals than just fun, there’s still a magic
circle that separates gameplay from reality. When a teacher uses a game to
illustrate an idea in a class, during play the gameplay creates separate
meanings and roles from imagination that make the goals and rules of the
game work. Once the game ends, the class might talk about the game from
outside the Magic Circle to connect it back to reality and learn a valuable
lesson. But the meanings and imagination that existed inside the game are
still separate.
This creates some interesting questions when people play games for
money. Is a professional football player playing a game, or is the thing
they’re doing more like a presidential election where the player is following
rules to achieve a real-world goal? I would argue that the NFL player’s
motivation might be outside of the Magic Circle, but during the game
they’re still agreeing that moving the ball a certain number of yards has
special meaning that it doesn’t have in reality. They may be motivated to
play a game for money, but they’re still playing a game. The experience of
football is trying to create a fun activity, even if outside goals are being
added outside of the Magic Circle.

Why Dissection?
This book uses dissection as a tool to discuss various game design topics.
But dissection is also a valuable skill in its own right. Game designers are
often asked to dig into other games and understand how they work. The
dissection tools discussed in this book will make you better at this type of
analysis.
Even if you’re not planning on becoming a game designer or game
developer, dissection skills can make your life better. When you’re playing
games for fun, you’ll get more out of them if you understand them better.
The more you know about the craft, the more you can appreciate the art.
Some people complain that analyzing games makes it harder to enjoy
games. I enjoy analysis, but the best analysis occurs after play. The only
time my analytical skills override my inherent ability to have fun in a game
is when the game isn’t doing its job. It’s the responsibility of the game to
keep the player immersed and engaged in the fun – inside the Magic Circle.
I only slip into deep analytical mode on accident when the game is failing at
that. And in that case, analysis is probably going to be more interesting than
the game itself. The best part about bad games (to me) is that they can
provide insights on what doesn’t work, which is immensely valuable as a
game designer.

Pre-Dissection
Dissection is also useful for games that don’t exist. When you’re trying to
come up with an idea for a new game, it’s useful to understand that idea in
great detail. The process of defining a game idea in enough detail to build it
is very similar to the dissection process. You need to see each piece of the
system on its own, and also see how they all fit together into something
greater.
The better you get at dissecting, the better you’ll be as a game designer.
Game development is just game dissection in reverse. If you can break an
existing game apart and identify everything that makes it work, you can
apply those insights into games that don’t exist yet.
Game Design Is Thievery
Part of the value of dissection is it makes you a better thief. All games are
built on concepts pioneered by previous games. A good game designer
needs to be able to identify good ideas from other games and convert them
into new forms for new games. This second part is important – you don’t
want to copy and paste an entire game. That is actually thievery. The best
game designers pick and choose carefully, taking interesting ideas and
presenting a new perspective on them by combining them in new ways. The
more you understand about dissection, the easier it is to identify which
ideas from other games make sense for yours.

Meta-Dissection
Part of the value of dissection is it helps you see how each game connects to
game design as a whole. If you can tear apart a game into its component
parts, you can see how those parts relate to the parts of other games. If
you’re working in an established genre with multiple successful games, it’s
important to understand the connective tissue that defines the genre and
makes it work. As a game designer, you don’t want to re-invent the wheel.
Being able to cross-reference successful features across multiple games is
one of the few ways to predict what’s going to work at the start of the
design process. So, embrace it and steal the best ideas you can find.
For years, game designers have been trying to call out these ideas and
determine the general rules for how games should work. This is sometimes
called “Game Design Patterns”, a term borrowed from programming. This
is an interesting line of thought that has yet to provide many universal rules.
Identifying patterns can provide good insights into what works and what
doesn’t for a particular game or genre. Patterns can be a useful way to
discuss and compare these insights. But it’s proven difficult to extrapolate
these learnings to all games.
Generalizing game design insights to encompass all games is a futile
endeavor that only fools would attempt. So, let’s go do it.
OceanofPDF.com
Chapter 2
Shovel Knight / Verbs
DOI: 10.1201/9781003346586-2

Shovel Knight is a great game.

Shovel Knight came out in 2014 when indie games were just becoming a
thing. Shovel Knight is a great example of the retro movement, where
games are created to honor the classic games of old. Shovel Knight evokes
the feel of old-school 2D platforms like Super Mario Bros or Castlevania
but also adds its own unique twists.
This dissection is specific to Shovel Knight: Shovel of Hope, the original
game in the series (Figure 2.1). Sequels and spinoffs usually require their
own dissections, even if some aspects may be the same.

Figure 2.1 Shovel Knight screenshot.

Shovel Knight: Gameplay Dissection
Verbs
To understand a game, the best place to start is the verbs. What actions can
the player take?
Games are different from other forms of entertainment like books and
movies. The biggest difference is that in games, the player gets to do things.
The player can change what happens in the experience through their choices
and other actions. When you’re reading Hamlet (Shakespeare 1604), it’s
always going to end the same way, and Hamlet’s always going to drag his
feet to get there. When you’re playing a game of Go with a friend, you
don’t know who’s going to win or how they’re going to do it. That
combination of agency and uncertainty is what makes games fun. Games
are interactive.
So, when you’re trying to understand a game, start with questions about
interactivity. What does the player do? What actions can the player take?
There’s more to understand than just that, but it’s a good place to start.
In Shovel Knight, the player can move around and hit things with a
shovel. Movement is a combination of running and jumping. The core
gameplay is Run, Jump, and Attack. The player can complete 90% of the
game by just doing these three things. These are the core actions of this
game.
But how do we know that? Let’s dig a little deeper.

Advanced Verb-ing
One theme throughout this book is that game dissections depend on your
goals as a dissector. If you’re just trying to get a rough understanding of
Shovel Knight, understanding that the player’s actions are Run, Jump, and
Attack is enough. Stop there, move on. That’s a completely valid approach
and you can pat yourself on the back.
But sometimes you need to know more. If you’re going to start work on
the a 2d platformer team, or you’re going to make a game inspired by
Shovel Knight, or you’re writing a dissertation on it, you need to understand
player actions at a deeper level.
And sometimes the verbs may not be as obvious. If you’re not sure what
counts as a core action, you may need to spend a little more time analyzing
before you can say that with authority.
When you’re doing a deep analysis of a new game, it’s useful to start
with every verb you can think of. Go through the controller diagram. Think
about what you do when you play the game, especially the choices that you
deliberate over. Ask your friends “What do you do in this game?” Read the
wiki if there is one. Write it all down as a starting point.
Let’s do that for Shovel Knight:

Run
Run Right (right on the controller while on the ground)
Run Left (left on the controller while on the ground)
Attack (press the attack button, officially called “Dig Slash”)
Jump (jump button while on the ground)
Jump Right (jump button while moving right)
Jump Left (jump button while moving left)
Fall (what happens after you jump or are in the air for another reason)
Air Control (moving right/left while in the air)
Drop Attack (a combination of falling and attacking triggered by
pressing down while in the air, officially called “Shovel Drop”)
Dodge (move or jump to avoid an enemy or other hazard)
Fight (move and attack against an enemy, with a combination of
different actions mixed together)
Boss Fight (fight, but with more complicated things going on)
Wait (don’t do anything, sometimes important to get timing right)
Dig (attack a pile of dirt to get things out of it)
Break (attack an object like a chest or wall to destroy it)
Loot (technically pulling things out of a chest is a different action than
breaking the chest open)
Juggle (hit something that is falling, which sends it up again)
Die (lose health or fall in a pit)
Talk (press up while near a Non-Player Character (NPC) to initiate
dialog)
Buy (choose a text option to spend money or otherwise acquire a
thing)
Fish (use an inventory item to try to get stuff)
Use Relic (use an inventory item at the cost of Magic, with about a
dozen different ones that can be acquired)
Drink (use an inventory item to gain a bonus)
Play Music (select music via a conversation with an NPC)
Charge Slash (a modified attack earned later in the game)
Uncover (sometimes when you destroy something, there is another
object or enemy behind it that enters the scene)
Mini-Game (there are places where you can play little sub-games with
different goals)
Explore (look around and move around to find something new)
Find (basically the same as Explore, but possibly referring to a specific
thing you were looking for)
 Gather (pick up coins or other things, which happens naturally when
the character touches them)
Collect (usually used to mean the same thing as Gather but can also
refer to Feats or other more abstract things)
Plan (connect a series of actions in your mind and think about using
them together, possibly with specific timing)
Position (place yourself in the right place)
Observe (watch how the objects in the world work)
Learn (uncover the systems of the game and how they work)
Predict (use what you’ve learned to come up with a theory about what
will happen next)
Watch (cut scenes)
Travel (move on the map view that is used to select level)
Showdown (a separate play mode that I’m ignoring for this dissection)
Challenge Mode (another separate play mode)
New Game (UI action to start a game)
Enter Name (UI action to label a game save)
Body Swap (change avatar appearance)

You can see how this list can get pretty long. Let’s talk about how to turn
this big list into a smaller list that is more useful.
To start, let’s analyze these words. Which verbs relate to each other?
Which do you think of together when you’re playing?
 Pre-Verbs
If you’re dissecting an idea for a new game you’re thinking about, you won’t have a controller
diagram to get you started. Instead, you need to identify the verbs based on the information you
do have.

When you’re describing the idea to people, what actions do you talk about?
When you play the game in your head, what is it that you’re doing?
Picture the game on a controller. What would the buttons do?

This might be hard. The inspiration for a game idea can start with a story or a cool new feature
or other things. But if you find that your game idea doesn’t have clear verbs, that’s a sign that
your idea isn’t really complete yet.
Understanding the core loop is a key part of early ideation. You can’t pitch a game if you
can’t describe what the player is going to actually do in that game. If this happens, you need to
spend some more time thinking this through before moving on.

Direct Action Verbs

Since we’re trying to understand interactivity, we want to focus on verbs
that are about the player directly doing something. Look for verbs where the
player does something that sends input to the computer – this can be
pressing a button, pushing on a stick, touching a touchpad, or anything else.
When the player does this, the game responds. The game rules recognize
the player’s action as something important, and something in the game’s
model of the world changes (Figure 2.2). These are key to interactivity.
 Figure 2.2 Flow chart showing how the player does something,
foreshadowing bigger flow charts coming up soon.

Once we identify things the player can do, we can look for verbs that all
relate to the same type of action. Any player actions that result in the same
response from the game can be grouped together. That’s what we’re looking
for.
Looking over the big list, we can group a bunch of movement-related
verbs together. All of these are used by the player to change the position of
their avatar character.

Run
Run Right (right on the controller while on the ground)
Run Left (left on the controller while on the ground)
Jump (jump button while on the ground)
Jump Right (jump button while moving right)
Jump Left (jump button while moving left)
Air Control (moving right/left while in the air)

The important thing here is to recognize that the player controls an avatar in
a 2D space by changing the avatar’s position (and orientation) in the world.
There are a lot of different movement controls with slight nuanced
differences. The diversity of specific movement controls is a good sign that
navigating through the world is important in this game.
Each of these is important to understand if you’re really trying to dig
deep on how Shovel Knight works. Doubly so for other games with more
complex controls to navigate more complex 3D scenes. But for right now,
we can group them all together as “Move”. We’ll talk about more ways to
differentiate these types of overlapping verbs later.
 Deeper Dives: Research
When I start working on a game that fits into an existing genre or has a clear relationship with
another game, it’s important to start by doing a deep analysis of that related game.

Play the game thoroughly yourself and reflect on the experience.

Play related games.
Watch other people play the game. Social media can be great for this.
If it happens to be a game with an active fandom, read up on what they think. Try to build
the knowledge that a fan would have. There will always be fans with deeper knowledge
than you – some fans are great with details – and that’s OK.

The next set of verbs is combat related:

Attack (press the attack button, officially called “Dig Slash”)

Drop Attack (a combination of falling and attacking triggered by
pressing down while in the air, officially called “Shovel Drop”)
Charge Slash (a modified attack earned later in the game)
Dodge (move or jump to avoid an enemy or other hazard)
Fight (move and attack against an enemy, with a combination of
different actions mixed together)
Boss Fight (fight, but with more complicated things going on)

There are lots of ways to inflict damage on enemies. The player provides an
input and then the game rules decide if an enemy takes damage and how
much. I think we can comfortably group these verbs together and talk about
“Attack” as one of the big concepts in Shovel Knight.
The final set of direct player actions are all object-related. The player has
an inventory of special actions they can take (by pressing up and attack).
There are about a dozen different Relics the player can acquire over the
course of the game with various different effects, ranging from healing to
attacks to earning money. Most of these deplete the player’s Magic stat
when used.

Use Relic (use an inventory item at the cost of Magic, with about a
dozen different ones that can be acquired)

Indirect Action Verbs

Action Verbs that describe what the player can do are the most important
for understanding the core of a game, but there are a variety of other verbs
that happen during a game that are important to understand.

Die (lose health or fall in a pit)

Gather (pick up coins or other things, which happens naturally when
the character touches them)
Collect (usually used to mean the same thing as Gather, but can also
refer to Feats or other more abstract things)
Uncover (sometimes when you destroy something, there is another
object or enemy behind it that enters the scene)

There are verbs that happen in the game but are not directly triggered by the
player. These are more of a consequence of a player action than the action
itself. The player never presses “Collect” in Shovel Knight – they just walk
over to an object and if it’s the right type of object, like a gem, it just gets
collected. The player might do that intentionally, but the “do” the player
does is just movement.
There are also object actions the player can take based on the items in the
world around them. These use the existing controls but change the results
based on the objects in the world when the player performs their action.
These are important to understand but are less core than the actions.
 Talk (press up while near an NPC to initiate dialog)
Fall (what happens after you jump or are in the air for another reason)
Dig (attack a pile of dirt to get things out of it)
Break (attack an object like a chest or wall to destroy it)
Juggle (hit something that you’ve previously hit to keep it in the air)
Buy (choose a text option to spend money or otherwise acquire a
thing)
Play Music (select music via a conversation with an NPC)
Loot (technically pulling things out of a chest is a different action than
breaking the chest open)

The most common type of object interaction is talking with NPCs. I know
it’s a bit rude to call NPCs objects, but from the perspective of the code
there’s not much difference.
Technically you could call combat an indirect action because it only
makes sense if there’s an enemy nearby and there isn’t always an enemy
nearby. This is a clever point, but given the frequency of enemies in the
level design, I think it’s fair to call “Attack” a core gameplay action. If
someone re-made Shovel Knight with significantly fewer enemies and more
depth to the NPC interactions, it would be an interesting way to make an
entirely different game in a different genre without changing the core
engine or code or player actions of Shovel Knight. But it wouldn’t be Shovel
Knight. More on content and how it changes the game in a later chapter.

Abstract Verbs
Dodge (move or jump to avoid an enemy or other hazards)
Explore (find something new)
 Find (basically the same as Explore, but possibly referring to a specific
thing you were looking for)
Plan (connect a series of actions in your mind and think about using
them together, possibly with specific timing)
Position (place yourself in the right place)
Observe (watch how the objects in the world work)
Learn (uncover the systems of the game and how they work)
Predict (use what you’ve learned to come up with a theory about what
will happen next)
Mini-Game (there are places where you can play little sub-games with
different goals)

There are a number of potential verbs we identified that aren’t things the
player can do. These are things the player thinks about when doing other
things – these are why the player does things, not what the player does. So,
these are important to consider, but I would categorize these as goals, not
actions. These are incredibly important to understand and are fundamental
to understanding the core game loop, but I’d put them in a different
category than direct player actions, and we’ll cover them in a later chapter.

Interface Verbs
Watch (cut scenes)
Travel (move on the map view that is used to select level)
Showdown (a separate play mode that I’m mostly ignoring for this
dissection)
Challenge Mode (another separate play mode)
New Game (UI action to start a game)
Enter Name (UI action to label a game save)
 Body Swap (change avatar appearance)

There are a number of actions that the player takes entirely in the UI (User
Interface), outside of core gameplay. In a game like Shovel Knight, it’s easy
to distinguish when the player is controlling an avatar (Shovel Knight) in
the game world versus when the player is selecting words in a menu. The
UI is something outside the illusion of the game, which is necessary but not
part of the fun. If you’re working on designing the UI, definitely think these
through and consider all the options. But for this game, they are not core
gameplay.

Deeper Dives: Frame by Frame

For fast-paced games, one technique I find very useful is to record a short video of the core
gameplay, then step through the video frame by frame. Watch closely, and you’ll see lots of little
things that the developer does to make things work. Visual effects that communicate important
information to the player, little tricks of physics to make things flow better, and complex camera
behaviors so the player’s attention is on the right things. These are great when you need to know
everything about a game. But most of this chapter is for when you’re just starting to understand
a game.

Games as Input/Output
To help understand why these different types of verbs are different, let’s talk
about how games work. As with most things, games can be understood in a
lot of ways. This is one of those ways.
When a player plays a video game, they are participating in an exchange
with a computer. The player provides inputs through a device of some sort –
a game controller, a keyboard, a touch screen – and the computer processes
that inputs and then provides an output for the player – pixels on a screen.
The computer uses a set of rules and logic and code to determine how the
player’s input changes the pixels. The rules and logic and code are what the
game designer controls. That is what we call the game. But everything else
is important to understand as it creates the full experience.
This can be represented in a diagram like Figure 2.3.

Figure 2.3 Input Output.

The left side of this image represents the player. What they see, what they
think, what they do. The right side of this image represents the computer.
All the electrical impulses tickle the microchips based on the code of the
game.
Let’s step through a simple single-game action based on this model. The
player is enjoying a pleasant time in Shovel Knight (Figure 2.4). They
notice some gems above and to the left, so they decide to jump to gather
them.
 Figure 2.4 Shovel Knight game screen showing the moment
before the player makes a decision.

We start by noting that the player has made a decision (Figure 2.5). This
is how things start in games – things happen in the player’s brain. The root
of all gameplay is imagination.

Figure 2.5 I want to jump.

 In order for the game to do anything, this thought needs to get from the
player’s brain to the computer (Figure 2.6). There are all sorts of useful
input devices designed to make this happen. The nature of the input device
has a large impact on the player’s experience, and new types of input
devices like the iPhone’s touch screen or the Wii’s motion controller can
lead to huge changes in the video game industry.

Figure 2.6 Press the button.

For our current discussion, it doesn’t really matter what input device is
being used. The important thing is that the player presses a button on the PC
or game controller that indicates that the player wants the character (Shovel
Knight) to jump.
This information then flows into the computer (Figure 2.7). A series of
electrical impulses flow from the controller through a wire into the
computer, where chips translate the signals into the code.
 Figure 2.7 Electrons flow.

The code then processes the new information (Figure 2.8). The code
knows that this sequence of electrical impulses means that this button was
pressed. And this button means jumping. And jumping means many things:

The position of the avatar character changes, moving them upward.

Most games have a physics engine that controls positions and motion,
so this is done by adding a force in a direction to the character and
letting the physics logic compute the exact change in position.
Because of the physics engine, gravity starts operating on the
character, changing their motion over time.
The character’s animation changes from a standing animation to a
jumping animation.
This changes the character’s state – the character is now in the air.
Now different inputs and outputs are possible. If the player presses
down now, the character will go into a downward attack state, as
defined by the game rules written into the code. Pressing the jump
button again won’t do anything.
 And all of this needs to be reflected in the pixels on the screen.

Figure 2.8 Character enters jump state.

This last point is important. The state changes in the code only matter if
the player can understand them. Most video games can only communicate
their internal changes through a screen. Any change to the internal state and
logic of the game needs to be expressed in pixels so the player can process
it (Figure 2.9). Setting up the right pixels is a combination of game art to
reflect changes in the game world and UI/UX design to reflect changes in
the data presented by the game. Both are important and ideally both work
together to produce the best result.
 Figure 2.9 Pixels change.

The player has their own complex system of input devices that they use
to process incoming information – eyes, ears, noses, etc. Good games will
reinforce their messages using as many senses as possible (Figure 2.10).
Starting with visual, often using audio, and occasionally adding some
vibration so the player can feel what is happening. Smell and taste don’t
come up as often, but I’m sure smell-o-vision will be everywhere in a few
years.
 Figure 2.10 Player notices change.

And then the cycle begins anew. Once the player processes this new
information, they want new things (Figure 2.11). For example, that frog
might be a problem, so the player may need to change their direction or
enter an attack state to avoid taking damage (Figure 2.12). The full game
experience is a long sequence of very quick input/output cycles that add up
to create something greater than each part.
 Figure 2.11 Player wants new thing.

Figure 2.12 Screenshot of Shovel Knight, showing a jump has

begun.

Verbed Input/Output
This Input/Output flow gives us a way to understand the differences in
different types of verbs. Each verb operates on a specific step of the
diagram as seen in Figure 2.13.
 Figure 2.13 Input verbs and output goals.

The Direct Action verbs are controller inputs delivered by the player.
The Indirect Action verbs are things that happen in the computer based
on the player actions.
The Abstract verbs are going on in the player’s head, helping them
plan out their next action but not happening as a direct result of a
controller input.
Interface verbs affect things the player sees in the Game Display that
are not part of the game world.

This helps explain why the Direct Action verbs are the most important for
an initial understanding of the game. They are the simplest and most clear
actions that shape the range of options for everything else. They are where
the player that exists in the real world interacts with the game. The game
rules only decide to break things after the player uses the attack action. And
the player’s desires are similarly shaped by the range of inputs the game
allows. The player may want to build a bridge or get a drink at the bar but
without a related button that desire is going to go unfulfilled.
 The actions the player can take shape everything about the game. So,
start there if you want to understand the game.

Understanding Verbs
Verbs are just the first step.
If Shovel Knight is about Moving and Attacking, so is Halo (2001). So is
Super Mario Bros (1985). So is God of War (2005). But these are all very
different games.
The second step in a game dissection is to understand the nature of the
verbs and how they fit together. We’ll talk more about systems later, but the
core ideas of systems thinking is that the relationships between things (in
this case verbs) can be as complex and interesting as the things themselves.
So, let’s start with our two core verbs.

Overlapping Verbs with Different Rules

So, we’ve broken Shovel Knight into the Core Verbs of “Move” and
“Attack” but want to understand more about them.
Going back to our definition of games in the first chapter can help. We’ve
already covered Interactivity – the choices and skills that make up the Big
Verbs. Next on the list is Rules.
So, our Verbs are what the player can do. The Rules are how the game
responds to those prompts from the player (Figure 2.14). “When the player
does X, the game does Y”
 Figure 2.14 Outputs.

In a board game, these are written out explicitly. “When you pass GO,
collect $200”. A core verb in Monopoly (1935) is move, so “pass GO” is
something that can happen naturally as a result of the player’s action.
The same sort of thing happens in a video game like Shovel Knight.
When you touch an enemy, you lose health. When you walk over a health
pack, there are fewer health packs in the level and you have more health.
The player does something, and the game responds.
Moving in Shovel Knight is about position, and position interacts with a
lot of game rules.

Where you are relative to an enemy determines whether or not you get
hit by an enemy attack.
Where you are and where you’re pointing determines if your attack
hits the enemy when you trigger an attack.
Where you are determines when you pick up gems, powerups, or other
items.
Where you are determines what you can see – the game only shows
you one screen at a time, based on player position.
 Where you are triggers progression through levels, including triggering
cut scenes that advance the story.

This is similar in a 3D first-person shooter but with some important

differences:

Where you are determines whether or not you get hit by an enemy
attack.
Where you are and where you’re pointing determines if your attack
hits the enemy.
Where you are determines when you pick up ammo, weapons,
powerups, or other items.
Where you are and where you’re pointing determines what you can
see, using a first-person perspective.
Where you are triggers progression through levels, including triggering
cut scenes that advance the story.

So, one way you can differentiate similar verbs is by looking at the rules
and consequences of those verbs. When there are mechanic differences,
those are indicators that the verbs may sound the same but are functionally
very different.

Overlapping Verbs with Different Goals

Looking at how verbs interact with rules is helpful. The processes going on
in the computer can change how the verbs work. But the processes going on
in the player’s head are even more important.

Why does the player use verbs?

Verbs are used to achieve goals.
 Verbs only mean something in the context of the player’s goals.

What are the player’s goals in Shovel Knight? Let’s do a quick brainstorm
like we did for verbs.

Rescue Shield Knight.

Help the characters you meet.
Defeat enemies.
Reach the end of the level.
Trigger the next cut scene.
Unlock new areas.
Get to the end of the game.
Don’t die.

Some of these (“Rescue Shield Knight” and “Help the characters you
meet”) are story, not gameplay. Story is an important part of most games,
but when we’re still assembling a core loop, it’s not the most important
thing to understand. Story motivates players to complete the game, it
doesn’t motivate the players’ actions from moment to moment. At least for
a game like this – some games like Baldur’s Gate (1998) or Life Is Strange
(2015) make story a key part of every level of gameplay. But a game like
Shovel Knight does not weave the story into every moment of gameplay
like this.
The goals that are woven into every moment of gameplay are the ones
like “Defeat enemies” or “Reach the end of the level” that guide player
actions in every moment. This is what the player is trying to do during
gameplay, and this is why the player engages with the verbs of the game.
The player attacks to defeat enemies and moves to reach the end of the
level.
 The next few goals (“Trigger the next cut scene”, “Unlock new areas”,
and “Get to the end of the game”) are just “Beat the level” phrased slightly
differently. When you’re considering goals or verbs, keep an eye out for
overlapping terms like this.
The final objective “Don’t die” isn’t really a goal per se but is more of a
fail case. If the player dies, they have to reset a little ways back. Death
never stops you from progressing, but it can definitely slow you down.

Verb to the Goal

So how does the player turn “Move” and “Attack” into completed levels?
They move through the level to find a route that gets to the next area while
avoiding or attacking any threats. Most areas have a little challenge or
puzzle to advance. Some areas require jumping over pits. Some require
jumping off objects to get up high. Each area has a different combination of
challenges that makes it interesting and different. More on that later.

Big Verbs
OK, so we’ve defined the big verbs in Shovel Knight as Move and Attack.
That feels right. When I play the game, I spend most of my time doing
those things. And thinking about those two things and how to do them
better.
If you want to understand a game, understand the things the player
spends most of their time doing. Sometimes there are other features that are
splashy and interesting and become the focus of the marketing and
discussion around the game. But a game designer needs to think about the
thing the player is actually investing their time on.
 You can see this in lots of different types of games. Platformers spend a
lot of time perfecting their jumping and running. In an open world game,
you spend most of your time moving from place to place. The open world
games that succeed are the ones that this travel fun and engaging.

Verb as Genre
As we’ll see in future chapters, verbs are a good way to differentiate and organize games. All
first-person shooters start with move and attack. Platformers like Mario focus heavily on move,
with varying amounts of attack depending on the platformer.
But a game like SimCity doesn’t have move as a verb at all – the primary verb there is build.
This makes sense – I think most people would agree that platformers have more overlap with
shooters than with builders.
This can be useful to know. If you’re looking for inspiration on your city builder, it might not
be best to start by playing some Halo. Maybe find a puzzle game like Triple Town that shares
“Build” as a verb to see how it’s done differently there.

More to Come
We’re going to stop our Shovel Knight gameplay analysis here. I know
we’ve barely scratched the surface of Shovel Knight and platformers in
general. But we’re building foundations here. It’ll get crazier in later
chapters.
The last part of this analysis pulls some ideas from future chapters, but I
like showing a version of the same diagram for each game, so forgive me
for the spoilers. This diagram represents each game as a formula. The
player has verbs they can use to reach goals but are blocked by challenges.
Figure 2.15 shows the simplest level of detail, breaking each step into a
single element. Figure 2.16 is a bit more useful in that it breaks each of
these categories into multiple entries, showing the different actions and
challenges of the game:
 Figure 2.15 Shovel Knight flow diagram showing Verbs,
Challenges, and Goals.

Figure 2.16 Shovel Knight flow diagram with more boxes.

Each level is a package of these different pieces, as shown in Figure 2.17.

In Shovel Knight, most of the variation comes from changing out the
challenges, but other games can handle this differently.
 Figure 2.17 Shovel Knight flow as a level.

Multiple levels can be added together to represent the full flow of the
game, as players work on larger goals and larger challenges. See Figure
2.18.

Figure 2.18 Shovel Knight flow from level to level.

We’ll dig into these ideas in more detail in later chapters, but before we
leave Shovel Knight, let’s talk about other things that influence gameplay.
Shovel Knight: Non-Gameplay Dissection
When dissecting, it’s not just about gameplay. Even for a game designer.
If you want to understand the gameplay of a game, you have to
understand everything about it. How do the non-gameplay aspects of the
game tie into the gameplay? Where did the gameplay ideas come from?
What led to the gameplay decisions you see in the game? The game itself
makes much more sense if you understand the context that led to those
decisions.
To be clear, I did not work on (most of) the games I talk about in this
book. I do not know how decisions were made by the teams that did. We’re
talking about these things as an outsider, piecing things together from what
we can see in the game and generally available knowledge. I’m sure there
are interesting stories there, and I encourage those who were present to
write them down so I can read those books. I love to know authorial intent,
but unless that’s you, the rest of us are left with the work itself. And there’s
plenty to see there.

Senses
Gameplay is influenced by the look and feel of the game.
Shovel Knight has a very clear and distinct style. It’s a retro game trying
to emulate the feel and visuals and sounds and complete experience of a
classic platformer. In this sense, the style fits very clearly with the
gameplay. People who are familiar with the references will expect the
gameplay and style to both align with their perception of those earlier
games. And people who don’t know the classics will still understand that
both gameplay and style evoke an earlier age, so don’t expect integrated AI
and vast open worlds and high-fidelity graphics.
 The simplicity of the game also makes it easy to see where gameplay and
graphics intersect. Remember our input/output diagram – the output part of
the diagram is made up of pixels on a screen that the computer manipulates.
This is the player’s only source of information on what is going on in the
game (well, that plus sound and sometimes vibration). As a game designer,
these pixels are all you have to get your ideas across to the player. A player
who might not really be paying attention and who most likely hasn’t played
every previous game that you (the game designer) did. This is a difficult
and momentous responsibility for a few measly pixels.
Every game has to create a visual vocabulary of what matters to the
player and to gameplay – using the same size, shape, color, or whatever
consistently helps the player learn what to do and then quickly process new
scenes. Shovel Knight does a great job of quickly establishing what the
player needs to know and sticks with that throughout the game:

The ground you walk on is flat – usually rectangles – and high contrast
against the rest of the background.
Objects you can interact with are also high contrast and often include
movement. The game intentionally doesn’t create separate
vocabularies for enemies and rewards, as it wants to play with that in
the gameplay. Sometimes bouncing on an enemy’s head is the only
way to reach a reward. Sometimes a reward is in a dangerous place
making it more of a problem than a pure benefit.
Combat actions and results are always clearly punctuated. Enemies
don’t just disappear – they vanish in a little cloud of smoke. If the
player missed the frame where they died, the game is still
communicating for a few frames after.
Shovel Knight’s shovel provides a nice strong indicator of where the
character is facing. This is especially handy for communicating the
 difference between “falling” and “falling with intent to kill”. The
downward shovel attack is important and might not be noticed without
the big bright object that moves around the character.

History and Influences

I’m a big fan of stealing ideas from other games. No game is an island.
Every game is inspired by the ideas and execution of games (and other
things) that came before.
There’s an important distinction to be made here. All games are inspired.
All games borrow and “steal”. But when you draw inspiration from
something, you need to be careful to not draw too heavily. If your game is
just a copy of someone else’s idea, that’s actual stealing, not the tongue-in-
cheek way I sometimes use the term. If your game gets all its features and
ideas from another game and doesn’t add or change them in a way to make
something new, then you need to rethink and start over to add your own
twist on the thing.
You can see this in Shovel Knight. Shovel Knight is a wonderful and
unique game. But almost every aspect of Shovel Knight comes from
somewhere else.

Super Mario Bros (1985) is generally credited with popularizing the

2D side-scrolling platformer genre.
But many of the ideas from Super Mario Bros were first developed in
the Mario Bros (1983) arcade game, and borrowed from earlier games
like Space Panic (1980).
Many 2D side-scrollers after Mario added attacks like Shovel Knight’s
shovel, including popular games like Metroid (1986) and Castlevania
(1986).
 Once the seeds of the genre had been planted there are many great
games that took those ideas and adapted them for PC gaming,
including Lode Runner (1983), Commander Keen (1990), and Cave
Story (2004).
Braid (2008) and Spelunky (2008) popularized the idea of using 2D
platformers as a simple and nostalgic base to experiment with
interesting indie game ideas.

As a game designer, it’s important to know the history of not just the games
but also the features. Jumping from platform to platform in Shovel Knight is
interesting but is informed by many other games before, and feeds into
many other games after. Dissecting each of those games, at least in terms of
that specific feature, helps you see how to use that feature in your own
games. Johnnemann Nordhagen took this idea even further and built a game
that collects all the different ways that a specific mechanic has been
executed over the years. Museum of Mechanics: Lockpicking (Dim Bulb
Games, 2022) is an interesting way to see how games borrow and share and
innovate on specific ideas.

Monetization
How a game makes money has a huge effect on the design of the game.
There are a few major trends in monetization that had huge impacts on
the types of games we associate with each era:

1. Arcade games. Monetization is quarter-based. (Technically

monetization for the developer is by selling arcade cabinets, but the
cabinet buyers were very interested in quarter throughput, so the
developers developed to that metric.) The goal was to get the player to
notice the game in a crowded arcade, and then plug in as many
 quarters as possible in as little time as possible. This led to game with
very quick session lengths, very steep difficulty curves, and little to no
thought about ongoing long-term play.
2. Console games. Monetization is based on selling boxes in stores, with
no internet to provide updates or live services after that. At first early
consoles tried to mimic games that players knew from the arcade, but
soon the focus became value. Games were reviewed based on hours of
play. Look at the difference between Mario Bros and Super Mario
Bros to see a clear example of this change.
3. Mobile games. Monetization is based on downloads and (after the first
few years) in-app purchases. The goal is to get a player in quickly and
easily, then keep them around for long enough that they get
comfortable spending money. The growth of big data plays a big part
here, as game developers start to analyze what actually gets people to
stick around and spend money. Game sessions are short, but players
are expected to stick around for months or years.

This is not an exhaustive list by any means. Early PC games resembled

console games but with a lot of their own twists. Console has been
changing with the growth of downloadable upgrades. Most of this doesn’t
apply to educational games or other games for change, which have entirely
separate monetization goals and pathways. Each game’s monetization needs
to be analyzed separately but with an understanding of how it fits into the
standard monetization schemes of its time.
Shovel Knight makes most of its money through downloads on PC game
services like Steam. This is currently the dominant sales platform on PC,
and the growth of independent routes to sell and market games without
interacting with big publishers and retailers is a big part of why indie games
like Shovel Knight exist and succeed. A successful Steam game can support
a small indie team with a good idea, but console games generally have to be
bigger.
Shovel Knight started out on Kickstarter, raising money for the initial
development through crowdfunding. Kickstarter and similar platforms are a
popular way for smaller indie games to generate some initial funds and
market an early idea. The initial money helps, but the initial fan base is as
or more important to making a small game with a small (or zero) marketing
budget a success.
Since its release, Shovel Knight has shipped multiple downloadable
content (DLC) packs, which have later been released as separate games.
Shovel Knight has become a series with explorations in multiple genres and
different playstyles.
Selling as a single purchase drives Shovel Knight to focus on providing a
complete package. Deeper progression and varied content make the single
purchase feel worth it, generating good word-of-mouth to keep the game
selling.

Narrative
Story and gameplay are friends. They work together to bring joy and
happiness to players. Different games make more or less use of this
friendship. Shovel Knight is a great example of the most common use of
story in games.
In Shovel Knight, the main use of story is to provide emotional context
for the player’s progression and emotional punctuation at key points. The
game establishes early on that Shovel Knight is trying to rescue their friend
Shield Knight. A classic rescue mission, which is a great way to make a
player want to reach the end. My friend needs help!
 Most of the time, this is just a small piece of background motivation.
During gameplay, the game doesn’t make reference to Shield Knight or
anything else about the characters. Bosses get a quick intro to establish their
character, during which the player has no input. The game stops to tell the
story. Then after a big level, there is a short dream sequence that reminds
the player of this long-term goal and provides some interactive rewards as
well. Some of the NPCs provide additional narration and backstory, but it’s
all contained in a few sentences so it never gets too involved or deep.
This is great. This is exactly what most games want. Story is there.
Emotions are triggered. The player knows why they’re doing all this, and if
the player puts the game down for too long, they feel a little guilty. Poor
Shield Knight. The story is fun and interesting, and the game is stronger
because of it.
Some games use story in deeper and more complicated ways. And I love
those games, too. But I’ve made some of them so I know how complicated
and difficult it can be to interweave story and gameplay in ways that
actually matter. More on that later.

Emotions
Story is one source of emotions, but games can evoke many emotions.
For action games, the most common emotion is frustration, which might
sound like a problem, but it really isn’t. Talking about games for the last
few decades, I’ve noticed that there are many flavors of frustration. Some
are good, and some are bad. The main difference is how and where the
player has agency in the frustration.

Outside Frustration: If the player has no agency, then it’s outside

frustration. It can be annoying, but it’s less personal so thus less
impactful. I can get frustrated watching a communication problem
unfold in an episode of a sitcom, but I don’t feel it personally.
Uncontrolled Frustration: Once the player has control of the action,
the potential for emotional involvement is higher. Give me buttons to
press and now any failure is my fault. So, the worst case here is when
the player has input but no real control. If the player expects to have
control but loses due to reasons outside of their control, that can be
very frustrating. The game made a promise and then failed to deliver.
This is often the worst form of frustration and game designer should be
careful to avoid situations like this. If the player needs to fail
regardless of their actions, make it very clear that the game knows it’s
not the player’s fault.
Confused Frustration: If the player has agency that actually matters,
that’s going to reduce the negative impact of the frustration. But this is
only true if the player understands all that. If the player doesn’t
understand how their actions influence the outcome, they might as well
not have agency. This is where many games fall apart, especially for
more casual audiences. It’s very frustrating to press buttons and not
understand why they matter. And that understanding is entirely the
responsibility of the game designer. Remember, the player only knows
what comes through that tiny funnel of information they see on the
pixels on the screen. If they aren’t able to construct a full mental model
of the logic going on inside the computer, to expect, they’re going to
have some negative frustration. In many cases, the hard part of being a
game designer isn’t coming up with good ideas; it’s getting the player
to understand those good ideas.
Difficult Frustration: Agency that matters and makes sense is not
enough. The challenges also have to be fair. Banging your head against
 an impossible challenge can be a positive type of frustration (cough
Elden Ring cough), but not for all players. Some players want more
manageable challenges and find extreme difficulty to be bad
frustration. But the inverse is also true – a player who enjoys mastering
a skill may consider a sequence of easy challenges to be frustrating in
a very different way. Finding the right balance for the audience of your
game can be difficult – more on that later.
Fun Frustration: Finally, in some cases, players have agency,
understand the consequences of their actions, and still fail. This is
called gameplay. If the player understands why they failed, the
frustration from the failure is much less. In the best cases, frustration is
a fuel that drives the player to come back and try again. Games are
built on iteration, both as a development practice and as a player
strategy. A good game will create positive feelings of frustration that
guide the player to an emotional response along the lines of “I can do
better”. Which is a great lesson to learn for life, even outside of the
magic circle.

Meaning
I think it’s important to understand what games mean. Since games are a
relatively new form of art, we don’t always think of them as having an
impact on the world. But they do. The person who starts playing a game is
different from the person after the game is done. Games make people think
about things. Games change the world.
For many games, this is not the primary goal of the game. It might be
money. It might be providing a simple fun time. But meaning happens
whether the creator intends it or not. I understand programming better
because I played Robo-Rally. Some of my love of creativity and creative
problem-solving comes from early Infocom text adventures. Minecraft has
made a generation of kids think differently about creation and resources and
exploration and problem-solving.
I don’t think any game has one discrete meaning. Even games with a
clear intended meaning will be interpreted differently by each player who
encounters it. So, when I talk about meaning in the games in this book, I
don’t intend to exclude other meanings. When we talk about this in class,
every semester has different answers and they’re all right.
For me, Shovel Knight is about persistence. It’s not the only game where
the main character is beaten down constantly and has to constantly get up,
but it’s a clear example of that style of play. I’m not the most skilled player,
so I tend to die a lot. But the only time I feel like I’m losing is when I put
the game away and give up on saving poor Shield Knight.

OceanofPDF.com
Chapter 3
Candy Crush Saga / Goals
DOI: 10.1201/9781003346586-3

Candy Crush Saga

Candy Crush Saga is a great game.
Candy Crush Saga (Figure 3.1) is a mobile game released by King (now
part of Microsoft) in 2012. Since launch, it has consistently been one of the
most popular games on mobile, with hundreds of thousands of monthly
users Curry, D. (2024).
 Figure 3.1 Candy Crush Saga screen.

I’ve never worked on Candy Crush Saga, but I did work on a competitor.
When I was at a studio called Zindagi, we designed a match-3 called Crazy
Cake Swap based on a similar puzzle game called Crazy Kitchen. The
studio got bought by Zynga and didn’t stay around for long after that, but I
know this genre better than I know most genres.

Candy Crush Saga – Gameplay Dissection

Verbs
To understand Candy Crush Saga, let’s start where we always do: verbs.
What can the player do in Candy Crush Saga?
In this case, it’s a simple answer: swap. The player can touch a candy in
the grid and then slide it to an adjacent candy. When the player does this,
the two candies swap positions. That is the entirety of the player’s active
contribution to the core loop (Figure 3.2). There are some interesting less
active things the player does while swapping, such as “Scan” and “Predict”,
but these aren’t the simple button actions, so we’ll save discussion of those
for later in this chapter.

Figure 3.2 Inputs.

 The core loop continues from there, though (Figure 3.3). After the player
makes a swap, the game rules check to see if either candy creates a line of
three or more. If it doesn’t, the swap is canceled. If it does, the game
removes those candies, possibly with an extra effect if the line is four or
more candies. Then the remaining candies fall to fill the empty space, which
can possibly create new matches and cause a chain reaction as each match
creates new matches.

Figure 3.3 Outputs.

The player performs a simple action, and the game rules result in
complex outcomes. This is one of the strengths of video games – the
processing burden of the game can be placed on the computer’s shoulders.
The rules of Candy Crush would be much less fun as a board game where
the players have to move each candy by hand and figure out the outcome of
each new position.
Swapping and the reactions to that swap is the heart of Candy Crush
Saga, and the core loop of any match-3 game.
And for completeness, Candy Crush Saga also has a few special case
verbs:
 1. Players can use powerups such as the lollipop to remove specific
candies or otherwise affect the grid outside of swap actions. Outside of
the tutorials that teach these interactions, they are never required to
complete a level, but they are always there for the player to use. But
they are a limited resource that can only be acquired as an occasional
reward or for money.
2. Late in the game, there are new board elements that include an
interactive choice by the player. For example, Candy Frogs let players
choose where to place them when they go off.
3. Secondary modes like events provide a choice of what to play and
sometimes have some basic decision-making in the event itself.

So outside of these caveats, the verb of Candy Crush Saga is “swap”.

Goals
The gameplay goal of Candy Crush Saga varies based on the level. All
goals revolve around clearing things off the board, but the nuances of
exactly what you’re trying to do can change:

1. Clear a certain number of candies.

2. Clear a certain number of candies of a particular type.
3. Clear a certain number of special objects, like cherries, by dropping
them to the bottom of the grid.
4. Remove a certain number of non-candy obstacles, such as jellies.
5. Etc.

More on this in a moment.

Challenge
The real variety in Candy Crush Saga (and most match-3 games) comes
from the randomness of the falling pieces and the variety of obstacles that
make reaching the goal difficult. Most levels have more than just candies on
the board – a large variety of different objects create additional challenges
and opportunities for the player. These range from simple inert objects that
are removed when matches are made near them (Frosting), to inert objects
that expand across the grid over time (Chocolate), to snakes that charge into
baskets after enough matches are made next to them (Candy Cobra). These
aren’t all dangerous – there are also advanced board elements that can be
useful, such as conveyor belts, or are almost purely beneficial, such as the
UFO that drops special candies.
These obstacles are key to Candy Crush Saga’s success. The basic
concepts of the match-3 genre existed before Candy Crush Saga.
Specifically, the concepts of the genre were solidified in Bejeweled (2001)
and Bejeweled Blitz (2009). Many of the core ideas were present in earlier
games like Shariki (1994). The core gameplay is very directly based on
previous games like Bejeweled. But those previous games used only a
single level – there were no obstacles and just one board. Candy Crush
Saga innovated by adding a variety of levels with different obstacles and a
different grid layout for each level. This gave the existing fun core
gameplay more variety and long-term appeal. Bejeweled was successful, but
Candy Crush Saga took that success to vast new heights by adding varied
challenges and a stronger sense of progression.
The sense of progression comes from the “Saga” in the title. One of the
big innovations of Candy Crush Saga was the Saga Map. It’s not just that
the player was given a variety of levels with different obstacles, it’s also
important that these levels were set up as a line – a path to a destination.
Once the player completes a level, they move on to the next one, heading
toward the end. The Saga Map organizes this progression in a simple line,
showing the player the next levels they’ll be playing. This came out when
social gaming was king, so the line also shows you where your friends are,
creating some social pressure to advance. But the key concept here is that
the player is always doing something new, and has a sense of progression
that makes them want to come back tomorrow and try again. Progression is
a great driver for players, in any form.
The basic flow of the game is simple – the player swaps to clear the
level, with various obstacles in the way. But the obstacles and goals change
in each level, as shown in Figure 3.4, making the simple flow work for a
long time.

Figure 3.4 Candy Crush Saga flow: Simple diagram showing

Actions, Challenges, and Goals, with lots of variation on the
Obstacles and Goals.
Candy Crush Saga – Non-Gameplay
Dissection
Technology
Candy Crush is a fairly simple game that relies on a vast network of
advanced technologies to work. This is one of the fun daily realities of the
world we live in. The rules of the game and the interactions the player
experiences are built on easy-to-understand combinations of simple shapes.
This is good game design and keeps the player engaged. But the rest of the
experience is built on touch screens and constant online communication and
vast behind-the-scenes databases that are way more powerful than the
technology that brought humans to the moon.
One thing that is important to think about as a game designer is how
these vast changes in the world of technology affect your game. New
technology comes along all the time, and some of it is a great fit for new
types of games. Sometimes there is a direct linkage between a new
technology and a new style of game, such as GPS-enabling Pokémon Go
(2016). Sometimes the changes are more subtle and need time and thought
to understand. It’s not obvious (except in retrospect) that improvements in
computer miniaturization and wireless communication would lead to
everyone having a powerful connected computer in their pocket at all times.
And even then, it’s less obvious that it would lead to people looking for
games that they can pick up and play at any time for very short play
sessions. Technology doesn’t only change what people can do with games,
it changes what they want to do and when they want to do it and for how
long. The changes to the rules of technology create changes to the dynamics
of how people interact with technology, but more on the word “dynamics”
in Chapter 4. These changes can be harder to predict but provide great value
for your game designs once you notice them.

Senses
Candy Crush Saga looks cartoony and childish. Everything is in bold,
bright colors. The candies are simple and iconic. The game is not subtle in
its communication; if it wants you to know something, it will shout it rather
than whisper it. At a glance, you might guess that this is a game for
children. But the majority of players are adults.
Clearly, this works for Candy Crush Saga. The game is immensely
successful, and there’s no reason to think that the bright cartoony style
holds it back. Various clones and imitators have tried to re-theme match-3
gameplay in darker, more adult styles and failed. Childish works for adults.
On one level, this could be pure nostalgia. Adults want a moment to relax
and escape from their everyday world, so something that evokes pleasant
childhood memories could be just the thing. The reason childish works
might be because adults want to feel like children.
On another level, childish might work because childish is good design.
Kids are not the only ones who like bold, bright colors. Iconic shapes are
easy to distinguish and help make gameplay easy to understand. Direct
clear communication is good for everyone, including adults who are trying
to just relax. The design principles that work for children also work for
adults. It’s not childish design, it’s just good design.
This is not to say that Candy Crush Saga is constructed in a childish way.
There is a lot of attention to detail and polish in Candy Crush Saga. When I
started working on match-3 games, one of the first things I did was spend a
week analyzing some of the top games, and some of the worst games. I
recorded a few levels of Candy Crush Saga and then literally watched the
recording frame by frame to see exactly how everything moved and all the
little tricks of presentation. It’s a very useful exercise if you’re trying to
understand polish and presentation. Make sure you do the same with some
less successful games to really see the differences.

Freeze Frame
Things I saw by watching every frame of Candy Crush Saga:

Be really careful with motion. Humans know how falling looks and expect things to
accelerate and decelerate in certain ways. Falling at a constant speed is easier to code but
looks weird.
You can cover up lots of things with explosions. Don’t bother with expensive
transformations if an explosion will feel just as good or better.
Squash and stretch. This is an animation term for changing the shape of an object to
emphasize an impact. It really works.
Anticipation. Building up to a reward is as important as paying off the reward. Adding a
few frames of buildup makes almost everything better.
Sequencing. Know the order of events and keep that holy. It’ll help players understand
what’s happening and it’ll keep your game logic and code understandable.
Compression. When you have lots of things that need to happen at once, you can break
the Sequencing rule and have things overlap a bit. But do it in a logical and constrained
way. Think through what the player needs to focus on, and do some hand-waving in other
places.

Emotion and Narrative

Candy Crush Saga is not a deeply emotional game. There is no complex
storyline to follow with deeply realized characters following narrative arcs
who the player grows to care about. There is a simple story that presents a
problem to solve, and then a few levels later congratulates you for solving
it. This is enough story to give you a reason to continue, but Candy Crush
Saga is not a game where the majority of players are advancing to find out
what happens next.
 Candy Crush Saga does hit emotional notes through the gameplay.
Candy Crush Saga is a game of constant rewards. Whenever you do
anything in the game, something explodes and the game tells you how great
you are. This is maybe not the most emotionally deep reward system, and I
wouldn’t trust the game’s praise to evaluate my self-worth, but sometimes
it’s just nice to hear praise. Constant rewards can be emotionally soothing.
A working mom who is having a hard day and just wants to catch their
breath and relax doesn’t need a game that constantly tells them they’ve died
and need to start everything over. Candy Crush Saga does have fail states,
and it does apply a little bit of emotional pressure in those moments (mostly
to encourage the player to spend money), but that’s the exception, not the
rule. Most of the time, playing Candy Crush Saga is soaking in a warm bed
of reward and praise. Some games are here to give people what they want.

Money
How does Candy Crush Saga make money?
The other big innovation Candy Crush Saga added to the match-3 genre,
other than the level variety demonstrated in the “Saga” part of the title, is
the monetization model. Candy Crush Saga is free to download and to play.
All ten thousand plus levels are free to access. But the game constantly asks
you to spend money on other things. Powerups. Unlimited lives. Extra
turns. Special modes.
Free-to-play business models drive most of the sales in the games
industry these days. And they work. Most of the players in these games pay
nothing. A small percentage pay something. And an even smaller
percentage pay a lot. These are the “whales” – the large spenders who keep
everyone else afloat.
 The morality of the free-to-play model is a hotly debated topic. There are
certainly companies trying to suck every penny out of their players, and
using some questionable approaches to do that. No one should be exploiting
gambling addictions or tricking children into spending their parent’s money
or promising things to players that will never materialize.
But most free-to-play games, such as Candy Crush Saga, are not that
bad. Every monetization model – arcade cabinets, CDs in boxes,
hypercasual ad-driven games – is trying to use human psychology to get
people to buy things. At some level, all marketing is manipulation. But that
doesn’t mean it has to be an entirely bad thing. Good marketing is
manipulating people to notice things that they want and will enjoy. And I
believe that most free-to-play games are actually providing a valuable
service to people. People need entertainment and games are generally a
good deal in the entertainment world. Free-to-play allows most people to
get their entertainment for free, with some advertisements as the cost. And
the people who want to spend more and can afford it fund the rest of the
players.
I’ve done interviews with whales. Most of them are very happy with their
purchases and would spend the money the same way again. They have
disposable income, and this is how they choose to use it. Rich people spend
money on much sillier things than mobile games. You can find people who
overspend or otherwise have a negative spending experience, but even
among whales most of the transactions are mutually beneficial.
Candy Crush Saga is a quality game. Spending some money to make the
experience better is not unreasonable. I’m personally a free-to-play player
(most of the time – I’ll spend a little money to support games I particularly
enjoy), but I understand why people would spend without being coerced.
The free-to-play model isn’t perfect, and if I could wave a magic wand and
let everyone play and make games and all be rich I would, but I believe that
free-to-play is only problematic when abused.

History
Mobile games are the core of the games industry these days. Console games
and PC games are still going strong, but in terms of dollars spent, mobile
games are bigger than everything else combined.
Mobile games took a while to get there, though. In the early days of the
mobile game market, there was a lot of experimentation and uncertainty. No
one knew what would succeed.

Some companies tried releasing console-quality games (Infinity Blade

(2010)) or literal ports of PC or console games (Return to Monkey
Island (2022, mobile 2023), Grand Theft Auto: San Andreas (2002,
mobile 2013), XCOM: Enemy Within (2013, mobile 2014)). Neither of
these approaches became dominant, but there were a few notable
successes, such as a little title called Minecraft (2009, mobile 2011).
Although even Minecraft dropped its price significantly compared to
other platforms.
Some games experimented with the novelty of touch screen inputs.
Games like Flight Control (2009) or Where’s My Water? (2011)
discovered new styles of game that wouldn’t work with a controller.
These generally did well and you can see new innovations come out
and succeed regularly. A lot of the modern innovation in this area is in
the hypercasual genre where innovative ideas are needed to grab
people’s attention (and then throw a lot of ads at them).
This was also the early days of the shift to digital distribution across
the industry. Games could now be services with new content flowing
to players as quickly as it could be approved by Apple. Pocket God
 (2009) was an interesting experiment in this space, updating as quickly
as possible with big new features changing the very nature of the game
regularly. Games that play like Pocket God aren’t common these days,
but the model of constant meaningful updates has become very
common and remains successful, even on other platforms like Roblox
or for other games like Fortnite (2017).
Along with other changes, monetization was also uncertain. At first,
companies thought that mobile gamers would pay something close to
console price, or at least similar to the downloadable markets that were
developing on consoles and on Steam. But as companies experimented
with pricing, they found that lower prices worked. The increase in
sales offset the drop in price, so soon Angry Birds (2009) was the top
of the Grossing chart at the lowest possible price – 99 cents. Then
Apple added the option to sell things inside the game, and in-app
purchases were born. Immediately, games like Smurf Village (2010)
discovered that free was even more appealing to people than 99 cents.
The modern free-to-play model was born.

In this era, mobile games were also learning a lot from social games.
Facebook was a huge platform for games for a few years, before changes to
the advertising structure and success of mobile killed it. Social games
pioneered many of the free-to-play principles that later dominated the
mobile market.
Candy Crush Saga was originally a Facebook game. And before that, it
was originally a web game. King, the company that makes Candy Crush
Saga, has used an interesting method for testing out new titles. They
generally release lots of new games first on their web browser and see how
they perform there. The best-performing titles are considered for conversion
to other platforms. At the time, Facebook was considered the first step and
mobile was a new and experimental platform. So, Candy Crush Saga first
appeared as Candy Crush, a web game without the level variety that made
Candy Crush Saga so successful. The Saga idea first appeared in Bubble
Witch Saga and proved successful. When King brought their score attack
match-3 Candy Crush over to Facebook, they added similar Saga elements.
Candy Crush Saga did great on Facebook, so within a few months it was
brought over to mobile as well.
By the time Candy Crush Saga came to mobile (2012), the battle over
pricing was mostly done. The free-to-play models that had been pioneered
in Facebook were dominating on mobile. So, Candy Crush Saga launched
for free, with strong social elements. Many of those social elements have
faded to the background over the years as the industry has continued to
evolve and updates have changed Candy Crush Saga. Competing for high
scores against your friends, or for progress along the Saga, used to be
prominently displayed but now is more of an afterthought.
So, Candy Crush Saga came into the market at the right time with an
innovative take on both the gameplay of match-3 and the free-to-play
monetization model. The game was a hit from day one. King already knew
that it had a good chance, based on its success on other platforms, but
nothing is certain in this industry. In this case, it worked. Candy Crush Saga
was so successful it spawned multiple sequels, a line of merchandizing, and
even a short-lived network television game show. And it became my wife’s
favorite video game.

Meaning
What is Candy Crush Saga about? What does Candy Crush Saga say about
the world? How does playing Candy Crush Saga change how you see the
world?
 To me, Candy Crush Saga is about the joy of unintended consequences.
The player makes a seemingly simple innocuous action, and suddenly the
screen is filled with colorful explosions and some guy with a deep voice is
calling out compliments. The player didn’t plan that out. The player didn’t
know exactly what was going to happen. But like many things in life, taking
that first step leads to all sorts of unknown things.
Candy Crush Saga is the type of game that doesn’t really challenge the
player. Sure, some levels are nearly impossible, but those levels generally
require luck to complete along with skill. Mostly Candy Crush Saga is a
warm bath of compliments and pleasant explosions. And I believe these
types of games make the world a better place. People need comfort. People
need more praise in their lives. If a game can help people be happy and get
through their lives, I support that. Since the dawn of time, humans have
sought escapism. That need is deep in the human psyche. Games are the
latest delivery method. I’m not going to fight against the human psyche. Let
people enjoy simple fun.

Goals
Verbs tend to work together. In Shovel Knight, I can move, and I can attack.
I’m generally more successful in the game when I combine these verbs. I
don’t move, pause, wait, then attack. I attack while I’m moving. I look
around while I’m moving. I don’t stop and pause; I’m constantly
performing all the actions at one time and that’s part of what makes these
games challenging and interesting. Not all players can manage the feats of
dexterity and rapid stimulus processing required to be successful at move-
shovel when they all have to be done together.
The player’s goals can also work this way. When I’m moving to line up a
shot, I’m also considering how to best avoid the line of fire of another
enemy at the same time. Goals can be combined just like actions can be
combined (Figure 3.5). But goals are even more likely to be layered rather
than combined.

Figure 3.5 Two Verbs merging into a bigger verb.

Look at Candy Crush Saga. If you chose a random second of a player’s

session in Candy Crush Saga, the verb is likely to be “swap”. Maybe “look”
if you accept that as a valid verb. But the player’s goals are going to hit
many different layers of analysis:

1. Making a match
2. Having this match lead to a chain reaction or create a power gem
3. Having this match lead to a better board state when it’s done
4. Removing lots of gems
5. Making progress toward the goal of this level
6. Scoring points
7. Beating this level
8. Beating this level with three stars
9. Completing an event or challenge goal
 10. Getting to the end of the level progression
11. Getting a better score than my friends
12. Relaxing
13. Helping the cute little puppet creature to complete its goal
14. Clearing the chocolate
15. Removing the blockers
16. A very large number of level-specific goal types that I could list but
then this chapter would become boring and repetitive and as a game
designer I try to avoid boring things

This is normal. In games as well as life, a single action rarely has a single
goal. Goals come in herds. Some with subtle nuanced distinctions, and
some coming from completely different directions. I want to beat the level.
I want to beat the level with three stars. I want to beat the level with a
higher score than my brother. These are expressions of the same base goal
but with different nuances. From a game design standpoint, these can be
merged into a larger goal: success in the level (Figure 3.6).

Figure 3.6 Two goals merging into a bigger goal.

 The bigger differences between goals are often based on scale. I want this
action to clear lots of gems. I want to beat the whole level. I want to get to
the end of the game progression. I want the action I’m performing now to
make me successful now, soon, and in the future. This sort of layering of
goals is very important to game design. Game designers like to talk about
goals as onions – one goal is contained within the other goal (Figure 3.7).
Progress toward one smaller goal also progresses you toward a larger goal.

Figure 3.7 One goal inside another goal.

So, both goals progress based on the same player action, but the second
goal is bigger so it takes longer. Completing the first goal is a step toward
completing the second goal. When you complete the first goal, you still
have the second goal to keep you driving forward. At no point does the
player complete one goal and have nothing left to do. For a game designer,
a player with “nothing left to do” is a problem – they’re likely to leave the
game and go do something else. Layered goals fix this by keeping the
player motivated at all times.
 In Candy Crush Saga, a player may say that their goal is to beat this
level. But to do that, they need to find and make matches. Once they
complete the small goal and make that match, they are (hopefully) closer to
their big goal of completing the level. But now they have to start another
loop of activity to find another match and complete that small goal again.
Only after a few dozen small goals is the large goal reached.
And once the large goal is reached, the player is taken back to the Saga
Map where they remember that the large goal is just a small step toward the
ultimate long-term goal of progressing to the next story bit or to beat a
friend or to reach “the end”.
So maybe goals aren’t onions. Maybe goals are garlic. One head of garlic
seems like a single unified thing but when you break into it, you find many
smaller versions inside (Figure 3.8).

Figure 3.8 Many goals inside another goal.

This is good psychology. Humans like to group things and draw lines
around them. Part of the fun of goals is getting to define a task and then
check it off your to-do list. Completing a goal is fun.
 And to be clear, this isn’t a monetization trick. This is what makes the
game fun. “Engagement” is a key goal of free-to-play games, and it does
drive monetization. But it’s also a sign that players are having fun. In my
opinion, the best way to drive monetization is to make sure the player is
having fun and engagement is one way to identify when that’s happening.
This is why layers are great. If there is always something more to do, the
player is never looking at an empty checklist. When they complete one
check box and get that hit of serotonin, there are still more checkboxes to
do. Not just the next one, but one that’s halfway finished. Or even better,
one that’s just about to be finished. If you’re at 4% progress toward a goal,
there’s no pressure to stay in the game and complete it. But if you’re at
96%, then you might as well play one more level to finish it off. And then
when you do, there will be another goal at 97% waiting for you.
This is present in many games:

Civilization and other 4x strategy games are well-known for promoting

“one more turn” gameplay where there are many different layers of
goals that overlap in complex ways, ensuring there is always one more
thing to do.
Boss levels are another way to promote players’ desire to check off
boxes. The boss is the hardest part of the level, but it always shows up
when you’re already past 90% complete. The temptation to check off
that box is part of why boss levels are so motivating.
Role-playing games (RPGs) offer multiple forms of progression –
story quests, side quests, character stats, faction relationships, skill
points, equipment, etc. This ensures that whenever a player completes
a goal, there are more things to do.

You can see this in the evolution of Candy Crush Saga.

 Match-3 games such as Bejeweled existed before Candy Crush Saga,
with very similar core gameplay loops, even down to the specific power
gem abilities. But when Candy Crush Saga came out, its sales and
engagement blew them all away. Candy Crush Saga succeeded in large part
because of the Saga goal structure. The Saga structure takes a fun core
gameplay loop that had been mastered in previous match-3 games (Figure
3.9) and adds another layer of goals on top of it. You still want to make
good matches and get a good score, but now there are two new layers.

Figure 3.9 Goals merge: Make match, Create combos, Gain

points.

In the time-battle-based style of games like Bejeweled, there’s no way to

“beat” a level. You can get a good score, but you can never check off a box
that says the level is done.
The Saga map gives a clearer sense of progress on an individual level.
You can “beat” that level, even if there’s always room to go back and
improve your score (Figure 3.10).
 Figure 3.10 Same, but with a bigger circle around it: Beat level.

And once you beat a level, you’re taken back to that map and shown the
true goal – long-term progression. In Bejeweled, the end of a level may or
may not check a personal goal box depending on your final score. In Candy
Crush Saga, you’re usually checking at least one box and maybe more
(Figure 3.11). And you’re reminded of all the remaining boxes you can
check in the future. That’s much more emotionally compelling and drives
long-term retention in a great way.

Figure 3.11 Same, but with a bigger circle around it: Complete
game.
Other Goals
Looking at the list of goals, some are groups of related tasks (various ways
to evaluate a move), some are garlic cloves of related progression (move,
level, game), but some don’t fit either of these categories. Things like
“relax” or “help the puppet” suggest that there is a lot of room for other
types of goals.
Helping the puppet is an example of a narrative goal. The player is
motivated by emotional goals created by the game’s story, especially when
mixed with the motivation to progress a narrative forward. These are
powerful goals for humans, as storytelling and completion are deeply
woven into humanity. For many games, just this goal motivation is enough
to invest time and effort into a full story with (expensive) cinematics. Core
gameplay drives players forward, and progression keeps them wanting
more, but deep engagement requires some emotional investment.
Other goals like “relax” suggest that the player may have goals for
playing the game that are outside the structure of the game. The game
doesn’t directly have a relax button, and none of the explosions that occur
when candies match are directly designed to ensure player relaxation.
Relaxation is the reason the player comes to the game, not something the
game directly produces. When all the parts of the game are working
correctly, this goal will be the result. These sorts of goals are important to
look for and understand in your games. Just because they’re not entirely
driven by game features doesn’t mean that you as a game designer can’t
influence them. Understanding what your players want inside and outside
the game is critical.
Player-Generated Goals
One type of goal that doesn’t show up in the goal list is player-generated
goals. The list focuses on goals that the game creates and enforces, but
humans love to generate and track goals on their own. Every time you
generate a to-do list and mentally check off items as they’re completed,
you’re acting as your own game designer to structure and motivate daily
life. Players do this during games, too. In Candy Crush, players might strive
to clear a certain area or create a certain combo even if the game doesn’t
ask them to. This is good.
As a game designer, you can’t force players to make their own goals. But
you can create an environment where goals are easy to make and track and
enjoy. The more freedom the game gives the player to try divergent actions,
the more players are going to find them and enjoy them. Players like
interacting with everything in the game world, which is something most
game designs should encourage. Finding ways to give strong feedback
during these actions helps the player know that their actions matter and that
the game recognizes what they did. When I was working on Paperboy 64
(1999) (my first full game design), we made sure that almost everything
reacted when being hit by a newspaper. The player was only meant to target
a few specific things, but the world is much more fun and engaging if
everything reacts. And if a player wants to focus on throwing papers at dogs
to keep them away from cats, they can do that. They don’t earn points for it,
but the game still creates these little scenarios with emotional potential and
shows the player that their actions have consequences.
Player-generated goals are also an important driver for gameplay goals.
During early playtests, keep an eye out for things players try to do. Ask
players what their goal is. If players want to do something, the game should
recognize and reward it. In some cases, take those personal goals and make
them into gameplay goals. But if that’s not appropriate, take those personal
goals and acknowledge that they matter. Keep track of them in the design to
make sure they remain fun as the game changes and iterates. If possible,
track them and give the player a small thumbs up in-game when they
complete that goal, even if it’s unrelated to the main goal. Players like a pat
on the back, so give them as much feedback as you can to make that
happen.

Events
Since launch, Candy Crush Saga has been finding ways to push the player’s
ability to complete goals even further. One of the most successful features
in many free-to-play games is the event. For example, in Candy Crush
Saga, they might have a special race where the player is rewarded for
completing a certain number of levels before other players do. In the current
builds of Candy Crush Saga, some of these “events” have become
important enough to bake back into the main game progression cycle. They
started as special events that only appeared occasionally but now are just
something that happens naturally during play.
There are a lot of good psychological hooks built into limited-time
events.

Players love getting more things “for free”.

They encourage coming back soon, so you can complete the goal
before the event ends.
They feel special and often emphasize the unique never-before-seen
qualities of this event.
If the goals are set up right, they scratch competitive urges in a way
that doesn’t automatically advantage expert players.
In Candy Crush Saga, events are a great way to add an intermediate layer of
goals. Candy Crush Saga does a great job of providing immediate goals,
and the Saga Map provides a great long-term goal. But the long-term of the
Saga Map can be pretty far off, especially now that there are thousands and
thousands of levels. The motivation to reach the end of the map now takes
multiple years to complete. So, there’s a need for more checkboxes between
the “complete a level” box and the “beat the game” box.
Events provide a goal that is bigger than a single level but is still realistic
to actually complete. A player may never reach level 20,000, but they can
earn the free powerup by collecting a few more levels to win the race. It
doesn’t even matter if they want a free powerup – it will still motivate them
to push forward just to check the box. Humans are weird that way.
Being able to see this in retrospect is a sign of a reasonably skilled
designer. But the real trick is identifying this before the feature is
implemented. A really talented game designer would be able to look at
Candy Crush Saga at launch and identify the goal layers. A really really
talented game designer would be able to do that, and then notice that the
goal layers are fairly far apart. Even with only 80 levels, the motivation to
complete a level and the motivation to complete the game have a big gap.
And it takes a really really really talented game designer to do all that, and
then design the right solution to fix that problem while also fixing a few
other problems along the way. I never worked at King, so I don’t know who
came up with the idea to add the events. But they deserve a raise.
(I’d be willing to guess it wasn’t a single person sitting down doing this
analysis and going “Aha!”. It was almost certainly a joint effort by many
people who fed each other small good ideas and then tested and iterated on
those ideas to find the best ones. Someone along the way probably noticed
the goal structures and might have advocated strongly for certain ideas
based on it, but in games good things rarely happen because of one
moment. Games are made by teams, and good games are made by
iteration.)

Secret Verbs
So good goals are garlic, with lots of goals and lots of layers in each goal.
Good goals also support good verbs.
When you just think about the verbs of Candy Crush Saga, it can make
the game sound a little boring. You’re always doing the same thing? Just
swapping over and over? But the experience of playing the game is
different. The game feels engaging and fresh even after playing for a long
time. How does that happen?
One reason for this is that Candy Crush Saga actually has another
important verb, but it’s not one that is obvious. It’s not even an action the
player takes. So, it’s a little bit harder to notice.
Before making a swap, the player spends a good amount of time scanning
the board, looking for good matches. The player isn’t actively touching
anything, but the act of searching is an important part of the gameplay
process. Searching is generally more engaging than swapping – swapping is
just the mechanical outcome of searching.
Searching involves multiple types of mental processes (Figure 3.12).
 Figure 3.12 Flow diagram from before, with expanded player
actions to account for scanning and evaluating.

The player has to scan the board to identify what matches are possible.
As a game designer, it’s important to build levels that support
interesting board states.
For each match, the player has to evaluate the value of that match.
This can involve lots of possible factors, including stepping ahead
multiple turns to see how this action affects future actions. This is
the hard part of a Match-3 game.
But for most players in most situations, this analysis isn’t very
deep. If you are more interested in relaxation than victory, a quick
scan without a lot of second-order analysis is enough. The ability
to choose how deep to go is a great quality in a game design, as it
allows different players to approach the game in their own way,
and also results in varying skill levels with different outcomes,
which is also good for your game.
Once matches are identified, the player has to compare the matches to
determine which to do.
There are many possible ways to evaluate – there’s no one right
answer for most board states.
This can include predicting a few steps ahead. The player can
predict how the candies will fall, allowing them to set up future
matches. This can’t be done too far into the future since the player
doesn’t know what random candies will fall from the top of the
screen.

Scanning is fun where there is an interesting board state. What makes a

good board state? I’d argue there are two main qualities that make a match-
3 board a quality board.
First, the board needs the right number of possibilities. When a player is
scanning, they only have so much mental capacity. This is actually a real
thing in psychology – human brains hold about seven things at a time
(Miller, 1956). So as a game designer, you want to give the player about
seven things to think about. Probably a bit less so you’re not quite at the
limit. And maybe even less than that for a casual game where the player’s
not signing up for advanced memory feats. In Candy Crush Saga, the
number of things the player needs to worry about at any one time is easy to
count – just go through the current board state and count the number of
possible swaps. If you do this, you’ll see that most board states on most
levels have three to five options. Many levels start in a constrained space
where there are only one to three options. It’s very rare to have seven or
more choices. This makes sense, given our understanding of the human
brain.
And in many cases, the choice in Candy Crush Saga is even easier than
that. When we’re tracking player’s mental capacity, we don’t need to
include every possible option. The only ones that matter are the ones that
take up mental space in the player’s brain. If the player sees a swap but
immediately dismisses it as a bad idea, it’s not causing them any stress. So
really, only potential swaps that have a real possibility to be the current best
move are worth counting. In that case, the number is much lower – in most
boards, there are only one or maybe two moves that are fully worth
considering. Note that this varies a lot by players. Players who think ahead
and track future moves may notice options that other players quickly
discard. And to a new player, every option takes a lot of mental capacity to
evaluate. So, while only meaningful options matter, “meaningful” is in the
eye of the beholder. Game designers need to keep all options in mind when
considering all possible players.
The second quality of a good board state is uniqueness. Level 462 should
feel different from level 463. And from level 4 and level 4,621. If the player
starts to notice that the evaluation that they’re doing in this level is the same
thing they’ve done countless times before, they might start to get bored. As
with meaningfulness, this varies greatly from player to player. Some players
would say that every level of Candy Crush Saga feels the same to them.
They may need a different type of game to find true happiness. But for
many players, small differences are enough. And some players seek
familiarity and repetition and wouldn’t mind if there were only tiny
differences between levels. As a game designer, you want to make as many
people happy as possible but you also need to acknowledge and accept that
not everyone wants the same thing. Create the right level of variety for your
game and your audience and your personal sensibility.
Searching makes Candy Crush Saga more interesting and compelling
than a game that is just about swapping. Searching is where interesting
mental processes happen.
 Counting Options
Counting the number of meaningful options the player has in other games can be more
challenging than it is in Candy Crush Saga, but it can be done.

Most card games give the player five to seven cards in their hand at a time. Even then,
many of the cards are situational so they can be ignored most of the time. A starting hand
of Magic: The Gathering (1993) is seven cards, but usually two to three of those are
lands and only one to two of the spells are available based on mana costs. So, the player
only has a small set of choices for each decision in a turn.
Big strategy games and big open world games are called “big” because there are lots of
options. But once the player selects a goal, the number of options that effectively move
the player toward that goal drops considerably. That’s the number to keep track of.
Understanding what makes an option “meaningful” is often the difference between a
novice player and a skilled player. Looking at a random turn in a game of chess, a novice
might count all the ways each piece could move and consider their set of options. But a
skilled player knows there are only a few moves that make sense at that time.

Verbs and Goals

The other reason that Candy Crush Saga is compelling with only one (now
two) simple verbs is that the verbs aren’t really all that simple.
Verbs don’t exist in a vacuum. Verbs only matter in the context of a goal.
The actions available to the player don’t matter except in how the player
can use them to reach goals. The goals are what make the verbs interesting.
In Shovel Knight, we talked about how the player might use the same
verb in different ways to accomplish different goals.
In Candy Crush Saga, the action the player takes can’t change much.
When described mechanically, one swap is basically the same as all others.
But each swap includes a multitude of different goals.
Every time the player finds a possible match, the player has to balance a
range of possible goals that might be affected by that swap:
 Does this help me complete the unique goal of this level?
Does this create a power gem?
Does this set up another match as part of the cascade that falls after
this match?
Does this remove a blocker that is in the way of my future progress?
Does this move a gem closer to a key position for a future turn? Can I
get a bigger match on a future turn? Can I clear something related to
the unique goal on a future turn?
Is the random fall of candies after this match likely to do something
good?

Within each possible match, a large number of different goals overlap and
come together, forcing the player to understand and evaluate all these goals
to make a decision. Evaluation is a big part of the skill required in Candy
Crush Saga, and many other strategic games. Chess is a game of evaluation
and thinking ahead. Poker skill depends on multiple types of evaluation all
merging together into a mechanically simple action. Good game design
finds ways to unite many different systems of evaluation into each decision,
so players are always finding new interesting moves that are rewarded in
the game, even within well-known rule sets. This is one way that game
designers build interesting decisions.

Goal Design
As a game designer, you want to make sure this happens in your game. This
means thinking about what goals might motivate the player, both short-term
and long-term. If there aren’t enough different layers, consider adding new
features to make the player think short, long, and medium terms. Garlic
doesn’t just happen; it’s something that game designers must actively
generate to ensure that the player stays motivated.
 If your game has lots of goals, look at how they relate to your actions.
When the player engages with the core loop of the game, are they also
engaging with all the varied goals? If not, you might need to rethink your
core loop or restructure your goals. The core actions in the game should
directly tie into the core goals and most if not all of the secondary goals.
Goals that aren’t connected to your core loop are ones you might want to
consider cutting. This can be a helpful way to evaluate early game design
concepts – how well do the actions and goals align? How can that
alignment be improved?
This is not to say that everything the player does needs to touch on all
goals at all times. Secondary actions outside the core loop can have their
own side goals. Sometimes the point of non-core gameplay is to give the
player a break from their big oppressive goals and let the player enjoy
something else for a while. In an RPG, every quest doesn’t need to advance
the story. In an action game, every area doesn’t need to provide big rewards.
Having these alternate goals can provide important variety to keep players
engaged.
A game is only as good as its goals. Designing goals that excite players
and keep them engaged over time is key to making a great game. If a game
isn’t fun, it’s often the verbs that are at fault. But if a game isn’t engaging
for a long time, it’s usually time to tweak the goals.

OceanofPDF.com
Chapter 4
Centipede / Dynamics
DOI: 10.1201/9781003346586-4

Centipede
Centipede is a great game.
In Centipede, the player controls a “Bug Blaster”, which is a weird
spaceship thing that exists alongside giant bugs for some reason. The
player’s Bug Blaster shoots at attacking centipedes and other insects and
creepy crawly things, who also plant mushrooms that block your way.
Mushrooms in reality aren’t actually planted by falling insects, but in the
context of the game, it works.
Centipede (Figure 4.1) is a classic of the arcade era. It was not the first or
the biggest arcade game, but it was a key player when video games were
just getting started. All video games that exist today owe a debt to our
arcade ancestors who were the first to figure out this crazy business.
 Figure 4.1 Centipede – screenshot of the game. Note: The image
used is licensed under the Creative Commons Attribution 2.0
Generic license. The author is Rob Boudon. The image was
cropped to focus on the screen.

Centipede – Gameplay Dissection

Verbs
What does the player do in Centipede?
This is actually pretty easy to determine in old arcade games. Most old
arcade games had semi-unique inputs. So, looking at the buttons and other
things the player manipulates gives you a good sense of the verbs. Ignore
the coin input and “Player One Start” buttons for this purpose – they’re
basically the wrapper User Interface (UI), not gameplay. In this case, there
are two verbs:

1. Move. The player can use a trackball input to move their ship around
on the screen. The ship is limited to certain portions of the screen, but
this isn’t visually communicated to the player. The trackball was a
great control for this game as it allowed for quick zipping around with
constant changes in direction and speed, leading to very precise
character placement for skilled players. Later ports to home consoles
and computers suffered a bit due to less precise inputs.
2. Shoot (aka Attack). There is one button and it fires the ship’s weapon.
The shots follow a rule used by a number of classic arcade games: the
player can only have one shot on the screen at a time, so shooting at
things close to you allows you to shoot very rapidly, but if a shot
misses and flies across the whole screen, the player has to wait for it to
finish before firing again. This is partially due to the limitations of
early computers, making it hard to display lots of shots at once, but it
led to good gameplay. Just because something started as a technical
limitation doesn’t make it bad game design.

You may notice that the two main verbs of Centipede are similar to the two
main verbs of Shovel Knight (and first-person shooters and many other
games). Moving and attacking are key verbs in many different games, many
of which are very different games. This is a little bit weird and demonstrates
that while verbs are a great place to start, they aren’t going to answer all
your questions.
Just from verbs alone, we can see some key differences in how Shovel
Knight and Centipede work:
 Move Attack
Shovel 2D gravity-based motion Jump on enemy heads, or
Knight scrolling to the right in a swing a shovel in a fixed
vast 3D world, constantly forward animation. Used
moving to new areas with against a wide variety of
new maps. enemies with different
behaviors.
Centipede 2D point-to-point Create a particle that travels
movement on the bottom up. Used against a wide
section of a fixed screen. variety of enemies with
different behaviors.

So, the games have some similarities, including lots of diverse enemies.
But they are clearly different in many ways – different motions, different
attacks, different worlds. So, while the similar verbs do indicate a similar
popular starting point – traversing a space in order to be victorious in battle
– the games take this starting impulse to very different places. Why is that?

Goals
Centipede and Shovel Knight are also similar but different in what they ask
the player to do. What are the goals of these games?

1. Don’t Die

In both games, the player reaches “Game Over” if their avatar character
dies. Centipede has a simpler health system – any contact with the enemy
kills the ship – but the player can have multiple lives before Game Over
appears. This is more of a negative anti-goal but definitely fundamental to
play in this case.

2. Collect Points/Gold
Points are used in Centipede to measure success and encourage players to
do things, similar to how gold motivates and rewards many player actions
in Shovel Knight. “Score Points” is almost always a meta-goal more than a
goal itself. You can’t score points unless you understand the actions that
score points, so the point-scoring actions are the real goal, and points/gold
is just the way of recording your progress.
Lots of different game actions tie into the point system, but in Centipede
points generally come from destroying enemies. Contrast this against the
Shovel Knight point system (expressed in Gold), which rewards players for
combat but also for reaching interesting, hard-to-reach, or hidden areas in
the level. Shovel Knight wants players to complete combat, but it also wants
the player to explore the space in a way that doesn’t happen in a single-
screen game like Centipede.
While the Centipede score system is very focused on vanquishing, it still
has a number of nice flourishes. The player earns more points for
destroying spiders based on the distance between the spider and the ship
when the spider is eliminated. This incentivizes the player to move into
risky positions with spiders to try for that 1,000-point kill.

3. Destroy Enemies

Winning combat is the main goal of Centipede. The player spends most of
their time destroying enemies and working out how to better destroy
enemies. Destroying enemies is a goal in itself, separate from points, as
players want to remove threats and enemies are set up as the antagonists.
Not surprisingly, this ties into our two previous goals. The best way to stay
alive is to remove the enemies before they can touch you. And the best way
to score points is to destroy a bunch of enemies.
 4. Gardening

There is a subtler goal going on in Centipede that isn’t as obvious or

directly stated. Optimizing for combat requires dealing with pesky
mushrooms. They not only get in the player’s way; they also speed up the
centipedes. And when the centipedes are in the bottom half of the screen
alongside the player, they become another threat the player has to dodge.
So, the logical next step as a player is to remove mushrooms to clear out the
space, especially in ways that sculpt the mushrooms to send bugs flying
down the screen to you. This becomes a goal that isn’t as important as
directly destroying enemies but fills the space between enemy-destroying
moments and encourages players to adjust their behavior a bit while
destroying.
Centipede is a clever game, and the ways that mushrooms interact with
centipedes is a good example of that. When a centipede hits a mushroom, it
moves down one level and turns around (Figure 4.2).

Figure 4.2 Centipede gardening dynamics – The Mushroom

Bounce.

This makes sense, and in a normal mushroom arrangement means that

centipedes work their way down the screen to eventually collide with the
player. But certain arrangements – one mushroom offset and below another
– cause the centipede to move down, turn around, and immediately hit
another mushroom (Figure 4.3).

Figure 4.3 Centipede gardening dynamics – The Mushroom

Tunnel.

This can send a centipede straight down very quickly. Which is generally
bad for the player. But since this sends the centipede straight down, it can
also create opportunities. As previously mentioned, the player gets one shot
at a time but can fire rapidly upward when the shot hits something. So when
the centipedes are traveling slowly enough, creating a tunnel like this can
give the player a chance to quickly eliminate a complete centipede before it
becomes a threat. Especially in the early to mid-game, creating and
maintaining these arrangements is a good thing. And by the time the
centipedes are too fast to track, the player usually has no time to garden
anyways.
And to make it even more interesting, when a centipede head is
destroyed, it becomes a mushroom. So, when players are shooting
centipedes in tunnels, the tunnels quickly become clogged with new
mushrooms. Tunnels that are bad for the player result in no dead heads, so
they stay around naturally forever. Tunnels the player is using to their
advantage require maintenance by the player to keep around. This is good
game design – players only get advantages if they’re paying attention and
putting in the effort to earn that reward.
There are more secondary goals than just these. There are more
gardening patterns, especially for advanced expert players. Players may
have longer-term goals like getting on a high-score list or advancing a
certain number of waves. As discussed in the last chapters, there are lots of
types of goals, and players are always coming up with new ones. But when
doing a general gameplay dissection, the core focus is on the goals
presented by the game and how those drive the core gameplay. For this
purpose, the four goals above are the key.

Action Goals
Applying these goals back to the verbs, we start to see how the goals shape
the player’s actions.
“Score Points” is a general meta-goal that drives the other goals, so it’s
more a motivation behind the scenes rather than a direct motivator.
“Destroy Enemies” is the main goal, and most of the player’s moments
are guided by this goal. Both verbs are used extensively when approaching
this goal – the player needs to position themselves just right and then shoot
at the enemies. This is clearly emphasized with the fleas, who come down
fast in a specific column, requiring precise positioning to destroy, which is
also a great subtle tutorial to encourage the player to try that with rapidly
descending centipedes, as discussed above. But even with the trackball, it’s
easy to misjudge by a pixel or two and call up the anti-goal of “Don’t Die”.
Different enemies engage the destroy goal and the player inputs in
different ways.
 Slow centipedes give players targets to shoot at from a safe distance.
Later centipedes encourage the player to develop advanced techniques
such as gardening.
Spiders force the player to learn how to dodge. Since spiders come
from the side and the player can only shoot up, the player often has to
avoid the spider before they can take the shot.
Learning to dodge is important because once the centipedes get low,
the player has to dodge them, too.
Fleas are more about lining up the shot and quickly responding. These
reward rapid precise movement and a single decisive shot.
Gardening mushrooms requires some positioning. But each mushroom
takes multiple shots, which helps the player learn about how shots
work and how to rapid fire.
Mushrooms also provide a constant backdrop, which means players
who fire indiscriminately or miss their shot can still feel like they hit
something.

The anti-goal of “Don’t Die” is mostly a modifier on the moments where

the player is trying to destroy enemies. Dying is something that happens
along the way while the player is trying to destroy enemies, so “Don’t Die”
is a factor in the player’s decisions more than a separate decision on its
own. “Don’t Die” only becomes a focus when things have gone horribly,
horribly wrong – when there are a lot of enemies in the lower part of the
screen.
“Gardening” can sometimes be a modifier on “Destroy Enemies”.
Sometimes the player will adjust their path or fire off a few extra shots
while crossing the screen to destroy a few errant mushrooms. But
“Gardening” can also be a primary goal. When the threat level from the
enemies is low, the player may decide to focus their attention on mushroom
removal for a few seconds. This is encouraged by the firing behavior
mentioned earlier – the most dangerous mushrooms are the ones inside the
player’s movement area, but the ability to fire rapidly at nearby targets
makes them the easiest mushrooms to remove. Park the ship right below a
mushroom and hold down the button and the mushroom will be gone in an
instant.
Allowing the player to sometimes focus on “Gardening” even has some
nice subtle nuances. The decision to garden versus attacking gives the
player an interesting meta layer to consider when not panicking. But even
within the gardening approach, it’s not always the best strategy to remove
all mushrooms. When the lower section is mostly clear, fleas are much
more likely. Which can be dangerous, but can also be profitable for players
who can spare a moment to rack up some points. It’s a judgment call that
individual players can make.

Challenge
So there really are two types of moments in Centipede.
Enemy Moments where you focus on efficiently removing enemies.
Gardening Moments where you focus on modifying the terrain to make
the other moments easier.
What makes these moments challenging?
The vast majority of the challenge in Centipede comes from the enemies.
The exact count of enemies depends on whether you consider centipede
heads and centipede bodies to be the same thing, but there are always at
least a few and they’re all unique and different.

Each level includes a main centipede that is working its way down the
screen. This is the most constant and explicit enemy in each level. It’s
so constant it’s even the name of the game. The player has to learn
 various techniques to deal with centipedes, and the strategies change
based on how many there are, where they are, and how many segments
they have remaining. Having a regular enemy makes for regular
patterns that the player can learn and keeps the feel of every level
consistent. An all-spider level would be an interesting variant bonus
level but would clearly not be the basic gameplay.
The spider is very different than the centipede. It wiggles around semi-
randomly in the player’s space. The spider is not the primary goal but
instead serves to make movement more interesting. The player can’t
freely move around their space because they have to constantly dodge
or destroy this erratic threat. And new spiders can appear from
anywhere, adding another layer of threat. Be careful around the edges!
Fleas are another different layer of threat. Fleas are an immediate
physical threat, as they fall fast. But the real problem with fleas is how
they affect the environment. This can change the nature of a level
quickly, and provide new behaviors for the centipedes that are coming
down. Fleas alter everything else the player is doing and also feed into
the Gardening mode discussed above.
Scorpions are a meta-threat. They appear in the top section of the
board and convert mushrooms into poison mushrooms. When a
centipede touches a poison mushroom, it zips down the screen quickly,
becoming an immediate problem. So, scorpions are not a direct threat
to the player, but they create a situation that can quickly get out of
hand. Since scorpions appear at the top of the screen, they might be
hard to shoot if you haven’t been gardening to keep some clear shots
available.

The different enemies provide different challenges that focus on different

goals and verbs. This is really good enemy design. The enemies also
interact with the mushrooms and with each other in interesting ways that
force the player to think about everything together.
During Gardening Moments, there is not an explicit danger that the
player faces. It can still be challenging, but it’s a different flavor of
challenge. The player still has to manipulate the controls carefully to
quickly get to the right place and shoot the right thing. But in the Gardening
case, the other thing isn’t moving or trying to kill the player. So, it’s a lower
level of challenge.
There is a second layer of challenge to Gardening. Just because you’re
positioning and shooting well doesn’t mean you’re gardening well.
Selecting targets and selecting the right goal are just as important. There are
decisions the player has to make based on their understanding of how the
game systems work and their assessment of the current state of the board.
These are a bit lower pressure than combat decisions but still require rapid
processing since the game never lets up and enemies are always about to
reappear. Quick thinking and fast reactions are required, even when taking
it easy.

Flow
Putting the Actions, Challenges, and Goals together creates a simple flow
diagram for Centipede, as seen in Figure 4.4.

Figure 4.4 Centipede flow from Verbs through Challenges to

Goals.
 But while clearing the board is always the player’s main goal, that goal
can be broken down into multiple smaller goals that direct the player’s
actions at any given time (Figure 4.5).

Figure 4.5 Centipede flow from Verbs through Challenges to

Goals, broken down into more lines.

In this case, these three goals can be clearly prioritized (Figure 4.6).
There’s no benefit to focusing on the centipede if you’re in immediate
danger. And if there’s an opportunity to take down the centipede, it’s not
best to be focusing on mushroom maintenance.

Figure 4.6 Centipede flow from Verbs through Challenges to

Goals, as questions.
 So, while gardening is the least important of these to the player’s
survival, it’s also the one that has the most options (Figure 4.7). Surviving a
spider requires precision and quick reflexes, but there’s not a wide range of
possible player actions that end in a good result. Taking out the centipede
allows for more range – players can try to snipe when the centipedes first
appear, arrange for a chance to quickly eliminate it as it descends, or engage
in brutal trench warfare with loose heads at the bottom of the screen.
Gardening is even more expansive – players can decide on what sort of
board they want and take steps to achieve that future. Some of these long-
term goals are natural outcomes of the game dynamics – clear out open
space in the bottom to move around, and create pathways for centipedes.
But players can choose what to focus on and what specific board layouts
work for their play style. There are many possibilities, but only if the player
can complete the top two actions quickly enough to have time to spare.

Figure 4.7 Centipede Goaly Goals: Same diagram, but with

expanded choice options on lower questions.

Rules
How does a game come together like this?
 Games don’t happen by accident. Every game has been crafted by a long
series of decisions, either by one person or a team of many. Each Verb,
Goal, and Challenge is made up of a lot of little things that lead to the
desired effect. There’s a lot going on behind the scenes to make each part
work.
We already talked a bit about how shots work. We call “shooting” the
verb, but the game is really based on the nuances of exactly how the shots
work. If you change even a small detail about the shot, the game would be
very different.
When we think about a board game, we’re used to seeing the rules
written out and directly stated to the player. But in a video game, it’s not so
obvious. You have to watch and pay attention to catch the details of how the
rules work. Let’s try to list out some of the rules for how shots work:

1. There can only be one shot on the screen at one time.

2. Shots move quickly.
3. Shots move straight up.
4. A shot is created when the player presses the shoot button.
5. When a shot is created, it appears at the top of the ship.
6. The player can’t shoot while there is a shot already on the screen.
7. If the player is holding down the shoot button, a new shot is created
when the old one ends.
8. A shot is destroyed when it hits anything.
9. When the shot collides with an enemy, the enemy is destroyed.
10. Centipedes create mushrooms when destroyed.
11. Centipedes split when a body segment is destroyed, creating a new
head.
12. When the shot collides with a mushroom, 1/4 of the mushroom is
destroyed.
 13. When the shot reaches the edge of the screen, the shot is removed.

I think these rules are a fairly complete description of how the shot behaves
for most cases. If I wanted to add more precision, I could add some pixel
lengths and exact times. When you’re describing a game that already exists,
that can be useful. When you’re writing out a design document for a new
idea, it’s not really worth detailing everything since those values are likely
to change during iteration. Sometimes when I’m writing those early docs,
I’ll make note of the desired variables so they get into the code, especially
variables I know I’m going to want to manipulate regularly during iteration.
Having all of those in a text file (or even better in an in-game editor) saves
lots of iteration time.
If you wrote out a list like this for every feature in the game, you’d have
the Game Design Document, or at least a solid start on the features part.
Sometimes people do that as a starting point for designing a game, but
usually it’s more iterative than that. The game designer might write out a
few of the key points and key features, but not define everything in this
level of detail. Or in some cases, every feature will get a document like this
but some of them won’t be written until halfway through production. As
with all documents, use lists like this when it helps you think things through
or helps you communicate with others, and don’t use lists like this when
they’re not useful.
 Break it Down
How do you generate a list like that? Breaking down a feature requires looking at the feature
from lots of perspectives.

Time. What are the steps that have to happen? In this case, creating a shot, the shot
moving, then the shot hitting something.
Visual. What does the thing look like? Can you break it apart spatially? Shots have a
clear visual, but not a lot of supporting visuals. There’s no visual effect when a shot hits
an enemy, but the consequence is pretty clear.
Goals. What does the player want out of this? In this case, the moment of destruction is
key. That’s how the player achieves their goal.
Physics. What are the operational qualities of the motion and reactions?
Interactions. What happens when X interacts with Y? What happens when a shot hits an
enemy? A non-enemy object?
Information. What does the player need to know about this? How is that shown? In this
case, the position of the shot is key and that is clear from the shot’s visuals.

Using Rules to Identify Changes

Breaking out the rules like this shows how dissection can be helpful when
making new games. Each rule suggests different changes that could be
made to the game with interesting possible effects. If you’re trying to
determine how to make your game interesting and new, a great place to start
is to look closely at similar games and see where you can change a few
things. The 2021 release “Centipede Recharged” includes a few of these
proposed features, which demonstrates the utility of this sort of exercise.
But since our dissection is limited to the original arcade game, let’s focus
on that for our Centipede feature ideas:

1. There can only be one shot on the screen at one time.

a. Allowing more than one shot, possibly as a powerup, would be
powerful. The bullet hell genre provides examples of how this can
 be fun but was only possible when technology got better at
displaying lots of things on screen at once.
2. Shots move quickly.
a. Altering the speed of the shot would have a big effect on play. I
wouldn’t want to slow things down permanently, but a temporary
effect based on enemies or certain level types could be
interesting. Frustrating, but interesting. I especially like that (if
handled correctly) this would call the player’s attention to the
“one shot at a time” rule in a way that would aid understanding
without needing a text explanation.
3. Shots move straight up.
a. Having shots wiggle around a little could be a way to make the
game less skill-intensive if you’re trying to make this appeal to a
broader audience. Giving shots a zig-zag path would make them
more likely to hit something but also less precise and predictable.
b. Currently, the shots only move straight up. If these were real
projectiles, they would follow physics and have some momentum
based on the movement of the ship. This could be interesting to
experiment with, but I suspect it would be terrible in a world
where you want to shoot exactly straight up to hit fleas. When
designing a game, it’s more important that your idea is fun than it
is that your idea is accurate to reality. Not every idea has to be a
winner; even the ideas that don’t end up changing the game help
you understand things better.
c. Physics could also slow the shots down over time to simulate
gravity. But this also doesn’t sound very fun. Often, realism is
less fun than simplicity.
 d. Shots could project in different directions. Some similar games
have powerups that add additional shots that fire in diagonals. But
if you only have one shot, straight up is a lot easier to understand
than diagonals.
4. A shot is created when the player presses the shoot button.
a. A powerup that creates multiple shots per press could be fun.
b. A powerup that creates different types of shots could be fun.
c. A version of Centipede where the Bug Blaster constantly fires
without player input would be interesting. Removing or
automating a control can make a game simpler, but in this case at
the cost of reducing precision. In this case, it would give the
player much less control over the (interesting) shot economy,
which doesn’t seem worth it.
5. When a shot is created, it appears at the top of the ship.
a. Shots could appear from different locations on the ship, but center
top is the simplest and easiest to understand so this doesn’t seem
like a great place to change, at least for the first shot.
b. An upgrade that allows multiple shots at once would need to
modify this rule, so shots appear in different positions.
6. The player can’t shoot while there is a shot on the screen.
a. This is the rule that limits the player’s ability to fill the screen
with shots.
b. While this is a key technical limitation, the rules allow rapid fire
under the correct circumstances.
c. This could also be controlled by a timer – the ship needs to wait a
second between shots. This would still limit the number on
screen, but it wouldn’t allow the rapid-fire fun that makes quickly
 eliminating a full centipede possible. So, I’d argue it would be
less fun.
d. The number of shots on the screen could also be controlled by
having the shots have a limited distance. But then you lose the
ability to handle centipedes when they first appear.
e. This rule is at least partially predicted by the technical limits of
the original arcade game. That is no longer a concern – modern
machines can manage more shots than the player could recognize
at once. Features like this that have outlived their technical
constraints are often worth examining and possibly changing. But
not always – this one has a lot of important effects on the rest of
the game so shouldn’t be changed lightly.
7. If the player is holding down the shoot button, a new shot is created
when the old one is removed.
a. Some games use this input to create a charged shot. As the player
holds down the shoot button, the shot projectile object gets bigger
and more powerful. When the player releases the button, the shot
is fired. This is getting pretty far from traditional Centipede
gameplay but could be an interesting way to diverge. The
question I would have is which is more fun – charging up a shot,
or rapid fire? That’s often the best lens to evaluate potential new
features.
8. The shot is destroyed when it hits anything.
a. There could be some targets that go away when hit without
removing the shot. The shot goes right through them and keeps
moving, allowing the player to hit multiple things with one shot.
b. There could be things that destroy the shot when it travels near
them. If the player misses, the shot is removed when it gets near
 the enemy. This might make certain areas of the screen more
challenging – the player would need to take out these enemies
before dealing with other threats behind them. This might be hard
to visually explain – maybe a beetle with wings that pop out but
aren’t damaged by attacks but is destroyed when its head is hit?
9. When the shot collides with an enemy, the enemy is destroyed.
a. For this rule, each part of the centipede is considered a separate
enemy.
b. Centipede doesn’t have any enemies that require multiple hits
(other than mushrooms, but they’re more scenery than an enemy).
If I were creating a boss level for Centipede, I’d probably start
with something like how mushrooms are destroyed. Or just a
multi-object entity where each part is destroyed separately, like
centipedes are.
10. Centipedes create mushrooms when destroyed.
a. The semi-sequel to Centipede (Millipede) adds bombs that create
a dangerous cloud when shot. So, the enemy is destroyed but
replaced with something new. This could be used for lots of
interesting effects. Replace the enemy with other threats, or
beneficial things, or things that alter the behavior or traits of other
enemies. Lots of potential design space there.
11. Centipedes split when a body segment is destroyed, creating a new
head.
a. This is an important feature that sets the tone for the rest of the
game, so I wouldn’t remove it entirely.
b. Having some segments that treat this rule differently would be
interesting. Maybe hitting certain body parts increases or reduces
the size of the resulting centipede, encouraging players to attack
 certain areas. Centipedes could vary based on the components
that make up their body.
c. Or hitting certain parts of the centipede causes different behaviors
for the new centipedes. Destroy an abdomen and the new
centipede is stunned for a second. Destroy the nervous system and
the new centipede head moves straight down.
12. When the shot collides with a mushroom, 1/4 of the mushroom is
destroyed.
a. I’m confident that when setting up the original Centipede, they
tried out different numbers here and settled on four hits being the
right pacing. This is the sort of thing that prototypes are good at
determining and setting this is important because it determines the
feel of the rest of the game.
b. Enemy design has a lot of great variety, but there are only two
types of mushrooms in the game, and from a Gardening
perspective they behave the same. It’d be interesting to have some
objects that clutter the space but behave differently. Bubbles that
take one shot to destroy and don’t remove the shot when hit.
Rocks that can’t be removed. Boulders that roll when hit. This has
a lot of potential to add variety.
13. When the shot reaches the edge of the screen, the shot is removed.
a. This is pretty important, given that the player only gets one shot.
b. The shot could bounce instead of just stopping immediately.
There would need to be some limits so it doesn’t just keep going
forever, but games like Arkanoid play in this space to good effect.

As a game designer, the rules of your game are up to you. As you can see,
some of these rules are critical and dangerous to adjust. Some are only
loosely held together and easy to poke at to get different effects. This is one
of the key roles of the game designer – to identify which rules can and
should change during development of a game.
Even a simple game like Centipede has lots of choices for a game
designer. Which of these changes are the right changes? It really depends on
what you’re trying to do. Identifying places where things can change is
easy. The hard part is identifying the right changes. And to do that, you
need to know what your goals are. Not the player goals like we’ve been
talking about, but rather the development goals. What type of game do you
want to make? What is the feel and mood you’re looking for? Or if you’re
looking at a more specific change, what problem in the game are you trying
to fix?
Once you’ve gotten out of the initial brainstorming phase of a game, you
shouldn’t make any change without a specific goal in mind. Change is
scary. As we’ll talk about, changing anything in the game can have lots of
unintended consequences. Change for change’s sake is a common mistake
that makes game development harder and makes games take forever to
finish. Don’t be afraid to make changes, but understand that every change
has a cost, and the benefit needs to outweigh that cost.
While any change matters, the cost gets higher the later you are in
development. Early in a game’s lifecycle, trying out new ideas just because
they sound cool can be a good idea. But once the game is halfway done and
all the content is designed, making a change to the core game loop is going
to be costly and likely requires reworking large chunks of the game. Only
do that if absolutely necessary.

Progression
Centipede has a simple progression system. Each level starts when one or
more centipedes appear at the top of the screen and ends when there are no
centipedes on the board. The board is not cleared between levels, which
means that as mushrooms start to fill the board, they can become a long-
term problem. Centipedes appear in a repeating pattern, with increasing
difficulty coming from the arrangement of starting centipedes, increases in
the speed of the game, the presence of more advanced enemies, and the
tendency for the board to fill up with mushrooms.

Centipede – Non-Gameplay Dissection

Senses
Centipede has a lot going on for only one screen. This is common in old
arcade games. In order to compete in a crowded arcade, games were forced
to be loud and garish. If you don’t get noticed, you don’t get quarters.
Centipede has a distinct visual style for the era, with more greens and
pastels in its color palette than the stark neon primary colors of Pac-Man
(1980) or the even starker whites of Asteroids (1979). This still fits the need
to stand out and be noticed but demonstrates how the arcade market was
starting to mature and expand a bit. If everyone is loud and garish, you need
to do more than just that to get those quarters.
The black screen and clear lines of early arcade games were mostly a
technical restriction, but one that led to very clear visual communication.
It’s easy to tell what is important when the restrictions mean only the
important things are drawn. And it’s easy to see boundaries and collisions
when everything is drawn with clear lines against a black background.
Even with the technical limitations of the times, Centipede still uses
sound and visual effects to emphasize when an enemy is defeated. A little
particle explosion persists in that location for a few frames. Similar to
Shovel Knight, it’s important for players to know the consequences of their
actions.

History
Centipede is a clear example of the golden age of arcade games. Centipede
came out in 1981. Arcade games started in 1972 with Pong but really took
off with Space Invaders in 1978. If you weren’t around in those days, it’s
hard to overstate how big arcade games were. This was the first time many
people around the world saw computers controlling games. Video game
arcades were huge successes and the media noticed. This boom only lasted
a few years – by the early 1980s home consoles started to take some of the
edge off the arcade, especially dedicated arcades with dozens or hundreds
of games.
Centipede was not the first big arcade game and not the last. But it was
big.
An interesting note about Centipede is that it is one of the few classic
arcade games coded by a woman. Dona Bailey was the only female
programmer at Atari and was a key member of the four-person team that
made Centipede. In interviews, Bailey and others describe an intent to make
Centipede more appealing to women, focusing on natural themes and less
human on human violence. It worked – Centipede had a large female
audience, especially compared to other games of the time (Guide, C., &
Gary, K. M. (1982)).

Money
Centipede made money through quarters. Technically the game developer
made money by selling arcade machines, but the demand for those
machines was entirely driven by how many quarters went into each, so it’s
fair to say that quarters are what matter. And you can see the importance of
quarters throughout the arcade era. Monetization isn’t a new thing that
changes how games are made; every game has to consider how it makes
money in the core design.

Arcade games need to attract attention quickly in a crowded market, so

they are loud and bright.
The player spends money at the start of the game, and no new quarters
go in until the player’s session ends. So, sessions try to be short –
maybe a couple of minutes. After a minute or so, the game is actively
trying to get rid of the player, overwhelming them with increasingly
deadly waves of enemies. Arcade games are aggressive and difficult,
and once they become difficult, they don’t give you a break.
Centipede’s progression system is built around this rapid difficulty
rise. It’s possible to stick around for a while, but only players with a lot
of talent and a lot of time investment can last more than a few minutes.
Arcade games are highly skill based. If you’re good, you can last much
longer than an unskilled player. This might seem to violate principle
#1, but the only way to get good is to play for a long time, so
rewarding skilled players creates a longer-term motivation for the
player to keep spending quarters. If the game is really hard but I have
no sense that I can master it, why play? If there’s another player who
shows me what greatness looks like, I am driven to emulate them (by
spending quarters).
High Score lists are a great example of a game feature that was added
to support monetization. Watching a great player makes me want to get
better. Typing in my name codifies that goal into something real.
Story doesn’t really matter. Players are only around for a few minutes
each, so it’s more about the thrill than guiding players through a deep
narrative.
Narrative and Emotion
Centipede is not a game that oozes with story. It’s not even very clear what
the player’s avatar is meant to represent. Official marketing materials refer
to it as a “Bug Blaster”, but even then, I’m not sure if it’s a bug or a
spaceship or what. The original version of the game contained no
explanatory text or cut scenes. Removing bugs is a fairly intuitive action
that resonates well for many people, so no additional context was provided.
Later games have added some narrative layers, but when the original arcade
cabinet came out there was none. Which was pretty common for the time.
Emotionally the game generally has the strong visceral emotions of most
action games. The game can be frustrating, both in a positive way when the
player reaches their skill limit and in a negative way when a spider spawns
in a position that doesn’t give the player time to react. The game is one of
the first to start to explore feelings of nurturing or shaping and building, but
it’s a bit overwhelmed by the rapid action going on and hard to really dig
into for most players. And calling this a nurturing game is a bit of a stretch.
So, there’s more emotional potential here than some competing games, but
it’s mostly untapped.

Meaning
What is Centipede about? What does it say about the world? How is a
player changed by the experience?
A player could see this game as saying that the natural world is the
enemy and must be controlled. Bugs are threats. Mushrooms get in the way.
A player could see this game as a metaphor for the frantic nature of the
time it was created. Many arcade games require constant motion and
constant work to delay an inevitable decline. The 1980s were a time when
people were struggling with the demands of work and the drive for
corporate success. Not that those pressures don’t exist today, but they were
discussed in many types of media and art at the time, and video games are
one expression of that drive in society.
To me, Centipede is really about the power of systems and dynamics. The
game has rules, but like a natural system the overall experience is much
greater than those rules by themselves. Mushrooms and centipedes and
spiders combine to create a unique and compelling experience that is more
than the sum of its parts. I think it’s not surprising that a game about nature
is one that expresses these system ideas very clearly. I have a degree in
Biology, and only later realized that my love of games and love of biology
are both expressions of my love of systems.

Dynamics
There’s a new word I want to introduce that is very useful when dissecting
games. When two or more rules combine to create an outcome not directly
stated by those rules, that is a dynamic.
We’ve already been talking about the dynamics of Centipede. The idea
that a player might want to remove mushrooms is not directly stated in the
rules of the game of Centipede. But it’s a logical outcome of a series of
other rules:

RULE: Centipedes move down and turn around when they collide with
a mushroom.
DYNAMIC: This means that mushrooms can accelerate the
centipede’s descent.
RULE: The ship can only move around the bottom of the screen.
RULE: The only way centipedes harm the ship is by colliding with it.
 DYNAMIC: These two together mean that the centipede is only a
threat when it’s near the bottom of the screen.
DYNAMIC: Since the centipede is only a threat near the bottom,
and mushrooms make the centipede reach the bottom faster, it
makes sense that the player would want to remove them.
This dynamic is complicated by other rules:
RULE: Fleas drop more mushrooms on the screen.
RULE: Centipede segments turn into mushrooms when they die.
RULE: Fleas spawn more often if there are fewer mushrooms.
And further complicated by other dynamics:
Fleas cause rapid descent of centipedes through a set of
dynamics:
DYNAMIC: There are certain patterns of mushrooms that
cause very rapid descent just due to the previously stated
rules. For example, if you have two columns of mushrooms
near each other, the centipede will bounce back and forth
quickly.
DYNAMIC: Fleas tend to make mushrooms in columns,
which encourages this pattern.
DYNAMIC: This encourages players to shoot fleas as
quickly as possible so they don’t make a long column of
mushrooms.

These examples show how individual rules can lead to more complex
results. Many of the example dynamics above are strategies – sets of actions
that the player is encouraged to take based on a combination of rules. And
these are very important to understand when you’re designing games – you
want to make sure that the strategies you’re encouraging the player to use
are fun and create the type of gameplay that you’re looking for. But
strategies are not the only type of dynamics. Any combination of rules that
leads to a new result counts.

When a game is described as “elegant”, that usually means that it takes

a relatively small set of rules and builds a lot of interesting dynamics
around those rules. The rules of Go are (basically) “On your turn, place
a stone of your color on a 19x19 grid. If a group of stones is
surrounded, it is removed. Whoever has the most stones wins”. But
this foundation leads to a lot of other dynamics.
You can build enemy AI out of very simple rules that result in a
behavior that feels complex. For example, if you want a group of AI-
controlled characters to move as a group, you can get very compelling
flocking behavior out of a few simple pointers that each entity
interprets on its own. Flocking is a dynamic.
The combat in Halo has a certain feel. It’s fast-paced but with an
interesting back and forth based on weapon types and distances. These
are not written in the rules or the code but are a consequence of the
code. The “feel” of the game is another important type of dynamic.

Talking about dynamics is a way to understand how different rules fit

together to make something bigger. It’s a way to read between the lines and
see the underlying systems that are built out of the individual rules.
Dynamics are important because the dynamic is often the target that the
game designer is trying to hit. Game designers don’t strive to make a game
with 0.7 seconds between sniper rifle shots (see the 2010 GDC talk by
Jaime Griesemer “Design in Detail: Changing the Time Between Shots for
the Sniper Rifle from 0.5 to 0.7 seconds for Halo 3”), they strive to make
combat feel fast-paced with an interesting back and forth, combat that feels
different every time because of different abilities that combine to create new
effects or to make a game that evokes a sense of wonder and exploration.
The game designer’s goal (separate from the player’s goal) is to make those
specific dynamics happen. But the game designer can’t just code in “make
the combat fast-paced” or “wonder”. They have to work with individual
rules and content values that combine to create the desired effect.
Understanding dynamics in games is a key step toward understanding
games.

Simple AI Rules
I worked on a game called Hunter: the Reckoning (2002), where the player fights zombies.
Early on, we built a robust state-based AI system with lots of controls and inputs to determine
how the zombies move and attack. Once we started testing it and playing with it, we ended up
only using a few of those commands in the final game.
In a game about killing hordes of zombies, the enemies don’t need to do much other than
“move toward the player’s character” and “attack when you’re close enough”. Since the
enemies are only supposed to be on the screen for a few seconds each, anything else was not
just unnecessary but often just confusing. Zombies turning around because another goal caught
their eye didn’t make sense to the player. Our zombies did not need robust internal narratives;
they needed to attack once or twice and then die.
The complex AI system was used in a few places. Some of the later enemies had more
complex behaviors, but they were still very simple things like “use this second attack every 3
seconds” or “the first time you move toward the player, do this special charge attack”. Bosses
were the only encounters that used the full AI, with multiple stages and ranges of attack. But
even the bosses were simpler than we had initially anticipated.

Dynamics in Centipede
So this gives us another important tool when we’re trying to understand a
game like Centipede. Understanding individual rules is important, but it’s
also important to understand when rules combine to become greater
together.
We already discussed mushroom gardening.
 And we discussed how the “one shot on screen at a time” rule leads to
some interesting consequences: players can rapid-fire on close-up targets
such as mushrooms, which makes mushroom removal more viable and fun.
If mushrooms behaved differently, or if shots behaved differently, this
would be different both in its strategic consequences and also in the visceral
feel of the game.
You can combine rules to create interesting, unexpected dynamics. I can
logically prove “The spider is your friend”:

1. RULE: The player can only move in a certain area at the bottom of the
screen.
2. RULE: The spider only moves in that same space.
3. RULE: The spider destroys mushrooms when it touches one.
4. RULE: Mushrooms block the player’s movement. (Which spawns a
DYNAMIC: “Mushrooms can be annoying and make you die when
they get in your way”.)
Thus,
5. DYNAMIC: Sometimes you might want to let the spider live for a
while before you destroy it.

Or “Rapidly descending centipedes are your friend”:

1. DYNAMIC: A combination of mushroom patterns and centipede logic

means the centipede can get into a state where it rapidly descends
directly down. This is usually a problem as the player doesn’t want the
centipede down low.
2. RULE: The player can only shoot directly up.
3. DYNAMIC: The player can fire rapidly in some cases due to the rules
around how shots are generated.
 4. DYNAMIC: When the centipede is descending downward, if the
player is in the right position, the player can rapid fire and destroy an
entire centipede while it descends.
Thus,
5. DYNAMIC: Sometimes it’s good to have columns of mushrooms that
send the centipede rapidly down.

A good game will have many dynamics, and some may be surprisingly
counter-intuitive.

Dynamics: Phases
Many games have “phases” where the player needs to adapt to changes in
the game world. In some games, this is direct and obvious – Mario Party
(1998) tells you when you’re starting a new mini-game. But even games
without big signposts have different phases that players recognize without
being told. These are dynamics that come about as a result of all the rules
and content in the game. For example, in Centipede there are two main
phases:

1. Descent: The centipede has appeared and is working its way down the
board. The player is encouraged to destroy the centipede quickly from
a distance to avoid entering the more dangerous other mode.
2. Crowded: The centipede or centipedes are in the player’s movement
space, making it crowded and thus endangering the player.

Then there are meta-phases between those two:

1. Between: There is a short pause between killing the last centipede and
the appearance of another. This gives the player a moment to do some
quick Gardening.
 2. Top Heavy: Eventually after many stages, the top of the screen will
usually be very densely packed with mushrooms. This will cause new
centipedes to immediately race down to the bottom. There is still a
Descent phase, which still gives the player a viable opportunity to
destroy the Centipede, due to the “rapidly descending centipedes are
your friend” dynamic discussed above. But the window to shoot them
down gets shorter and shorter as the centipede speeds up.

These stages aren’t built into the rules or hard-coded, but they’re natural
progressions from the rules and code. That’s another example of dynamics.

Systems
Dynamics are an example of how games are built as systems.
A system is any set of things where the relationships between the things
are as important as the things themselves.

The world economy is a big system where the products made in one
area affect the products and prices in other areas.
The natural world is a system where species get food in various ways
and the populations of one species affect the populations of other
species nearby.
Humans (and all living beings) are made up of many organs which
have clear functions but depend on the functioning of other organs.
A family is a system where the actions and moods of one family
member can have a huge effect on the actions and moods of the others.

Systems are everywhere, and understanding systems is key to

understanding everything.
 The study of systems is a fairly new and growing field. All systems have
certain qualities and relationships that are shared, and understanding these
meta-rules for how systems work can help in understanding any system. I
recommend books like “Thinking in Systems” (Donella H Meadows, 2008)
for any game designer interested in better understanding how games work.

Centipede Systems
So, if games are systems, what are the systems of Centipede?
This can be a difficult question to answer, and it really depends on what
you’re looking for. The needs of your dissection determine where you want
to look for systems and how detailed you want to look. If you’re just trying
to understand the overall game, a high-level systems analysis will do. If
your game draws inspiration from one particular feature of another game,
your analysis might have a different focus. You can take any of the high-
level systems and break them down into multiple sub-systems and do an
analysis at that level. It works better with bigger games where there is more
going on in each system, but the principle applies in either case. Systems
are not a universal truth, they’re a lens you can use to look at the things that
matter to you.
High-Level Systems
At the top level, you can break down most games into a few high-level
systems. These usually correspond to the things we’ve been talking about
already:

1. Actions
a. Movement Systems
i. How does the player’s avatar move?
 ii. What are the controls the player uses to influence
movement?
iii. Once the avatar has moved based on player input, how does
the physics engine move the avatar? If you want to call
physics a separate system, you can. Usually not worth it on a
simple game like Centipede, but it can be done.
b. Attack System
i. See the detailed bullet list rules above. That’s the majority of
the attack system.
2. Goals
a. Score System
i. How does the player earn points?
ii. How are points displayed?
b. Combat System
i. For a big game like Halo or God of War, I would consider
the attack system to be how the player creates attacks and
how they’re presented up until the moment that the attack
scores a hit, and the combat system to be all the other things
about what happens when something is hit, and also how
enemies behave in combat.
ii. In the case of Centipede, the combat system and the attack
system are simple enough that they could be considered to
be just one system. The attack rules list above already
includes a number of combat rules that aren’t technically
about how the player attacks.
3. Challenge
a. Enemies
i. Each enemy has different rules and AI behaviors.
 ii. This includes how they appear, how they move, how they
interact with things they touch when they move, how they
die, and so on.
iii. As discussed, these relatively simple rules lead to interesting
dynamics, especially when enemies interact with other
qualities of the game such as mushrooms.
iv. Some qualities of enemy are better described as content
rather than part of the system, but we’ll talk about that in a
later chapter. For now, we’re just grouping everything here.
b. Mushrooms
i. Mushrooms have their own rules for how they work. If you
want to call that a system, you can. If you want to consider
that part of the level, you can. It just depends on the level of
fidelity you need for the dissection you’re currently doing.
c. Levels
i. There are rules for the generation of levels and the
progression of levels as time goes on.

Note that some of the lines between things are fuzzy. This is fine. If you
want to say enemy behavior is entirely contained in the AI System, great. If
you’d rather group Enemy movement and Enemy attacks as part of the
Enemy AI System, great. If you want to throw in all the other qualities and
rules about enemies and just do one big Enemy System, that works, too. It
just depends on what’s important to you and your analysis. The underlying
rules don’t change. Systems are just ways to group those rules together to
make them easier to talk about and compare. Which is why it’s hard to get
people to agree on what the systems of any particular game actually are.
System Dynamics
We can describe pretty much everything happening in Centipede with five
to seven high-level systems. If we wrote this out, we could describe the
entire game in maybe ten pages – all the rules for how things appear, move,
combine, and end. Most modern games are bigger and more complicated
than this, especially once you include all the content, but the principles are
the same.
The fun thing about systems is that those ten pages would be just the start
to a true understanding of the game. The real depth of the game comes from
how those rules – and those systems – interact. Combining two or more
rules together will often lead to one or more dynamics, each of which
creates new gameplay. Each system has complexity within it, but the true
complexity comes when they are combined.

(Mushroom System + itself) The rules for how mushrooms come and
go create the gardening goal that we discussed above.
(Mushroom System + Attack System) The way shots work combines
nicely with the way mushrooms work, encouraging players to blast
nearby ones but making far-away ones a challenge.
(Movement System + Enemy System) The trackball-based movement
combines nicely with the spiders to keep players moving around at all
times, creating an interesting frantic mood to the game.
(Point System + Enemy System) Spiders giving you more points when
you’re close to them encourages risk-taking.

Understanding the game requires an understanding of all of these

intersections and consequences. A full and complete understanding of every
possible (rule + rule) combination would be huge. Even for a simple game
like Centipede, that would be impossible to fully map out and document.
That’s why it’s important to prioritize. A good game designer can play a
game for a bit and map out the key dynamics and systems that are important
to understand. They’re not going to map out every little thing, just the top
priorities for their needs.

One possible need is Industry Analysis: If you’re just trying to

understand the game and why it’s a success, identify the top five to ten
systems and any unique dynamics that have a big effect on how this
game feels.
One possible need is System Analysis: If you’re looking at a game
because it has a system similar to one you want to make, then you need
to dig deeper into that particular system and map out more of those
dynamics. Other systems only matter in the way they relate to this
system.
One possible need is Genre Analysis: If you’re looking at a game to
better understand how a genre works, then focus on how this game
interacts with standards of the genre. For most genres, the systems are
going to be the same game to game, but there may be an added one in
some games, and each game will have its own twists on specific
dynamics to create a different feel.

System Visualization
It can be useful to create a visual map of the systems of a game.
Honestly, I don’t normally do this very often. I map things in my head
way more than I do on paper. But partially that’s because I’ve been doing
this for a long time. Early in my career, I did draw things out and draw lines
to connect them and think things through in that way. These sorts of
diagrams can be useful to communicate a game design to others, but
personally I find that doesn’t often work. As discussed, a dissection or
system outline is very personal to the perspective of its creator. Other
people aren’t necessarily going to have the same interests or goals in their
dissections, so diagrams may fall flat. Where they are very useful is in
thinking these things through. Visualizing the system can help you as a
game designer identifies systems and how they work in ways that text or
mental imagery cannot. Find the diagram that works for you and use it
when you really need to deeply understand.
Venn diagrams can be a nice way to think about how systems interact, as
seen in Figure 4.8.

Figure 4.8 Centipede represented as a Venn diagram.

The first, and maybe hardest part, is coming up with your boxes.
Identifying the systems that matter for your current analysis requires you to
go through the whole process described in this chapter to tease out that
information.
Once you have your boxes, start to think about the most important
connections. Put those boxes near each other and start to overlap them to
show where they touch. Label those intersections to clarify what happens
there. Keep doing this until you get to some interactions you might not have
considered already. What happens when mushrooms touch fleas? I bet we
could add something interesting there. This was a key part of the design
process for Where’s My Water? We knew that every new feature had to have
an interesting interaction with every other feature, so we would brainstorm
different ways each intersection could play out in the game.
One weakness of a standard Venn diagram is that representing one on
paper only allows for two dimensions. Once you’ve placed two boxes near
each other, it’s hard to squeeze in too many more near that intersection. A
true deep analysis would consider how every box touches every other box
(and combinations of more than two). But with only two dimensions, this
becomes challenging once you get past three boxes. Make sure you’re
paying attention to this in your diagram. Shuffle things up and try multiple
diagrams with multiple perspectives. Or use the diagram to spark ideas, the
document every combo somewhere else.
Flowcharts give you another way to think about the relationship between
systems, as seen in Figure 4.9. Instead of intersections, here we have arrows
showing things flowing from one system to another. If you do use
flowcharts like this, make sure you apply your standards consistently. If
bubbles are nouns and lines are verbs, stick with that throughout.
 Figure 4.9 Centipede represented as a Flowchart.

This is a great way to map out certain types of systems, such as your
economy. Use multiple arrows, where each arrow represents one type of
resource. Show where resources convert into other resources. Throw in
some numbers if you’re comfortable doing that.
Flowcharts can also be used to map other types of directional
relationships. Show where players are putting their time or attention. Show
the flow of difficulty and comprehension over time. Systems are about
relationships, and these relationships can come in many different forms.
Make the diagram adapt to your needs. Just be careful to be consistent in
how you use each element in the diagram. In the Centipede flowchart
above, circles represent objects in the game world and arrows represent
actions they can take on each other. That’s just one way to turn Centipede
into a flowchart.
It can be helpful to think about your video game as if it was a board
game, as seen in Figure 4.10. If you had to write out the rules as text, how
would you phrase them? How would you describe this system to a new
player? What diagrams and examples would help people see how this is
meant to work?

Figure 4.10 Centipede represented as a Rulebook.

Writing out rules as text is a useful first step toward creating the design
documentation for your video game. But the diagram part is often even
more useful. A good board game rulebook isn’t just text – there are a lot of
diagrams and images and examples to show how things should work, and
handle some of the more confusing cases. This gets you thinking about your
systems from more of a player perspective, which is key. Systems can get
very abstract and conceptual, so bringing them back home to a practical
experience can really help. Understanding the system doesn’t matter if you
don’t understand how it’s going to influence actual players.

Dissecting Systems
How do you turn a game into a set of rules, dynamics, and systems?
First, play the game. There’s no substitute for experiencing the game on
your own in your own head. Watching other people play or reading other
people’s analysis can provide insights, but all dissection is predicated on
you having a mental model of the game that includes the player experience
– what draws the player’s attention, what strategies seem useful, and how it
all feels. That’s hard to replicate without actually playing the game yourself
first. Ideally, you should play in the same circumstances as your players
will, but that’s usually impossible. You’re never going to be a suburban
mom who finally got the kid to sleep and is sitting down for the first time
all day, or whatever your target audience is. Unless you actually are a
suburban mom who finally got the kid to sleep. In which case, you’re a hero
and I have so much respect for you. Getting kids to sleep is hard.
Once you’ve played the game, start to break out what you noticed. What
features or game design decisions stood out to you? What made this game
interesting and different compared to others you’ve played? As you start to
break out those key features, you can start to identify which are rules, which
are dynamics, and which are systems (and which are content, which we’ll
discuss in a future chapter).
If you’re familiar with the genre and previous similar games, you
probably have at least a vague idea of what the key systems are. If not, start
to generate that high-level overview. As shown above, this can normally
flow from our definition – Actions, Goals, and Challenges. When you
identify what those are, that’s also where your systems are going to be.
Bigger games will take a little more work to do even a high-level
analysis. Many big games will have multiple different phases or states that
can include entirely different gameplay in them. So, a game like Oblivion
will have a system mapping to describe combat and exploration, but also a
whole different mapping with different Actions, Goals, and Challenges just
for conversations. And for lockpicking. And (depending on how you want
to break things out) for character progression. Bigger games have more
systems and often more games inside them.
 And games can be big in different ways. A fighting game like Street
Fighter 2 is all about combat, but any serious fighting game will have many
sub-systems within combat, for combos and blocking and hit boxes and
move chaining, and so on. These features may be present in games with
simpler combat, but in a fighting game they are a key part of the experience
and are more likely to matter. As mentioned, any dissection depends on the
perspective of the dissector.

Mapping Non-Existent Systems

So, you’ve read the past few pages and now you’re an expert at dissecting
system. So, you go to start working on a new game idea, and the first step is
to play the game. Wait a minute! The game doesn’t exist yet, so how can
you build a dissection off of playing it? That’s crazy talk.
Dissecting games that don’t exist is a critical skill for a game designer.
And it can be hard.
The best way to do this is to play the theoretical game in your head.
When you’re thinking about a new game concept, there’s something that
draws you to it. There’s something that you can picture that makes it
compelling. Take that picture and expand on it. Close your eyes and think
of that amazing new moment, then keep playing past that. Add on the rest
of the rules and systems and features that are needed to make the game a
complete experience. Map that like you would a real game.
Once you’ve got something to start with, you can reverse engineer the
systems of the game.

If your picture of the game is a single moment, you probably know

something about the actions of the game, and the moment-to-moment
goals and challenges. Think about how those moments can be
expanded into a full level, and many levels, and a full story, and a full
 game. Shamelessly steal progression systems from other games. The
moment is what makes this game shine, so it’s fine if the long-term
systems are ones people have seen before.
If your picture is a story or character or mood, think about how that
can be expressed through player actions. Build a core game loop that
strengthens the thing that inspires you. If the thing that inspires you
isn’t a player interaction, maybe consider if it would be better to turn it
into a book or movie or something else. The inspiration for a game can
come from anywhere, but if interaction isn’t at the heart, it might be
better served elsewhere.
If your picture is something outside of the core game loop – a
progression system or deep strategy or new play pattern – then map
out that feature but quickly get back to the core loop. Everything
depends on the actions and goals and challenges the player experiences
over and over. If your idea for a great new progression system doesn’t
have fun core gameplay, you’re in trouble. But in a case like this, you
might look for a popular genre where your core loop is similar to
previous games, and then add your exciting new twist. The first games
to let the player manipulate time didn’t also come up with completely
new gameplay to surround it – they were first-person shooters or side-
scrolling adventure games or other classic core loops with time
twisting as the exciting new part.

This is part of why playing a lot of games is important for game designers.
You need to be able to identify the systems and rules and dynamics of other
games so you can identify them in yours. And then fill in the rest of your
game with things that are proven to work. If you understand how rules
combine into dynamics and systems, you’ll be able to start to predict how
your exciting new bits will combine with the classic old bits. This is a big
part of the work of a game designer. Predicting how rules will interact is the
value that game designers provide to game teams. No designer can do that
perfectly. But a good designer can anticipate problems and design optimal
solutions to reduce the workload on the rest of the team. Knowing how to
design enemies and levels to fit your game’s mix of new and old reduces
the amount of trial-and-error iteration and puts the team on a fast track to
success, or at least improves the odds by skipping a few of the early rough
iterations.
Systems: To Be Continued
Systems are very important, and I want to keep talking about them. But I
also want to dissect a lot of games in this book. So, we’ll come back to
systems and talk about them in more detail in a later chapter.
OceanofPDF.com
Chapter 5
Settlers of Catan / Uncertainty
DOI: 10.1201/9781003346586-5

Catan – Gameplay Dissection

Settlers of Catan (also known as just Catan) is a great game.
Catan is a board game based on building settlements and managing
resources on a randomly generated island (Figure 5.1). Catan is one of the
best-selling board games of all times, selling over 25 million copies since
its release in 1995. Some credit Catan with inspiring the golden age of
board games that we are currently experiencing – many of the successful
games that followed it have similar qualities. I worked at Mayfair Games in
1995 and 1996 and helped Jay Tummelson (and many others) bring Catan
to America, and I wrote the first US rule book.
 Figure 5.1 Catan photo, by the author.

Verbs
This is the first board game we’re discussing, but the fundamentals of
dissection are the same. All games are about interactivity, including board
games, so we start with the verbs that define that interactivity.
The key verb in Catan is “Build”. The player builds settlements and
roads that provide resources and score victory points. “Upgrade” is a
closely related verb, as upgrading a settlement to a city is just building with
some restrictions and costs. And buying cards is just building with variable
outcomes. The heart of Catan is turning resource cards into stuff that gets
the player more resource cards.
“Trade” is another key verb in Catan. Players aren’t ever required to
trade, but success in the game comes largely from identifying the correct
resources and trading with other players (and the ports) to acquire those
resources. This could also be framed as “Negotiate” or “Beg” or some other
phrasing, but it’s still about trading intelligently.
“Gather” is almost a verb in Catan, but it’s not exactly what we’re
looking for here. The player has no action that they can regularly take to
directly gain resources. Building settlements and cities increases the
player’s gathering power, but the actual production of resources is
determined by the roll of the dice, not the player’s choices. “Gather” is
definitely something players do, and it’s certainly a goal players have, but
it’s not a player-directed action like the Verbs we talk about. A first pass
verb list will often include a number of goal verbs along with the action
verbs.
There are a few verbs centered around the Robber. Triggering the robber
is not a player action unless they’re using a card. But once the robber starts,
the player has to “Place the Robber” and “Steal” with relevant choices as
part of that. And some players may have to “Discard” if they have too many
cards. These are all legitimate meaningful verbs because the player has to
make a decision with consequences. But for the purposes of most types of
dissection, these can be grouped as “Robber” verbs. They only happen
occasionally during a game, so they’re not the core gameplay loop.
Similarly, “Play Card” is a verb with lots of different specific sub-verbs,
but we can group them together for most analyses.
Things that are not verbs in Catan:

“Roll” – This is the interface for gathering, not the action itself. In
Centipede, we don’t call “Press Button” a verb, so this doesn’t count
either.
“Settle” – There’s no actual settling occurring, other than how “Build”
is related to this word. If you want to play a game with its action verbs
in the title, try Minecraft.

Things that are verbs in Catan but aren’t active parts of the core gameplay
loop:

“Assess” – Knowing what spots matter and will provide the best return
is key to success in Catan. Many games are decided by thoughtful
initial placement of villages.
“Compute” – Knowing the probabilities and how the math works out
helps with assessment, but it isn’t an action the player does, it’s
contained within Assess.
“Defeat” and “Win” and “Score Points” – All verbs that players may
do, but they’re tied to achieving goals rather than specific actions
players take, so they’re a different category of verbs.
“Gather”, as mentioned, is a goal verb but not an action verb.

Goals
The goal in Catan is technically to acquire victory points.
 But in order to do this, the player needs to build an empire. Gaining
victory points is done by building. And building is done by gathering
resources.
The gameplay goals of Catan align nicely with what the player naturally
wants to do, which is important in any game. Building little empires is fun.
Gaining more resources is fun. When the action and the goal don’t flow
naturally together, you can get players who think they’re doing well, but
then realize too late that they’re not actually doing the right thing. But in
Catan, everyone wants to build more things, so rewarding that impulse is
good.
Building has multiple goals within that create a nice set of garlic layers
(Figure 5.2). Each turn, even other players’ turns, provides an opportunity
to gather resources. Which is a clever trick to get people to pay attention
throughout the game, also seen in games like Monopoly. Getting the right
resources leads to the opportunity to build. Depending on what the player
wants to build, they may need to build other things first. Can’t have a city
without a village, and (usually) can’t have a village without new roads.
There’s always something new the player wants to build or gather to reach
the long-term goal of points.
 Figure 5.2 Catan goals as garlic layers.

The pressure between player impulse and game goals can be seen with
the Longest Road card. Players like building the biggest and best of things,
and this card is a nice way to reward that. But I’ve been in plenty of game
sessions where one or more players focus on the Longest Road card to the
detriment of their overall victory (I’d like to say hi to my kids here). This is
especially true when players get competitive, trying to top each other with
increasingly convoluted roads. While there’s a reasonable argument that the
game is asking you to exercise your strategic judgment to avoid this
problem, it can lead to feel-bad situations where players who focus on roads
are disappointed at the end of the game, which is not desirable for a game
designer. If the player thinks their actions are going to lead to victory, the
game should find a way to make that happen.

Challenges
So, the player is trying to build an empire to gain victory points.
What makes this challenging?
 There are no enemies. There are no pits to jump over or bugs attacking
you. There’s barely any conflict between players. So, challenge and
uncertainty in Catan (and many board games) come from a different place.
The primary barrier to success is the economy. The player can’t build
whatever they want – they have to build based on various restrictions,
including the resources needed for each build. The rules tell the player that
they can only do the fun thing in certain ways – you can only build here, not
there; you can’t build until you have enough of this; you need this before
you can build that. These rules tell the player when resources are created,
and this production is highly variable based on die rolls. The player has to
be creative to find ways to get the most production possible out of these
limited resources and apply their resources to the best methods to gain more
in the future. Limited resources and restricted building rules are the true
enemy in a game like Catan.
There are other subtler limitations on the resources in the game. There
are only so many places to build on the island, and many of them are
terrible for resource production. You can’t build close to other villages. You
have a limited number of each building object. Most of the rules in the
game are restrictions put on the player’s ability to build whatever they want
whenever they want. Limits come in different forms.
The other major barrier to success is the true enemy in many games –
other people. Navigating resources only gets you so far. The player needs to
manage other players both in the standard ways – long-term strategy,
planning around future actions, etc. – and in personal and persuasive ways.
Convincing everyone else that you’re not a threat and it’s OK to trade with
you (again, I’d like to emphasize how much I love my kids) can be as
important to victory as careful resource planning.
Core Loop
This creates a simple diagram to understand how the core verbs and the
non-verb gathering action create the core loop of the game. The player
gains resources that they can spend to increase their ability to gain
resources in the future. These actions also increase the player’s victory
points, and eventually that value reaches a critical mass and that player
wins.
This is a common and successful loop for lots of resource- and economy-
based games (Figure 5.3). It’s essentially a variant of the power fantasy
seen in the character progression systems of many video games, but in
economic terms. The player keeps getting bigger and exerting more control
over the world around them and then invests that power to gain more power
in the future. In a combat-driven game like Shovel Knight, the player battles
enemies to get gold, then invests gold in RPG stat upgrades so they can get
more gold in the future. In Catan, players gather resources to build villages
to gather more resources in the future.

Figure 5.3 Catan Gather and Build flow.

Note that all of the uncertainty is on the gathering side. As with many
economic games (SimCity (1989), Agricola (2007), Civilization (1991), Pet
Simulator (2018), The Game of Life (1960), Roller Coaster Tycoon (1999)),
the challenge for the player is coming up with the resources. A player with
resources has clear verbs with clear focus. There is intellectual challenge in
picking the right building choices, but no uncertainty in knowing what will
happen when a choice is made. If the player who wanted to build had to
spend their resource cards and then had to roll to see if the village was
successfully built, that would be a very frustrating failure. It’s better to
provide variable rewards than variable outcomes.
Every game can be mapped to verbs leading to a goal with challenges in
the way (Figure 5.4). This simple flow diagram can be expanded (Figure
5.5) to show how a Catan player turns that core economic engine into
victory. The player has to Gather in order to Build, so there’s a simple
dependency there. And within Build, there are specific player actions that
also have dependencies. The player can’t build cities without villages to
upgrade and can’t build villages near each other, requiring roads. This
mandates a standard flow that makes building interesting.

Figure 5.4 Catan flow: Simple diagram of Actions, Challenges,

and Goals for Catan.
 Figure 5.5 Catan expanded: Same diagram with Gather before
Build and more options inside Build.

And from there, most build actions feed back into both the gathering
system and the point system (Figure 5.6). Villages and Cities are worth
points, moving the player directly toward victory. But they also give the
player new ways to generate resources or to generate more resources.
Similarly, cards have a random chance of either providing direct victory
points or providing economic benefits. Roads are the only building action
that doesn’t directly provide either victory points or other resources, unless
the player manages to build the longest road.
 Figure 5.6 Catan ConnectedPlus: Same diagram, showing links
between build actions, Gather, and Points.

Catan – Non-Gameplay Dissection

History
Settlers of Catan came out in 1995 in Germany, and 1996 in the US. It was
an immediate success, winning the prestigious Spiel des Jahres for 1995
(Figure 5.7. shows the 2021 Spiel des Jahres winner). Which isn’t a big deal
to most people in America but is a very big deal in Germany and to board
gamers around the world.

Figure 5.7 Spiel des Jahres. Note: The image is licensed under the
Creative Commons Attribution-Share Alike 4.0 International
license. The author is Thomas Ecke für Spiel des Jahres e.V. No
changes were made.
 It’s hard to understate the effect of Settlers of Catan. Germany has a long
tradition of great board games. Post-war Germany had a large market for
thoughtful family games that can maintain the interest of both children and
parents, with short playtimes and non-violent themes. Games there were
and still are bigger in Germany than they are in most countries, especially
in the US. In the US, even today, the winner of a prestigious board game
award would get a little notice and might sell a few extra copies. In
Germany, the winner of the Spiel des Jahres sells in the millions. Catan was
just one of the first such games to see big success in America.
America had a board game industry before Catan, but it was smaller and
less inviting. There were the traditional games that were sold in toy stores
and department stores – games like Monopoly or The Game of Life. And
there were the hardcore gamers who played war games and Dungeons and
Dragons (1974) and small quirky games from Steve Jackson Games or
Cheapass Games. When Catan came out, the hobbyist game industry was
going through a growth period due to the explosion of Magic: The
Gathering a few years earlier (e.g., that’s how I got a job in the games
industry). There was a revitalized hardcore audience, so when Settlers of
Catan came out and offered them a game they could play with their non-
hardcore friends, everything came together. From there, Catan spread
across the world and became a huge influence on games to come.
Catan heralded a golden age of board games that continues today. I think
it’s reasonable to assume that even if Catan never existed, the world would
still have discovered the beauty of German board games and this
renaissance would have happened, but in a different form. In this fictional
scenario, my money would be on Carcassone (2000), but El Grande (1995)
or even Manhattan (1994) could have done it. Or just about anything by
Reiner Knizia – Tigris and Euphrates (1997)? Medici (1995)? “What if” is
fun. But the unique qualities of Catan have helped shape a board game
culture that respects randomness, respects non-violent play, and has strong
system game design principles at its core. Gamers owe a lot to Catan.

Monetization
Catan sells in a box. A rather expensive box, compared with many other
games, especially the mass market toy store games of 1996. In some ways,
this becomes a selling point. The game is high quality. The pieces are high
quality – sturdy wood instead of flimsy plastic. The box is heavy. If you
invest in this quality product, you can play it over and over again, so while
it’s a higher initial investment, it’s a great deal in the long run. That’s not to
say that a higher price point makes it sell MORE copies, but at least there’s
a legitimate reason for the customer to consider it.
Catan also sells things beyond the box. There are expansions and video
game adaptations and merchandise. You can buy T-shirts with sheep jokes
on them. I take great pleasure in seeing Catan jokes. I made sure to include
a few simple jokes in the original American rulebook to set the right mood
and get people to see this as something light and fun, even if the game
requires thought. I like to think it worked, and that it helped at least a little
in the game’s success. But the core of the business is that box.
In some ways, Catan set the standard for monetization in the modern
board game industry. Appeal to a broad audience. Keep the game short –
under an hour. Invest in production quality to make the game feel valuable.
There are plenty of successful games that cater well to smaller audiences,
but Catan showed that a broadly appealing game can make a significant
amount of money over multiple decades.
Senses
Graphics mean something different for a board game compared to a video
game. In a board game, it’s about the look of the game and the components
and the box. Board games even get an extra sense – touch. The feel of the
components is a big part of the sensory experience of a board game, and
people like something that feels hefty and solid and real. A good game has a
good feel and mood, and the senses set the tone for that.
Visuals also help explain how to play the game – a good game should
guide the player to understand everything without even touching the
rulebook. One of my greatest regrets as a board game designer/developer is
not pushing for better visual clarity on the first US edition of Settlers of
Catan. I was young. I was fresh out of school. I didn’t know any better. But
as the gameplay developer on that edition, it was my responsibility to guide
the talented artists at Mayfair and I didn’t do a great job. Compare the first
US edition with the later editions and you can see the difference – the colors
and images on the hexes and cards in my version were too similar and it
became hard to tell the difference in some cases. That’s a problem. Not a
big enough problem to prevent the game from succeeding, but looking back
I know I could do better with what I know now.

Narrative
In Settlers of Catan, the player is settling on an unexplored island,
competing with other players who are also building on the same island at
the same time. The story is not deep and not specific, but evokes historical
periods of growth and construction, using generally European resources and
society.
In recent years, the board game world has been wrestling with some
difficult questions. Many of the most popular board games are based on
themes of expansion and exploration and discovery. These make for great
gameplay and inspirational goals, but when combined with historical
themes, they raise some questions. Historical exploration and colonization
were usually based on exploitation and destruction. In reality, Europeans
settling on a “new world” had consequences for the people who lived on
that “new world” before them. Games that build on historic and economic
models based on the positives of never-ending growth are increasingly
being called into question.
Games are systems. Games boil concepts down into understandable bits
that players can work with. Games that try to recreate reality inherently
need to simplify reality to make that work, and it’s worth considering the
biases and assumptions inherent in that simplification. All human endeavors
include bias and assumptions, so that’s not a flaw. But it’s similarly not a
flaw to question this and poke at it with a stick.

Understanding why we use such settings, and why they can be

unsettling, is necessary, but I don’t intend to condemn it and I don’t
think we, game designers, should stop making boardgames that
caricature the Orient or Ancient Greece, no more than we should stop
making games that caricature barnyard or jungle.
– BRUNO FAIDUTTI

Catan presents an island for the player to settle. There’s no mention of

previous inhabitants or a historical location. This lack of specificity allows
the player to set these sorts of questions aside and not worry about them.
Klaus Teuber, the designer of the game, says he was inspired by Vikings
reaching uninhabited Iceland (Raphel, A (2017)). Catan is meant as a fun
moment of escapism, not a deep rumination on history. Not every piece of
fiction needs to bring up and resolve every possible complication implied
by its premise. Superhero comics don’t need to explain why building suits
of armor with vast military power doesn’t change the geopolitics of the
world. Illumination’s Minions franchise never needs to tell me what they
were doing in the mid-1900s. Part of the joy of games, and movies, and
books, is the escapism, and it’s not every game’s responsibility to address
every social issue.
But I do love games that do confront social issues head-on. The lack of
colonial specificity in Catan is as an opportunity for future games to dig
deeper rather than a problem with Catan itself. Simplified models of the
world that allow escapism are great and are never going to go away. But I
would love to see more games that wrestle with the tough problems of
history to make players think. History is messy, but it’s also a great source
of inspiration.

Meaning
Catan is about expansion and growth.
How do you take an irregular flow of various resources and turn that into
the things you want?
Can you build an empire out of a dribble of scraps?
When life gives you sheep, make sheep-ade. Or on second thought, don’t.
There’s also an interesting discussion of cooperation and competition.
Most games of Catan involve a lot of trading, but at some point, a player
starts to take the lead and other players become reluctant to trade with them.
Players work together collaboratively, up to a point. Having a strictly
defined zero-sum goal encourages this behavior, even if that’s not how
reality works. I like to think that the early game of Catan says nice things
about humanity, and the end game is more about the artificial systems of the
game. But I’m an optimist.
Uncertainty
We’ve talked a lot about verbs and a bit about goals. Now let’s talk about
the middle section of the game moments: challenge.
Challenge is really where the fun lies in a game. Toys are pure
interactivity with no goal, and that can be fun. But the difference between a
toy and a game is a goal. A game points you in a direction and makes it
difficult but not impossible to get there. The fun of a game is the struggle. If
you know you’re going to get to the goal, it’s too easy and thus no fun. If
you know you’re not going to get to the goal, then it’s also no fun. Fun is a
balancing act.
This is where games resemble the flow state from psychology (Figure
5.8). For something to achieve a flow state, it needs to be difficult but not
impossible. Humans want to push themselves against difficult but no
impossible odds. We play games because they provide artificial challenges
that let us practice and play with difficult things but with low stakes.
Remember our definition of games – goals and interactivity but inherently
separate from the real world. Without some uncertainty, this artificial
challenge isn’t challenging.
 Figure 5.8 Adapted from Flow by Mihaly Csikszentmihalyi.

Types of Uncertainty
Maintaining the balance of fun requires uncertainty. If the player knows the
outcome, the game collapses into either too hard or too easy and breaks
down. Not knowing the outcome is what makes the outcome interesting.
Different games create uncertainty in different ways.

PEOPLE: Multiplayer games, including most board games, have it

easy. Other humans create a lot of uncertainty. Understanding the
actions of other humans is so difficult we have whole fields of
academia devoted to it (psychology) and sub-fields within other major
fields (game theory as a sub-field of economics). Other people provide
plenty of uncertainty to drive plenty of games, from Rock Paper
Scissors to Diplomacy (1959).
HIDDEN INFORMATION: If something is certain and fixed but the
player has no way to know it, it’s uncertain to them. Video games rely
on this for the fun of exploration – when I turn down this hallway in
Doom (1993), I don’t know what I’m going to find, even if every
player sees the same hallway once they get there. Shovel Knight is
similarly built on the fun uncertainty of exploration. Many party games
such as Werewolf rely on information that some players know but
others do not. Managing which players get access to which
information at what time is an important way for game designers to
balance the flow of fun, especially in multiplayer games.
PHYSICAL SKILL: Most action games create uncertainty from
physical actions that are hard to perform precisely. For example, most
sports require difficult physical actions such as throwing or kicking or
running where it’s hard to predict if a specific person can perform that
 action correctly or not. Basketball is fun to play or to watch because
you never know if that shot is going to go in.
Some physical actions are more predictable than others, making
them less about uncertainty and more about pure skill. If you
know that you’re playing a tug of war against a body builder, it’s
not very uncertain. But some physical actions are harder to
predict, like balancing a wooden block on other wooden blocks,
bringing uncertainty back to the forefront.
MENTAL SKILL: Some games create uncertainty by giving the player
so much to think about or process, making it effectively impossible to
look too far ahead. This is actually quite similar to the physical
challenges that action games provide, but less obvious – in either case,
the uncertainty stems from the player’s ability (physical or mental) to
complete the challenge, which is up to the player to provide. You could
argue that all people-based uncertainty is just a variant of mental skill
uncertainty, with the ability to predict people being the skill.
RANDOMNESS: Many games create uncertainty by using
randomness. Dice, decks of cards, or just random number generators
give you a certifiably random result. More on that in a moment.

Randomness is the purest form of uncertainty. Other uncertainties stem

from things that have patterns or predictability but are hard to know – other
people, individual skill, or just things you haven’t seen yet. But if
randomness is done correctly, there is literally no way to know the result
ahead of time. All games need uncertainty, but randomness is the king of
uncertainty.
Uncertainty in Catan
You can see many of these types of uncertainty playing out in Settlers of
Catan. The balance of different types of challenges based on different types
of uncertainty is part of what makes Catan a great game.

PEOPLE: Understanding your opponents and adjusting your strategies

to account for other players’ actions is a critical skill in Catan. This is
true when competing for the top production spots but is even more
clear in trading. Trading is a highly social activity where persuasion
and pleading can advance your cause in the game. And on the other
side, knowing when to give in to someone else’s pleas and when to
stand firm often decides the fate of a game.
HIDDEN INFORMATION: Catan has multiple types of hidden
information. Any deck of cards relies on hidden information, both the
order of cards in the deck and also the cards in other player’s hands.
The Victory Point cards are specifically designed to play off this, as
they can be revealed at the last minute to create a sudden win. There
are often signs that another player is holding onto a Victory Point card,
but there’s always some uncertainty that makes victory totals more
interesting.
PHYSICAL SKILL: Physical skill is absent from Catan, other than
being able to balance the city tokens on top of the road tokens to
amuse yourself when someone else is doing a long trade negotiation
during their turn. Games based on physical skill are usually grouped
into the genre of games we call “Sports”, with a few exceptions like
Marbles or making a goal post with your fingers for someone to flick a
paper triangle into.
 MENTAL SKILL: Catan is very much a game of mental skill. A good
player will have a higher winning percentage than a bad player. They
won’t win 90% of the time like they would in a game like chess, but
they’ll win 65–75% of the time against less skilled players. There are
lots of individual decisions that have a significant impact on the final
victor. The starting layout is the largest and most obvious of these. A
bad starting placement can be very difficult to recover from, making it
a key way to distinguish mental skill.
RANDOMNESS: Catan has lots of randomness, but the randomness is
used in the right ways. See below for a detailed breakdown.

How Uncertainty Works

To understand the role of uncertainty, let’s go back to a previous diagram,
modified for Catan (Figure 5.9).

Figure 5.9 Input—Output diagram, Catan version.

This diagram is beautiful and perfect but does not include everything.
Inside the Player Action box, there is more going on. When the player
chooses an action, there is normally more than one verb available to choose.
And even a simple verb may have multiple different options on how to use
it.
For example, in a game of Catan, many turns have a core decision the
player has to make. After the player rolls the dice and (maybe) gathers some
resources, what other action or actions do they want to do? They can always
pass, but with the right resources they could build or buy various things.
And there’s always the option to try to trade with other players. Most turns
in Catan end in a pass, but the other options are always on the player’s
mind (Figure 5.10).

Figure 5.10 Choices in a turn during a game of Catan.

Each of these choices has the player choosing a verb that is allowed by
the game of Catan – the verbs we start our dissections with. But the other
parts of the dissection are also present.
When the player is making a choice, they’re doing so because they have
a goal in mind. There is something the player wants, and the action is their
attempt to achieve it. So, a player who passes their turn in a game of Catan
may not have any resources to build with, or they may be doing it because
they want to save up their resources to build something bigger in a later
turn. Similarly, a player who builds something is usually doing so because
the game gives them rewards for doing so – victory points, more resources
on future rolls, or blocking off other players. Different players may be
motivated by different goals even when they’re taking the same action.
And between the player’s action and the eventual goal, there are various
challenges, either direct barrier like enemies in Shovel Knight or indirect
barriers like the economic system of dice rolls in Catan. These challenges
are what make the connection between the action and the goal uncertain and
interesting.
This is all happening inside the Player Choice box in our diagram. The
player can choose an action, and the game result follows directly from that
action. If I discard four cards to build a village, the outcome is always that I
build that village. There are lots of possible player actions, but each player
action ends with only one response from the game rules. Even if the player
passes, the outcome is always certain – the play passes to the next player.
For the sake of simplicity, the next few diagrams will focus on two options
– pass or build.
A player’s choice between pass and build can be represented by adding a
branch to the input/output diagram we’ve used before, as seen in Figure
5.11.
 Figure 5.11 One input can lead to multiple possible outputs.

Adding a branch changes the diagram. The player may go down one path
with one game state as the result (build a village and now the board layout
is different) or down another path that results in a different game state (pass
the turn and have more resources in hand). Every game of Catan is the
result of an endless series of branches like this, with an endless possibility
space of different game states from every choice.
Other events in the game work differently. At the start of each player’s
turn, they are required to roll the dice. There are no options there, no
variation for the player to choose. But the outcome of this action is variable
based on a randomizer – the dice. So, one input produces many possible
outcomes (Figure 5.12).
 Figure 5.12 One input can lead to many outputs.

As you can see, this can lead to explosive possibility spaces. Even just 11
outcomes make it hard to manage all the possibilities (Figure 5.13).
Thankfully, our brains are good at managing this sort of possibility space.
In most cases in Catan, the results can be boiled down to a binary good or
bad. Did the player produce resources or not? There is definitely more
nuance to this outcome, but for the sake of a simple analysis of this game,
boiling the results down to two possibilities makes it manageable.
 Figure 5.13 Many outputs can seem like only two based on how
the brain processes it.

Trimming choices like this is an important thing to consider when

thinking about the player’s choices. It’s OK to have lots of options if some
of them are obviously terrible choices. And there are other ways players can
remove choices from the “meaningful” list:

Options that only make sense at certain times

In Ticket to Ride (2004), it’s generally not a good idea to gather
more route cards unless either (A) it’s early in the game and you
think you can handle more, or (B) you just finished or are about
to finish a route or two. On most turns, this action isn’t even
considered.
In our Catan example, I don’t need to consider whether or not to
play my Knight if the robber is currently in a place I like,
especially if I just put the Robber there recently.
Options that only work when you have the right resources
 In a game of Magic: The Gathering or Marvel Snap (2022), you
can ignore cards that cost more than the resources you have that
turn. But an expert player will consider how those options might
influence their current play. That complicates things, but only for
players who choose to complicate things, which is great. This is
an example of “lenticular design”, a phrase coined by the lead
designer of Magic: The Gathering, Mark Rosewater, to explain
decisions that are interesting for advanced players but don’t even
register as a complication for beginning players (Rosewater,
2014).
In Catan, I don’t need to consider most of the build options for
most turns because they simply aren’t possibly with my current
resources.
Options that only work when combined with other choices
In Pandemic (2008), I can’t use the Treat Disease action unless
I’m in a city with disease cubes there. So, I can ignore that much
of the time.
In Catan, I might have made a deal with another player so I don’t
consider placing the Robber on that space. This can also go the
other way and add complexity to my decision, if there isn’t a clear
deal and I want to make one in the future, or if there is a deal but I
might betray them.

Game State
Once the player has condensed their choices down to a manageable number,
they pick an action and change the state of the game. These changes add up.
Each time the player makes a choice, they change the state of the game as a
result of their choice. And each time the dice are rolled, they change the
state of the game as a result of randomness.
In a game with clear winners and losers such as Catan, one way to
consider the state of the game is who is winning. Each action and result that
changes the game state in any way can be tracked based on whether it
increases or decreases the player’s chance of winning. Actions that add
victory points move the players closer to a win, as do actions that gain them
resources either now or in the future. The same can be said for each random
result (Figure 5.14). The player’s goal is to get to the win, and change in
state is designed to move in that direction.

Figure 5.14 Consequences lead to conclusions.

Eventually, the entire game boils down to a series of choices and random
events, each moving the state of the game one way or the other (Figure
5.15). Until eventually one player wins and the other players lose.
 Figure 5.15 Consequences lead to more conclusions.

This is how the actions, goals, and challenges of a game add up to a full
experience (Figure 5.16). These are the core of the game and result in the
final outcome.

Figure 5.16 Choices lead to conclusion.

If the player knew the full results of each of their actions ahead of time,
the choice between two actions would always be trivial. A game where
every outcome is certain is not really a game at all. Every game does
something to ensure that players can’t just breeze through every decision.
As Sid Meier said, a game is a series of interesting decisions. If the
decisions are known, they’re not interesting.
For Catan, most of that uncertainty comes from randomness. The player
can make a decision to build or pass, but the player doesn’t know what rolls
are going to occur in the future, so the outcome of the player’s action is
unclear, and thus interesting. Uncertainty creates a cloud of uncertainty that
prevents players from perfectly predicting the outcome of each game action
(Figure 5.17). That uncertainty keeps the game unpredictable and thus fun.
A good game has many of these clouds to ensure that players don’t worry
about perfect optimization. Knowing who’s going to win is not fun. Waiting
for 45 minutes for the other player to compute who is going to win is even
less fun. Clouds of uncertainty remove the incentive to over-optimize.

Figure 5.17 Uncertain actions – luck.

Trading is similar. Trading is an uncertain action that may or may not

benefit the player. But in this case, the uncertainty comes from something
other than randomness – it comes from other players (Figure 5.18). The
player knows how the trade will benefit them, but does not know what the
other player has in mind. Good players know to not only consider the
benefit of a trade but to also consider the cost – the benefit to the other
player. If the other player gets more out of the trade than the first player, the
first player may want to reconsider. This gets even more complicated when
we consider the effects of all the players. In Catan, a trade that is better for
the other player may be a good idea if the other player is nowhere near
winning and all the first player needs is a small benefit on their own side to
secure a win. This is the type of uncertainty that makes games like chess
fun. Chess has no randomness, but every player action results in an
interesting array of possible responses from the other player.

Figure 5.18 Uncertain actions – people.

The final type of uncertainty seen in many games is skill. Catan doesn’t
have any player actions where the result depends on the player’s ability to
move a mouse precisely, or run quickly between bases, or wiggle a joystick
at just the right time, or physically overpower another person. But plenty of
other games are filled with actions where the results are uncertain due to
strength or dexterity or other demonstrations of physical or mental skill
(Figure 5.19). These actions are uncertain because the player doesn’t know
if their skill is enough to pass the challenge.

Figure 5.19 Uncertain actions – other games.

Clouds of uncertainty come in many flavors, but all of them have the
same effects – keeping players focused on the moment and ensuring the
game retains the necessary uncertainty.

Benefits of Uncertainty
Uncertainty Tends to Avoid Analysis Paralysis
Analysis paralysis is a term originated by H. Igor Ansoff in 1965 for
corporate strategy, but used by gamers to describe cognitive overload.
When a player has too many options, or too many factors to consider for
each option, that player can get overwhelmed. They may take a long time
trying to analyze each option, or they may just shut down and give up.
Neither is a good result. As a game designer, your goal is to create decisions
that are interesting for the player. That means each choice needs a certain
level of challenge, but not so much that it becomes overwhelming.
 At any point in a game, the player has a certain number of options on
what to do next. A simple choice might only have two options. You can
build a whole game around this level of decision-making. Love Letter
(2012) is a successful and fun game where the player never has more than
two options (ignoring targeting and a few other secondary decisions). But
most games have more. If you dissect a bunch of games, you’ll find that
most of the successful games for average gamers try to avoid giving the
player more than about five options to consider at any one time.

Most card games give each player a hand of about five cards
Magic: The Gathering: seven cards at most, but on most turns it’s
actually five or less
Robo Rally (1994) gives the player up to nine cards, but each is
very simple
Dominion (2008) gives the player five cards each turn, but with
ways to draw more
Some games give a player a list of things they can do each turn,
usually three to four actions with three to five choices for each
Pandemic gives the player four actions per turn with eight options
(but many of these are similar or rarely used)
Ticket to Ride gives player one action with three options (but one
is rarely used)

This aligns nicely with psychology. Memory research generally shows that
people can store about seven things in memory at one time (Miller, 1956).
There are ways to get around this, and there is newer research that questions
the details of this. The actual number might be smaller. But the general
concept that people can only store a finite number of things in their head at
one time is still generally accepted.
 So, if the human limit is seven (or maybe a little less), it makes sense for
most games to err on the side of caution and go with about five things to
consider at one time.
In a game like chess, the number of possible game states is more than
seven. Even if the player can manage to trim down the number of possible
options as discussed before, the possibility space still gets very big very fast
(Figure 5.20). For many people, thinking this far ahead is no fun. Chess is a
popular game, but for many people playing at a high level is more
headache-inducing than fun. Most people play games for a light good time,
not for grueling intellectual intensity. And there’s nothing wrong with either
preference.

Figure 5.20 The future is busy.

And even for players who enjoy grueling intellectual intensity, the
process of doing so can create analysis paralysis. There’s a reason high-
level chess uses timers to keep players from spending days on each turn. If
you want to reduce analysis paralysis, you need to stop the player from
analyzing too far in advance. Ideally, you want to discourage the player
from worrying about more options than they can keep in their head.
Randomness is great for this. If you put out a branch that the player
literally cannot predict, with more possibilities than it would ever be
reasonable to compute through, the player is strongly discouraged from
thinking too hard about the future.
It’s like you put up a barrier in the diagram and prevent the player from
seeing past that (Figure 5.21).

Figure 5.21 The future is blurry.

The dice have too many outcomes to reasonably predict. And when the
player draws from the development deck, the possibilities are even harder
to track. This effectively puts a limit to how far the player can predict. This
can also be true for other types of uncertainties, if there are enough
possibilities. In a game of pool, there are enough skill-based nuances to
where each ball moves that it’s impossible to predict specific turns far in
advance. Pool is not a game of long-term strategy.
 You can see this clearly in highly random games like Munchkin (2001) or
Fluxx (1997). There’s no reason to plan out a long-term strategy in these
games because there is no way to predict the state of the game even one
turn in advance. And even if you could, the game doesn’t give the player
verbs that will affect that future state – actions are all about the current
state, so the player stays focused on that.
Catan strikes a nice balance here. There isn’t value in predicting the
moment-by-moment play out too far. The player can’t know exactly what
they’ll do on their next turn, but they can plan where they want to build
next. Building is a flexible enough goal structure that it can handle some
randomness and still feel good to complete. Players can predict how general
trends are likely to play out over multiple turns. And over multiple turns,
the dice are more likely to conform to predictable mathematical patterns.
Players know enough to know that if a player tends to produce a lot of
wheat, a wheat port is a good idea, even if they don’t know exactly when it
will pay off.

Uncertainty Makes Games More Varied

Good games have lots of variety. Players enjoy it when they can play a
game many times and/or for many hours and still find something new and
interesting each time. Replayability is a key metric for measuring the
quality of games. For linear games, it’s less about replay and more about
filling hours of fun. But it’s still the same principle. Variety makes games
good.
All of the forms of uncertainty we have discussed are also source of
varied outcomes. A variant of Catan where each tile produced resources on
a fixed schedule could replicate the same probabilities and frequencies of
the current game. But part of the fun of Catan is that crazy things can
happen. In most games of Catan, the dice are rolled often enough that the
probability curve starts to take shape, but not often enough that it always
collapses down to the ideal form. If every game of Catan involved hundreds
of resource rolls, then things would average out over the course of a game
and players could depend on probability to work out by the end. But with
the current game, odd things can happen. I still remember the game of
Catan where we literally never rolled a six. That game was memorable and
fun and interesting and while predictability might sound nice, I’d rather
occasionally lose like that and come out with a good story.
You can see this in games like Dungeons & Dragons. Tabletop role-
playing games would be pretty boring if attacks were always the same.
Rolling dice to determine success makes things a little different every time.
When a critical hit is rolled, it’s a big moment that players can celebrate.
This gives players different things to react to each session.
Mixing things up doesn’t absolve the game designer of responsibility for
making things interesting. Randomly generating boring enemies is still
boring. Maybe a little less boring, but still boring. The range of possible
results needs to have enough meaningful variation so the randomization
matters. But mixing them up through randomness lets the game stretch
those elements out longer. To make the best variety, you need quality parts
and the right amount of mixing.

Benefits of Randomness
All forms of uncertainty provide a way around analysis paralysis. And they
all provide good variety. But to further understand the pros and cons, we
need to break uncertainty into two main camps. Not all uncertainty is due to
true randomness.
Skill is the type of uncertainty where the root underlying the uncertainty
is something solid. There’s an actual difference between the players that is
hard to see but is actually there.
Randomness is a type of uncertainty where the root underlying the
uncertainty is not based on anything. It’s just uncertainty for uncertainty’s
sake. Which sounds terrible, but is actually better than skill in many cases.

Randomness Tends to Make a Game Easier

The more decisions the player has to make, the harder the game is for new
players. As mentioned above, new players can easily be overwhelmed by
too many decisions, especially if they haven’t learned the coping
mechanisms to simplify those decisions. One way to improve this as a game
designer is to take some of those decisions out of the player’s hands.
Replacing a choice with a random event makes the game simpler and easier
for a wider range of players.
Imagine Catan where you could pick what resources to buy each turn
from a budget with no randomness. It would give players more control but
would also add a big decision to each turn and thus slow the game down.
And it would make games more predictable – players would tend to buy
exactly what they need and nothing else, so they’d do that every time. But
mostly it would make the game harder. You’d need to make better
purchasing decisions than the other players if you want to win. The real
Catan is easier and thus more fun for more players.
This is where understanding your game and your audience are important.
Maybe you want to attract players who like complicated decisions. Maybe
your vision for your game is based on players learning to manage decisions
like this one. But if neither of those are true, and you’re hoping to get as
many people as possible to play your game, you may want to consider
streamlining any choices except the most important ones. If it’s not your
core game loop, maybe it should be a die roll instead of more things players
need to think about.
Randomness Allows Mixed Groups to Have Fun Together
Casual players prefer randomness over skill. If you don’t have the skill,
randomness is your best chance to win. This may sound terrible, but it’s
important. Most people in the world use games for entertainment and don’t
invest enough time to master every game they play. But they still want to
play games with a variety of different friends and family and have a good
time. Playing a skill-based game with people of varying skill levels is not a
great experience for either side. The best player gets bored and the worst
players get frustrated. This is why I don’t play a lot of multiplayer first-
person shooters these days – I’m terrible at them and just die over and over
again. Dice fix this. For a mixed group of varied skill levels or mastery
levels, a little randomness provides a way for all players to engage with the
game, even if the skilled players win a little more often.

Randomness Tends to Make a Game More Widely Appealing to

a Large Audience
There are more people out there who enjoy a quick game every now and
then than there are experts who master the skills of a game. If you want
your game to have a large audience, rather than small but passionate
audience, it’s generally easier to succeed with a game where everyone feels
like they have a chance to win.
The sweet spot is when you make a game that everyone feels they can
win, but the skilled players feel they’re more likely to win. This is where
Catan shines. It feels both random and strategic at the same time. This is a
big part of why Catan was incredibly successful as a gateway game,
bringing hardcore gamers together with casual gamers around the same
table. The casual players feel like they understand what’s going on and have
a chance, and the hardcore players feel like they are more likely to win
because they understand even more. And that provides a way to bring new
players into the board game community, which is a wonderful thing.

Randomness Provides a Catch-Up Mechanic

Randomness provides a way for players who are behind to realistically
expect that they might catch up. Whether they’re behind due to some bad
choices early on, the superior skill of the opponent, or just bad luck, there’s
always hope that the next die roll or the next card draw or the next random
loot item will turn that all around.
This has a larger impact on the success of a game than you might think.
Giving players psychological armor to make them feel good and enjoy the
game at every moment is a huge benefit. Part of the reason many people
find Monopoly to be an imperfect gameplay experience is that despite the
heavy use of random probability, Monopoly does not have a good catch-up
mechanic. If you’re behind in a game of Monopoly, it feels like you’re just
going through the motions until you lose. That’s not a positive play
experience. Compare the feeling of being behind in Monopoly to the feeling
of being behind in Catan. In Catan, it always feels like you’re just a few die
rolls away from doing something big. Maybe not winning immediately, but
at least building another town. Building creates good goal structures. And
with the right combination of the robber and a few bad die rolls, the player
in the lead can be toppled. I don’t know the statistics for how many come-
from-behind wins occur in Monopoly versus Catan, but when we’re talking
about emotional feel, it doesn’t really matter. The important thing is that the
player feels better about the bad result in Catan, whether they actually come
back or not.
Downsides of Randomness
Uncertainty is part of what makes a challenge interesting. Adding
randomness randomly to a bad core loop isn’t going to improve things, but
finding the right places to add a little bit of uncertainty can help most
games. This depends on your audience. Uncertainty and especially
randomness work for certain types of games for certain types of audiences,
and not for others.
All games want some flavor of uncertainty, but skill-based uncertainty
and random-based uncertainty attract very different sorts of players,
especially for competitive games. Competitive gamers want outcomes
decided by skill and skill alone. This is true for mental skill games like
chess and games that include physical skill, either shooters like Call of Duty
(2003) or a professional sport. Removing all traces of probability from your
game is a strong statement that you want it to be decided purely by skill.

Punishes the Best Players

The highly skilled players who spend a lot of time developing and
mastering a skill tend to get annoyed when someone else wins based on
random events. Expert players tend to frown on luck. High-level Super
Smash Bros (1999) tournaments usually turn off items, which are a feature
that adds random powerful items to the map at certain times, creating a lot
of chaos that might make a less skilled player win. Without items, the game
is more skill-based and allows experts to pit skill against skill. The more
complicated a board game gets, the less chance affects the outcomes.

Takes Away Earned Rewards

This actually demonstrates what I consider to be the worst use of
randomness in Catan. When anyone rolls a seven, all players with more
than seven resource cards have to discard half of those cards. This doesn’t
feel great. The player worked hard for those cards, and they are taken away
with no recourse. There is no way for the player to react to this event and
salvage their hard-earned resources.
This is generally a psychological problem with uncertainty. When there is
something that the player can’t know, and it does something negative to the
player, that feels unfair. Especially when the penalty is immediate and
there’s no way for the player to mitigate the negative impact.
Players get especially angry when the penalty takes away something that
they consider “theirs” either by right or by earned effort. Humans tend to
use a very broad definition of “mine” and get very touchy when this sense
of ownership is violated.
In defense of this rule, it’s right there in the rulebook and everyone
knows it can happen at any time. Players can take steps to avoid this
penalty. If you have more than seven cards, or are close to it, this rule puts a
lot of pressure on you to do whatever you can to build on your turn. Which
is the point of this rule. But there are still plenty of times when you take
every possible step but end up with a good income right before a seven, or
when no good builds are possible and you have to hold only six cards and
hope. It’s a necessary and useful rule but leads to less-than-perfect
moments.
Can Overwhelm Choices
Games are about choices between the actions the player can take. Random
events are not verbs, because they’re not actions the player takes (unless the
game is set up so a player initiates the random event, which is a whole
different thing). If a game has too much randomness, it can take away from
the core of the game. If there’s so much randomness that my actions don’t
really matter, then why even play a game? Games like Munchkin and Fluxx
are not popular with some players for this very reason. Games are only fun
if the player’s actions are meaningful, in the context of the game.

Downsides of Uncertainty
The cloud of uncertainty that we talked about as a positive before can also
be a negative if not handled correctly. If you use uncertainty to prevent
long-term prediction, but you cover up too much of the game with
uncertainty, players might feel like nothing is ever knowable and strategy is
not worth it. If the game’s catch-up mechanic is too strong, it can feel like
there’s no benefit to getting into the lead. This can be another problem with
Munchkin – players who do well early are severely punished, so some
players will pace themselves and not move to win until one or two other
players have tried. Which some players consider to be an interesting
political strategy, but feels wrong to some players.

Building Randomness in Catan

Catan is a master class in how to use randomness to make a great game.
Let’s look at each place where the game introduces random elements and
discuss the principles that make the randomness work.

Initial Board Setup

Each game of Catan starts by setting up a randomly generated board. This
includes both randomly arranging the tiles and also semi-randomly
arranging the numbers on the tiles and the ports around the island.
A different board each time is a big part of the appeal of this game. Each
board offers a new puzzle for players to analyze and react to. As discussed,
uncertainty provides variety.
One reason randomizing the starting layout of the game works so well is
that it’s done before the player makes any decisions. Randomness can’t take
away earned rewards if the players haven’t started the game yet. And this
randomness is fair – everyone is on the same footing and reacts to the same
information.

Resource Production
Every time a player rolls the dice to produce resources, there’s no way to
know what’s going to happen. But over the course of a whole game, you
can say what is likely to happen. One important factor when setting up a
random event in your game is to have an idea of how often that event
occurs.
Things that only happen once across the whole game make the random
nature very obvious and hard to manage. If you get a weird result in
something that only happens once, it’s really going to stand out. If you get a
weird result once and it determines the outcome of the game, that’s a
problem. These things should only be randomized if they meet the other
criteria for good randomness – for example, the random board in Catan.
Things that happen many times over the course of a game are going to
spread that randomness around and display the full range of random
outcomes to the players. This gives players the tools to understand and
manage the randomness long-term. The resources in Catan are a good
example of this. Resources are rolled many dozens of times every game.
The number of different outcomes the players see makes it likely that
they’ll see a good range of outcomes. Catan’s resource production is very
nicely balanced where the probabilities tend to even out over the course of a
game, but not always perfectly. There is enough variation from game to
game to keep things fresh, but there is enough consistency within a game
that players who understand probabilities will have an appropriate
advantage.
 That said, the probability doesn’t always work. The most frustrating
games of Catan that I’ve played are the games where a player makes an
intelligent informed choice early in the game, but the dice decide to not
behave that day. There are feel-bad moments when a player does everything
right but the probability doesn’t align appropriately. Most players are able
to weather this psychological trauma, and probability makes that easier to
manage since there’s something else to blame. In an entirely skill-based
game, players have nothing to blame but themselves. In a game with
random elements, the dice (or cards, or RNG) becomes a convenient
scapegoat. This sort of mental crutch may sound silly but it’s a very useful
tool as a game designer. The more you can do to make the player feel good
after a loss, the better overall experience your audience will have. A game
where the winner walks out happy but the loser walks out sad is a net zero
fun. But a game where the winner walks out happy and the loser walks out
feeling fine is a net positive. Make the world a happier place – let players
blame chance.

Development Deck
The development deck is a classic randomizer. Decks are so ubiquitous as a
randomizer that people sometimes forget that they’re random. Losing a card
game because you didn’t draw the right cards feels different than losing a
card game because a card did something that randomly picks targets (yes,
I’m looking at you, Hearthstone (2014)).
The development deck is also a strong example of some of the nuances of
randomness. The deck is mostly soldiers, which have a useful but not huge
effect. The rest of the deck is either victory point cards, which make you
win, or cards with varied but powerful effects. So, players who draw cards
are expecting to draw a soldier, but sometimes get something better. If the
non-soldier cards were weaker than soldiers, that could be a very negative
moment. Paying resources to draw something but getting something worse
doesn’t feel great. With the deck as constructed, even when you don’t get
what you expect, you get something you’re happy with.

Rolling a Seven
One result of the resource roll is special. When a seven is rolled, that player
gets to move the Robber. This does a few varied things: the hex the Robber
is on doesn’t produce, the player gets to steal a resource card from someone
near there, and all players with more than seven cards in hand have to
discard half of them.
Now, I worked briefly on Catan but I never met Klaus Teuber, the
designer of the game. I don’t know any details of his process or how the
game changed during iteration. But I have strong theories about why the
Robber works the way it does.
First, there is a common complaint about Catan (any many board games)
that the game is not very interactive. In this context, this isn’t looking at the
player’s interactions with the game – there are clearly lots of meaningful
choices in a game of Catan – but instead at the players’ interactions with
each other. Trading is very interactive, but it’s always mutually beneficial.
There aren’t a lot of ways to slow down the progress of another player in a
game of Catan. I suspect that Mr. Teuber had an early version of the game
with no robber, and he quite wisely noticed that there was not enough
interaction. So, he found a way to add some in. Both the production halt and
the card theft are good ways to weaken the top player.
The third effect – players with lots of cards have to discard some – also
feels like something that would crop up during iteration. I can picture Mr.
Teuber hosting a few playtests where one player has a good run and ends up
with a hand of dozens of cards, then doesn’t do anything with them for a
long time. This doesn’t feel great for anyone, so it makes sense to add a
way to encourage players to build with cards instead of hoarding them.
Based solely on my thorough analysis of circumstantial evidence, I believe
that the Robber came about as a game design solution for two problems in
early versions of Catan. The fact that this doesn’t seem obvious when first
playing the game is a testament to the elegance of these solutions.

Certainty about Uncertainty

Catan provides a great model on how to use uncertainty and specifically
randomness. Dissecting the verbs and goals helps us see how uncertainty
works here, and why the randomness of Catan feels better to many players
than randomness in other games.

Uncertainty helps avoid analysis paralysis by preventing players from

analyzing too far in the future. But the build goals still allow players to
look ahead and think long-term.
Uncertainty keeps the game interesting for the largest number of
players by providing a balance for players with greater skill at
analyzing resources.
Uncertainty is used in places where it’s either not disrupting players or
providing players with ways to understand and analyze the randomness
and change their actions accordingly. Randomness doesn’t destroy
player agency.

Understanding how randomness is used in Catan can help many games

improve their play experience and appeal to more people.
OceanofPDF.com
Chapter 6
Spelunky / Systems
DOI: 10.1201/9781003346586-6

Spelunky
Spelunky is a great game.
Spelunky (Figure 6.1) was first released as a web game in 2008, then
attained larger success on Xbox and other platforms with a 2012
update/remake. Spelunky is part of the first wave of indie games, which
have spawned a thriving market.

Figure 6.1 Spelunky screenshot.

Spelunky – Gameplay Dissection
Verbs
What do you do in Spelunky?
You run, you jump, you attack, and you use items that act as powerups.
Yes, there are more verbs here. But as discussed, the depth of dissection
depends on your goals for the dissections. And this meets my goals for this
chapter.
Run? Jump? Sound familiar?
Spelunky is an indie 2D side-scrolling platformer. Shovel Knight is an
indie 2D side-scrolling platformer. These games have many similarities, but
also many differences. More on that in a moment.

Goals
As a platformer, the goal of Spelunky has a fairly typical set of layers:

1. Explore the lost temple or whatever it is that you’re exploring

(narrative framework).
2. Get to the end of the game (long-term gameplay goal).
3. Get to the end of the level (medium-term gameplay goal).
4. Get lots of treasure (gameplay goal that ranges from short-term to
long-term).
5. Rescue your friend (medium-term gameplay goal, with gameplay
rewards but no strong connection to the story; rescuing people is such
a popular trope in games like this that it needs no introduction).
6. Don’t die (short-term gameplay goal).

Again, these mirror the goals of Shovel Knight. Games in the same genre
tend to have a lot of similarities.
Challenges
Levels in Spelunky are challenging because of difficult jumps, varied
enemies, and lots of different hazards. The exact set of dangers and their
powers is different in the two games, but the concept is the same – every
level is a sprinkling of varied challenges that keep the player on their toes at
all times (Figure 6.2).

Figure 6.2 Simple diagram showing Actions, Challenges, and

Goals for Spelunky.

Progression
This is where things start to change a bit.
In Shovel Knight, the player plays through a sequence of levels that are
pre-defined and never change. The player might need to re-play a section of
a level after the player dies, but that section will always be the same and the
section after it, too. Every player sees the same content.
In Spelunky, the levels are procedurally generated each time the player
plays. If the player dies, they have to face an entirely new level with new
hazards arranged in new ways. There is no level progression in a traditional
sense – the player plays as many random levels as they can until they die,
and then they start over again. Until you get far enough to get to an ending,
but I don’t want to spoil it all. This leads to a number of different dynamics
that affect the overall feel of the game. More on that in the systems
discussion, below.
This style of play, with random levels that reset on death, is called
“rogue-like”. More on that in the history discussion, below.

Spelunky – Non-Gameplay Dissection

Senses
Spelunky is a sprite-based retro-styled game. Simple graphics are important,
as procedural generation puts lots of extra requirements on the content of
the game. Everything has to be flexible enough to be placed anywhere, and
still perform its intended function. Platformers are notorious for requiring
precise placement of items and platforms to get the desired difficulty – the
length of a jump has a huge impact on how it plays. So, making a wide
range of content that all fits those requirements and still looks great is no
easy task.
Spelunky content creation is extra double challenging because not only
does everything have to work in the generation system, but it also has to
clearly communicate how it works. The game is built on the player’s
understanding of how every hazard and enemy behaves. In a normal game,
that understanding would be built partially on carefully designed
introduction scenes where the player learns the dangers of a new item and
how the character is meant to interact with it. Because the player’s first
encounter with a new hazard in Spelunky is randomly generated, the object
itself needs to communicate everything about how it works based solely on
itself. An enemy that shoots projectiles might be firing short-range, might
be firing at a walking character or a falling character, and might be the first
thing the player sees or the end of a long sequence of threats.
Communicating how gameplay elements work is always hard.
Communicating like this is like doing trapeze acts without a net.
Given these constraints and requirements, it’s kind of amazing that
Spelunky works at all. It’s especially amazing that it’s cute and compelling
and has a strong style of its own.

History
Spelunky is an indie game. As discussed with Shovel Knight, indie games
are generally PC games that generally have a strong artistic view and
generally are smaller and more personal than big console or mobile games.
Talking about “indie” is usually a contrast to differentiate a game from the
big AAA studios.
Spelunky is a rogue-like. This sub-genre gets its name from an old
computer game called Rogue (1980) that used ASCII graphics to build
dungeon levels out of letters and keyboard symbols. Players would explore
a random ASCII floor of a randomly generated dungeon, fighting monsters
and acquiring treasure in a simple Dungeons and Dragon style adventure.
Despite the simple graphics, Rogue inspired a number of similar games
(Figure 6.3) and became a moderately popular sub-genre of computer role-
playing games. Rogue inspired the Mystery Dungeon (1993) series in Japan,
which drew inspiration from Rogue but added real pixel graphics instead of
the simpler ASCII approach. Diablo (1997) also explored these ideas to
great commercial success.
 Figure 6.3 NetHack, my favorite early rogue-like.

Spelunky was a conscious attempt to infuse the rogue-like structure into a

genre that is not normally created with procedural generation. Previous
explorations into the rogue-like structure had generally kept the dungeon
setting and Dungeons & Dragon styling with magic and monsters and so
on. Spelunky went a different route, with platformer gameplay based on
running and jumping more than character progression and stats-based
battle. Spelunky’s success led to a revival of rogue-likes, especially in the
indie scene. Games like Slay the Spire (2017) and Hades (2018) and FTL:
Faster than Light (2012) took that progression structure and applied it to all
sorts of different core loops, leading to a wide range of interesting games.
Some people call rogue-likes that play with the core loop “rogue-lites”. I
call them fun.

Money
How does Spelunky make money?
The primary financial structure of Spelunky is the download. Spelunky
initially launched as PC freeware, without huge commercial aspirations.
When this version proved successful, the developer reworked the game for
Xbox Arcade. While most indie games are thought of as PC games, console
download services were a key factor in popularizing many early indie titles,
including Spelunky, Castle Crashers (2008), and World of Goo (2008). This
provided a stronger revenue stream than early PC download platforms
could offer. Since then, Steam and other PC downloads have boomed, so
console downloads remain important but are less necessary.

Emotions
The emotional pang of losing all your progress and having to start over
from the start is part of the appeal of rogue-likes. In my opinion, it’s also a
limiting factor in the appeal of rogue-likes. Many players (myself included)
love that extra emotional kick in the pants, especially when it’s clear that
the game is built around that and repeated playthroughs are expected. I
suspect that not everyone agrees with that. Losing everything can feel
terrible, even if you know it’s just a temporary setback. Which is why
rogue-likes and rogue-lites are critical darlings but can struggle to find huge
commercial success. The ones that have the greatest commercial success
(Diablo, Mystery Dungeon) are the ones that take the edge off that loss by
maintaining a good chunk of the player’s assets and not doing a full reset.
Spelunky also hits another emotion common in games – feeling clever.
There’s a certain joy in realizing something that you hadn’t noticed before
and using that understanding to reach new heights. This is the driving factor
in most puzzle games, which Spelunky shares.

Meaning
What is Spelunky about? What does Spelunky say about the world? How
does Spelunky change you?
 There’s something special about defeating a level in Spelunky and
realizing that no one has ever beaten that level before. No one has seen the
exact combination of things that you just defeated. There’s power in
uniqueness. It feels good to discover and defeat something new.
Spelunky is also a great example of something I firmly believe. Learning
is fun. Not in an elementary school book fair kind of way (although I also
believe in those), but in the act of learning itself. For many people, play is
the best way to learn. Experimenting with an idea in a safe, structured space
gives learners a sense of ownership and accomplishment, especially if the
space includes positive feedback on key actions. Games have always taught
through play, and humans have been learning through play for tens of
thousands of years. Our modern education system works very hard to
remove this impulse from most students, but for the few who manage to
maintain that perspective, there is a lot to enjoy and savor in a game like
Spelunky. Your battles feel more rewarding when you know that you
discovered your solutions yourself.

Systems
So how does a game like Spelunky do it?
In the Centipede chapter, we talked about dynamics and systems. Now
let’s talk about them in more detail and look at how game designers can use
systems to make better games.
Building a level in Spelunky takes a lot of clever ideas. It’s not just
randomly generating a bunch of hazards. Randomness would result in lots
of dead ends and impossible levels and unfun layouts. Incorporating
randomness into a process like level generation takes some careful crafting
and balancing. It requires a thoughtful approach on where to include
randomness and where to ensure good game design. If you’re dissecting
how to do this for a game like Spelunky, you’re in luck. Derek Yu, the
creator, has done a great job explaining his process in the book Spelunky
(Yu, D. (2016)). It’s interesting and clever and will help anyone planning on
making a procedural generation system. I recommend you read it.
For this book, I’m interested in talking about how a game like Spelunky
works from more of a player perspective. How do players figure out a big
complex level? How is that still fun?

Understanding Fun
To start off, the player needs to understand how a single element of the
game works. This is true in every game – we already broke down some of
the rules for games like Centipede. Let’s do this again for Spelunky.
Spelunky has lots of different hazards, but one of the early ones is the dart
gun, also known as an arrow trap.
Here’s a basic set of rules for an arrow trap:

1. When the player moves in front of the arrow trap, it shoots an arrow.
2. The arrow moves in a straight line for a short distance, or until it hits
something.
3. The arrow damages the player if it hits him.
4. Each arrow trap only has one arrow, after which it doesn’t do anything.

That’s the basic idea. As a casual reader, this works fine. But whoever codes
this functionally needs to dig deeper into details to actually make it work –
What is “something”? What is “in front of”? How much damage does an
arrow do?
All these only work if the player knows what’s going on. If the player
thinks that arrows will track their position and aim at them, then that player
will not be able to devise the correct strategies to survive. The game needs
to make sure that everything in the sentences above is very easy to learn
and understand and notice. The arrow trap is intended as a “trap”, so it
doesn’t need to clearly communicate everything to the player when it’s first
encountered, but once the player has seen it happen, they should be able to
predict what will happen in future encounters. There are various ways that
the presentation of the arrow trap helps the player understand the rules up
above:

1. An arrow trap is a different color than the ground, so it stands out.

2. An arrow trap has a face on it that indicates the direction it’s facing.
3. The mouth on the face shows where the arrow comes out.
4. When an arrow fires, there are visual and audio effects to make it clear
what’s happening.
5. The arrow is an object that follows the rules of the world and persists
after it is launched.

(As a game designer, I’m actually a little surprised that there is not a clear
visual distinction between “arrow trap ready to fire” and “arrow trap has no
arrow”. Even a small texture change or slight color swap would be a nice
way to emphasize this distinction, which is critical to understanding how to
manage the arrow trap as a player. Maybe that would be too obvious, and
it’s meant to be one more thing for the player to keep in mind.)
Following on the dynamics concept that we discussed back in the
Centipede chapter, we can start to see some dynamics that are logical
consequences of these rules:

1. The player can make the arrow trap fire to make it not a threat.
2. Since the arrow follows the rules of physics, if the player is falling fast
enough when the arrow is triggered, the arrow will miss the player.
 3. Use an arrow trap to make a weapon – the arrow can be thrown like a
rock but does more damage.
4. Levels with arrow traps become less dangerous over time – players
need to be cautious when first exploring an area but can relax a bit
when returning.

How do you achieve nice clear visual connections between items and their
function for your game?

First, be careful with what you add to the game. If something doesn’t
have an obvious function, it might not be worth including. Simple
direct categories of things are best. If it’s a game about inflicting
violence, things that do damage make sense. If it’s a game about
planting and tending a garden, brass knuckles might not be a good
thing to include. Adding weird machines to your game that do
interesting things but have no visual connection to their function can
be fun, but it’s dangerous.
Second, clear and consistent visual design can help make things make
sense. Think through what the player needs to know and simple ways
to get that across. Adding a face to the arrow trap is a great example –
faces are something that everyone naturally perceives as oriented in a
particular direction, and decorative statues make sense in the
environment of Spelunky.
Next, look at ways to enhance visual communication beyond just the
look of the object itself. When something happens, add sound and
visual effects to emphasize that action and call the player’s attention
there. Make the VFX persist for a little bit so the visual sticks around if
the player doesn’t notice it immediately. Have things flash different
 colors or outlines when they take damage. Decide what is important
for the player to know, and spend your time and effort there.
Finally, come up with clear and consistent visual communication that
all objects share. When the arrow hits the player, there isn’t a unique
visual and sound to indicate that the player takes damage. All damage
is communicated the same way. When other things later in the game
shoot out other projectiles, use the same or similar sound and visual
effects so the player associates them.
Building to create consistent player expectations is true for visuals and
also true for rules – other objects that shoot should follow the same
shooting rules as arrow traps. They can break these established
expectations, but only carefully. Ideally don’t break more than one rule
at a time. So future things should shoot a single projectile in a straight
line triggered by motion in front of the creation object. If you want to
shoot lots of things that move erratically and trigger when anything
gets near the creation object in any direction, that may be harder for
the player to understand and require more careful tutorialization or
difficulty balancing to make it work.

So, arrow traps make sense. Arrow traps follow a consistent set of rules that
define their behavior, and lead to interesting dynamics. But that’s just the
start.

Procedural Fun
Procedural-level generation isn’t just a fun party trick. When the player
doesn’t know what’s next, and literally can’t know what’s next, they have to
approach things differently. When the player dies in Shovel Knight, the
player (hopefully) knows what killed them and what they did wrong. When
they respawn at the save point, they are ready to try again. But in Spelunky,
each challenge is unique and disappears when the player dies. The player
can’t just memorize what to do and then execute on that memory. The
player has to analyze every situation as a new set of problems.
This puts a lot more pressure on the player to really understand what’s
going on. The player needs to be able to read the room and piece together
how all the parts fit together, and then come up with a plan of attack that fits
those problems. In Shovel Knight, the player can rely more on standard
patterns of play – jump over everything – and correct their course when
something new disrupts that pattern – a spike above the place you want to
jump, for example. In Spelunky, the player only gets one shot and if they
don’t read it right the first time, it’s game over.
To make this work, the player needs to be able to look at any
combination of objects in the Spelunky universe and be able to predict
what’s going to happen with that combination in those precise positions.
While this is possible in most games, Shovel Knight included, it’s critical in
a procedurally generated game like Spelunky. If there are combinations that
are unpredictable or impossible, the game falls apart.
This means each element in the Spelunky universe needs to behave in
consistent, predictable ways. And to complicate things even more, each
thing needs to interact with each other thing in consistent, predictable ways.
This is true in all games – building the grid of “what happens when this
touches that” is a big part of content design in many games. And many of
those grid entries are going to be fun and interesting and lead to new types
of gameplay and inspiration for levels. That’s where most level ideas in
Where’s My Water? came from, for example. But some combinations just
don’t work right. Or aren’t very fun. In a hand-crafted game like Shovel
Knight, if there’s a certain object interaction the game doesn’t handle well,
the level designer can just make sure those two elements aren’t near each
other. Problem solved! In a procedural game like Spelunky, everything
needs to work with every other thing.
There’s a great essay by Liz England called “The Door Problem”. In it,
she discusses all the things that a game designer needs to think about if their
game has doors. Seems simple, but there’s actually a surprising amount of
nuance in even something seemingly obvious like doors. When you’re
playing a game, you focus on the interesting and complicated things. When
you’re making a game, everything is complicated. Simple things like doors
are simple by themselves. Writing the code to make a door work is fairly
straightforward and can be done quickly. But any element, doors included,
may potentially interact with any other element. And all of those
interactions only exist if the code knows what to do in that case. And the
more interactivity and choice the game gives to the player, the more
possible interactions that the development team needs to code. Freedom has
a price.

Systems Thinking
What do we call a bunch of things that follow their own consistent rules,
and also have consistent rules for how they interact with each other? We
call that a system.
The study of systems as a whole – the shared understanding of how all
systems work – is a new field that is just starting to make some big
advancements in our understanding. If you’re interested in this, I suggest
starting with Thinking in Systems: A Primer by Donella Meadows as a good
book to start with.
Once we have Spelunky’s arrow trap start interacting with everything
around it, the game becomes a system. Individual rules for each object
interact with individual rules for each other object to create new rules
unique to the interactions. And since we know that rules create dynamics,
we also know that rules for object relationships are going to create
dynamics unique to those interactions.

Does the arrow trap trigger only when the player moves in front of it,
or can other things trigger it?
If inanimate objects trigger it, then throw things to make the
arrow trap fire without getting hurt. This increases the value of
inanimate objects. In early levels where arrow traps are prevalent,
a rock is a very useful thing.
If enemies trigger it, then knock an enemy into an arrow trap to
kill two birds with one stone.
What happens when an arrow hits a solid object?
If the arrow destroys itself, things are clean and simple but the
player can’t use it as a weapon later on.
If the arrow sticks in the wall and can be picked up after, it’s a
nice safe way to make it available to players later, but it might be
a challenge to reach it.
If the arrow bounces off the wall, it’s available for use later but
may still be a threat while it’s bouncing around. This is the actual
answer in the published game.
What can the arrow damage?
If the arrow only damages the player, that’s simple and clear but
removes the possibility of some fun dynamics.
If the arrow damages anything alive, that makes sense and allows
players to use arrows to remove other threats like enemies.
If the arrow damages every type of object, then level design needs
to consider that since placing a pot in front of an arrow trap
basically removes the trap as a threat.
 And if an arrow damages the terrain, that’s interesting and opens
up some interesting dynamics where the player can use an arrow
trap to access new areas. But it doesn’t really feel right for an
arrow. And also might be too powerful if the player can pick up
an arrow and destroy any terrain anywhere. Maybe if it was a
cannon, this would make more sense. That might be an interesting
variant of the arrow trap to add in later levels.

Making the game rely more on system interactions opens up new

possibilities. In a game like Shovel Knight, the game designers craft
interesting and fun interactions that the player discovers and solves. In a
game like Spelunky, the systems of the game allow for interesting and fun
interaction that the player has to create on their own. This may sound like
the game designer is less involved, but actually the opposite is true. The
interactions between objects in a game like Spelunky are only interesting if
they are carefully thought out to ensure that the player can find fun in
almost any interaction. And the level design may be procedural, but there is
a lot of thought put into how things are placed to ensure that they are fun
and not too difficult or frustrating. Derek Yu’s book talks about all the ways
he ensures that players don’t get stuck in dead ends or have exit doors in
impossible-to-reach locations. Good game design doesn’t happen by
accident.

Dissecting Systems
To understand a game, you have to understand all the elements of the game.
But the tricky part is understanding how all those elements fit together into
systems.
Game development teams often talk about any collection of features and
rules as a system. There’s a combat system and a magic system and an AI
system. This is a loose and casual use of the term, but it works. When you
have a lot of elements that closely relate to each other, their rules are more
likely to butt up against each other and lead to systematic interactions.
So, what are the systems in a game like Spelunky? We can do a big
brainstorm like we did with verbs and other things before. The idea here is
to think of any collection of smaller features that might affect each other.
Anything that someone would have to code or build to make the game what
it is today.

Movement: The player’s avatar can move around, including running.

Jump: The player’s avatar can jump and control their motion in the air.
Depending on the purpose of my dissection, I might just group this in
with movement.
Terrain: There are platforms and other types of terrain that characters
can stand on. Terrain can be destroyed.
Items: The player can acquire items that basically act like powerups,
giving the player different interactions with the game. This includes
things like Bombs and Climbing Gloves.
Traps: There are objects placed in the game world that actively try to
harm the player. This includes things like Arrow Traps.
Objects: There are objects placed in the game world that the player can
pick up or otherwise interact with. This includes things like Rocks and
Stone Blocks.
Monsters: Enemies that try to harm the player.
Bosses: Enemies with more complex behavior patterns and (often)
specific levels built for them.
Combat: The player can interact with enemies to defeat them or take
damage.
Damage: The player and monsters have defined health points that can
be removed. When they reach zero, that thing dies.
Death: What happens when the player’s avatar dies?
Time: The game tracks how long the player spends in each level.
Specter of Death: If the player spends too long in a level, a ghost
appears to encourage the player to move on.
Money: When the player collects items, they are translated into a
number. The player can spend this number at shops.
Damsels: The player can find characters in the levels. Rescuing them
earns hearts.
Idol: The player can find special treasures in the levels. Picking this up
triggers a boulder trap.
Shops: The player can find shops where they can buy things. Attacking
the shopkeeper has consequences.
Altar: The player can find altars where things can be sacrificed for
rewards.
Level Generation: There is a system of rules that determine how levels
are created.
Tutorial: There is a tutorial that the player goes through the first time
they play that introduces the basics of the game loop.
Progression: The player advances through levels, heading toward the
end.
Game Modes – Daily Challenge, Deathmatch: There are different ways
to play that this book isn’t going to cover in detail.
Customization: The player can select the appearance of key objects in
the game.
Cinematics: At key moments, the game presents non-interactive
sequences of information. This includes things like the opening and the
 ending.
User Interface (UI): The game lets players select things in menus to
start games and select other things.
Options Menu: There is an options menu where the player can set
various things.
Leaderboard: There is a leaderboard to track progress relative to
friends.
Achievements: There are achievements the player can unlock.

As with verbs, we can group these into meaningful categories (Figure 6.4).

Figure 6.4 Diagram showing the features from the list as boxes.

We can start with the core loop. What can the player do? What has to
exist for the player to be able to do anything?

Movement
Jump
Terrain
Items
Then we can add obstacles and challenges. What makes the player’s
progress uncertain and difficult?

Traps
Objects
Monsters
Bosses

What are the rules for how the player interacts with these challenges? What
are the consequences for failure?

Combat
Damage
Lives/Death
Time/Specter of Death

Then there are goals – things the player wants, and ways to turn one
resource into another.

Money
Damsels
Idol
Shops
Altar

Once you have verbs, challenges, and goals, you have a full core loop.
The idol is interesting because it bakes risk into its reward. And the same
can be said for many of the threats mentioned above – sometimes the risk
and the reward are separate features that need good level design to fit
together, but sometimes the risk and reward are baked into the feature itself.
These features can defy simple categorization into binary good/bad buckets,
but for a quick review I don’t mind simplifying a bit. Just remember that
these categories are not absolute.
Then there’s the stuff you need to make the core loop work over multiple
cycles.

Level Generation
Tutorial
Progression
Game Modes – Daily Challenge, Deathmatch

Finally, we have the stuff that happens outside the gameplay.

Customization
Cinematics
UI
Options Menu
Leaderboard
Achievements

This last list includes all the connective tissue to guide the player through
and tie everything together (Figure 6.5). When we talk about the features or
systems of the game, sometimes these are left out because they’re not really
the “game” in a technical sense. But they’re very important to consider
when you’re thinking about building a game, as they take time and money
and need thought. But when we’re just trying to understand the gameplay,
we can set these aside.
 Figure 6.5 Diagram showing features grouped into categories.

Similarly, I debated including “Game Engine” and “Graphic Engine” and

“Sound Engine” and “Tools” in the above lists. These are a layer removed
from even the UI setup when considering the game itself but are important
to the development process. When I started in the industry, these would be
the first things to be considered and would require months of work.
Nowadays, many games use off-the-shelf engines to completely bypass
these issues. Most of the time, when I talk about the “systems of the game”
I wouldn’t include these, but if we’re looking at the list as a to-do list for
game development, these are important things to consider.

System Link
Identifying the systems is useful. Once we have a list, we can start to
consider how these systems fit together.

1. Jump + Combat (Figure 6.6)

 Figure 6.6 Diagram showing the previous feature list,
focused on these two.

a. To defeat an enemy, you need to position yourself in a place

where you can damage them.
b. When you land on an enemy from above, you damage the enemy.
2. Combat + Items (Figure 6.7)

Figure 6.7 Diagram showing the previous feature list,

focused on these two.

a. Items can be thrown to damage enemies.

 b. Items can contain enemies that are released when the item is
destroyed.

This can lead to nearly infinite possible combinations, so drawing out a

complete diagram of every possible connection is an impossible task. When
thinking through connections, it’s often best to approach a diagram like this
with a particular lens. For example, focusing your attention on Level
Generation (Figure 6.8) allows you to consider how each other system is
used to generate new levels. Some features determine the base terrain of
each area, and some features determine the smaller items that are placed
inside each area. This focus can help when thinking through how the level
generation system works.

Figure 6.8 The system list from before, with items relevant to
Level Generation called out.

Removing irrelevant features can help call attention to the relationships

relevant to the current lens. Rearrange things and add new groupings or
connections to make these relationships clear (Figure 6.9).
 Figure 6.9 Level Generation systems grouped in the ways
relevant to Level Generation.

The most common system diagram is based on the economic model.

When thinking through how to control the player’s resources, it’s important
to understand where those resources appear and disappear in the game. So,
mapping out the possible economic impact of each feature is an important
way to understand how the game’s economy works. Spelunky’s economy is
fairly simple, with Gold as the only real currency the player interacts with.
There are a few other resources that can tie to Gold, such as when the
player trades Health or Time (or the risk of losing Health or Time) for Gold
by attempting a risky maneuver. But for the most part, analyzing where
Gold is gained or lost is the key to the economy and thus a good place to
start an economic analysis (Figure 6.10).
 Figure 6.10 Diagram showing systems that add or remove Gold.

When designing a game, thinking through system connections can be a

useful exercise – compare each system with each other system to see how
they relate. This will often expose things that can easily be overlooked in a
game design. This can also spark new ideas – what if this system related
back to this other system? Small simple games often don’t need a lot of
systems or a lot of connections, but a robust dynamic game wants to give
players ways to mix everything together. Systems that don’t relate to other
systems are often not really doing much in the game and sometimes are best
removed.

Systems Thinking
Once you understand the systems of the game and how they start to fit
together, you can start applying more advanced understanding of how
systems work to better understand your game. There are rules that guide
how systems work, and a strong game designer can use these rules to their
advantage.

Systems Thinking – Feedback Loops

Cybernetic systems aren’t robots or AI-based systems. It’s a term for any
system where one element can regulate another. The classic example is the
thermostat in a house. It senses the temperature in the house and adjusts the
temperature based on what it detects.
If the temperature is too high, the thermostat will turn on the air
conditioning to cool things down (Figure 6.11).
 Figure 6.11 Too Hot: Diagram showing temperature pushed
down.

If the temperature is too low, the thermostat will turn on the heat to bring
things back to the desired range (Figure 6.12).

Figure 6.12 Too Cold: Diagram showing temperature pushed up.

This is called a negative feedback loop – if things go in one direction, the

feedback pushes back the other way. You could picture a thermostat that did
the opposite – if the temperature is high, it turns on the heater to make
things even hotter. That would be a positive feedback loop. This would not
be desirable in a thermostat but has uses in other places.
How can you use feedback loops in your games?
The classic example of a negative feedback loop in games is the
interactions between the powerup system and the race rank system in Mario
Kart (and other similar racing games). Powerups are semi-randomly
generated, but the range of options varies based on the player’s current
position in the race. Players who are doing well get only simple small items,
while players in the back of the pack get much more powerful items. This
creates a dynamic where players who are in the lead tend to fall back a bit
and players who are in the back have a way to get back into the race. This is
a desirable dynamic for the game designers, as a close race where you see
your opponents near you is more exciting than a race where one player gets
a huge lead at the start and never loses it.
There are lots of similar ways that system interactions can create negative
feedback loops that make the game better:

Give players who are behind ways to catch up. There’s a Magic: The
Gathering spinoff called Duelmasters (2002) or Kaijudo (2012) (I
worked on a PS2 version) in which players don’t have life points, they
have shields. Shields are represented by face-down cards that the
player puts in front of them at the start of the game. Whenever the
opponent has a successful attack, the player loses one of their shields,
and if the player has no shield when attacked they lose the game. But
the genius game design is that when a player loses a shield, they pick
up the card and put it in their hand. The player who is behind in victory
gets a big resource advantage, which might enable them to come back
and win, or at least make the game closer and more exciting.
 Any multiplayer game with a reasonable amount of information
sharing will automatically have negative feedback loops. In Catan,
when a player gets close to winning, other players notice and take
action. Robbers are placed on that player’s tiles. Trades with that
player become much more punishing. In the start of the game,
everyone is equal so everyone is happy to perform mutually beneficial
actions. But if the mutually beneficial action gives you the victory but
only gets me to the halfway mark, then it’s not in my best interest to do
so. Negative feedback is built into any system with a slight amount of
politics.

And don’t forget positive feedback loops. Positive feedback loops can be
dangerous, because they feed on themselves and keep getting bigger. But
sometimes you want that.

Most progression systems have a bit of positive feedback built in.

Players who collect enough resources to advance in stats or money or
research or whatever are rewarded with powerful items or abilities that
make it easier for them to collect more resources. Left unchecked, this
can make a game way too easy once the player gets started. Which is
why most games check progression by increasing the power and
danger of the threats to match. In some cases, this is done naturally
through level design, but open world games can struggle with this.
Early Bethesda games got criticized for artificially increasing the
power of enemies to match the player, which made players feel like
advancing wasn’t worth it.
When you want a board game to end quickly, give the player who’s in
the lead an advantage. Many board games have powerful abilities
 players can acquire in the late game, which are meant to bring the
game to a close.
When you want a board game to feel out of control, a positive
feedback loop is just the thing. Pandemic does a great job of using
multiple positive feedback loops to ensure that players always feel like
the world’s about to end. And often, that’s exactly what happens.

Deep Systems
One key lesson in Thinking in Systems: A Primer is the idea that changing a
system is hard. And the mechanisms to make a change are often not the
ones that people find intuitive. The book ranks the effectiveness of different
leverage points on their ability to bring about real change. “Numbers” is the
worst. Just changing around values in an existing system is unlikely to fix
your problem. This is often the first place game designers (including
myself) go when trying to fix things – can we rebalance the game by
adjusting the hit points and damage of a few enemies? If it can work, it’s
quick and easy. But if the problem is more fundamental than just a local
balance problem, these types of changes aren’t going to be enough.
Delays are more important than values. If the impact of an action in one
element isn’t felt until multiple steps later, that can drastically change how
that action interacts with other things. You can see this in any game
economy – long-term gains don’t matter if the player dies on the next level,
and players tend to understand that. Similarly, delayed benefits in a combat
system are much harder to make valuable. Every Pokémon game ever has
had various status effect that are supposed to be useful and meaningful. But
the delayed benefit is hard to balance against the immediate benefit of
doing more damage. Especially in a combat system where most fights only
take a few turns. Spending a whole turn for a delayed benefit is rarely better
than just attacking again and ending the combat earlier.
Feedback loops are fairly impactful – they rank around the middle of the
list.
To really enact change, the change has to be fundamental to the system.
Change the rules of the system. Change the goals of the system. Change
how the system organizes itself. When you’re looking at a problem in your
game that involves the whole core loop or multiple elements in your
progression system or other systematic issues, try to look deeper than some
number changes. Look at how the system elements fit together. What are
the system rules that guide that interaction? What are the dynamics that
appear as a result? What dynamics would be better for the game design?
How do I change the rules so those dynamics happen? That’s where game
design gets challenging and interesting and (in my opinion) fun.

Systems Are Fun

Understanding systems is understanding games. Games are systems and the
systems are where many of the most important qualities of the game live.
As a game designer, you’re not trying to achieve certain rules, you’re trying
to achieve a feel or a strategy or make the player have an emotional
response. These are all dynamic responses to sets of features, not one thing
that the designer can mandate. Understanding how to create fun dynamics
requires understanding systems more than anything.

OceanofPDF.com
Chapter 7
Magic: The Gathering / Content
DOI: 10.1201/9781003346586-7

Magic: The Gathering

Magic: The Gathering (also known as just Magic) is a great game.
In Magic, players build decks of spells and duel as wizards. Decks are
made from cards, and each card has its own self-contained set of rules that
are printed on the card. Some cards are fairly simple, but others can
radically change the rules or even the goal of the game. These cards are
acquired in randomized packs, with new sets appearing on a regular basis,
which add new cards and thus new rules.
Magic (Figure 7.1) is one of the most successful games of all time. It has
sold billions of cards over the 30+ years it has been on sale. It has spawned
the genre of collectible card games that continues on today, and inspired
hugely successful digital games such as Hearthstone and Marvel Snap.
 Figure 7.1 Magic: The Gathering photo – my Griffin commander
deck.

I have played Magic since it first came out, but I have never worked on
it. The closest I’ve come is a PS2 version of Duelmasters, which was a
slightly simplified version of Magic from the Magic developers. That card
game took off in Japan but never managed to find the right niche in
America, which is a shame since it was a fun game. Many of the ideas from
Duelmasters are starting to appear in newer collectible card games like One
Piece Card Game (2022) or Star Wars Unlimited (2024).

Magic: The Gathering – Gameplay

Dissection
Verbs
What do you do in this game?
Magic is a little tricky here. Since every card can have unique rules and
abilities, it’s hard to really say every possible action the player can take. So
in a case like this, it’s necessary to abstract a bit in order to have a clear
discussion.
Abstracting a bit, the things a player can do in Magic are:

Play Cards
Play lands (one per turn)
Play other cards (constrained by available mana, which generally
comes from lands)
Attack
Declare Attackers (and declare targets if there are planeswalkers
or battles in play)
Declare Blockers
Activate Abilities
Cards in play can do all sorts of crazy things, often requiring
“tapping” so that card can only be used once per turn

There are other types of verbs that are the wrong type of action. “Tap” is
just the mechanical representation of how activating abilities and attacking
work. There are other types of verbs that state player goals rather than direct
actions – “Gain Power”, “Destroy Enemy Things”, “Win”. These are
expressions of goals and we’ll cover those in a minute.
So “Play Cards”, “Combat”, and “Activate Abilities” are the complete set
of (abstracted) actions that the player can take in Magic. But that doesn’t
really tell us much about how the game works. As before, Verbs only really
make sense in the context of Goals.

Goals
Why is the player performing these actions?
The goal of Magic is to win the game. There are a variety of ways to do
that, and individual cards can create entirely new victory conditions.
Similar to actions, it’s hard to pin Magic down to just one interpretation.
And you can go to higher levels of motivation that span across multiple
games. Why does the player play Magic? What are they trying to get out of
this as a hobby? If you’re interested in this level, go read Mark Rosewater’s
essay “Timmy, Johnny, Spike” for an insightful discussion of player
motivation.
When you look at the things the player wants in Magic, there are two
main buckets, which align nicely with the actions they can take. And with
things found in other games.

Build: Gather resources, increase your ability to do whatever you want

Gain sources of mana (Land)
Enhance your board presence (Creatures, Artifacts,
Enchantments)
Make your things more powerful (Auras, Equipment)
Attack: Various ways to reduce the opponent’s life total
Initiating combat with creatures
Casting direct damage spells

To be clear, this is not complete. There are plenty of things you can do with
cards in Magic that don’t fit cleanly in these buckets. Cards have various
abilities, and players might use an ability as part of a multistep process that
leads to a positive result. Combo decks have clear Build actions but often
use other paths to victory that skip Attack entirely. If you really want a
complete list, you can rephrase “Attack” as “Win” and group non-combat
victory conditions in there.

Build: Gather resources

Win: Progress toward a victory condition
Doing damage
Milling (removing cards from the opponent’s deck – if there are
no cards left, they lose)
Playing cards that change the win condition

That’s more precise, but we’re still missing a large category of Magic
actions. Spells that kill creatures or counter spells are a key part of the game
and aren’t technically part of Build or Win actions. You could argue that
these are part of Build since it’s a zero-sum game, any action that reduces
the enemy’s power increases my power. Players can build winning decks
that have very few positive build actions and lots of negative destroy
actions. And this is a valid way to look at it. But there are definitely
differences between ramp actions and removal actions that are worth
considering.

Build: Gather resources

Destroy: Remove the opponent’s resources
Kill creatures
Counter spells
Weaken creatures
Make it harder for the opponent to do things
Win: Progress toward a victory condition, such as doing damage
Similarly, if we want to dig a little deeper, we can break Build into two
types of Build. Some building focuses on directly building ways to win –
adding power to creatures, creating new ways to attack without being
blocked, etc. Some building is focused on economic goals – increasing your
ability to make more stuff. Building a strong mana base can be understood
as a related but separate action from using that mana base to add creatures
to the battlefield.

Build Economy: Gain the ability to make more things.

Play lands.
Play artifacts and creatures that produce mana.
Build Power: Gain the ability to win.
Play creatures.
Play auras and other power boosts.
Destroy: Remove the opponent’s resources.
Win: Progress toward a victory condition, such as doing damage.

This seems like a good model for Magic’s actions in the context of their
goals. We could keep going and finding more exceptions and breaking out
sub-categories for everything. Once you start abstracting up from direct
player actions like button presses, there’s some level of subjectivity and
interpretation that is inherent to the process. The “right” level of abstraction
depends on a lot of things. Why are you dissecting? Are you building a
similar game and want to compare how your competitors work? Are you
trying to understand player psychology? The level of abstraction and the
types of groups you create depend on your needs. Feel free to experiment
with different approaches and different lenses until you find the perspective
you need.
Challenges
Magic creates challenge and uncertainty through other players. Magic is
(almost) always played with other players who have the same potential
abilities and power as every other player. They might have a different deck,
but nothing (other than money) is preventing them from having the same
deck. Having a different deck is one of the choices that make up the game –
“deckbuild” is an important and interesting verb. But during play, every
player follows the same rules and has the same victory conditions available
to them. It’s like a video game deathmatch in this way.
Magic also creates challenges due to the economic systems of the game
(Figure 7.2). The mana system is a very clever way to prevent players from
playing all their best cards on the first turn. These rules create challenges
players have to plan around and optimize in order to win. Magic resembles
Settlers of Catan in this way – the biggest challenges are often the
limitations created by the rules, not the creatures or players.

Figure 7.2 Magic Flow: Simple diagram of Actions, Challenges,

and Goals for Magic: The Gathering.

Magic also resembles Settlers of Catan in that randomness adds a lot of

uncertainty. Magic generally frowns upon putting purely random effects on
cards. But every turn starts with drawing a card from a randomized deck.
Even when playing the same deck over and over, the player doesn’t know
what’s going to happen next. Long-term strategies are possible, especially
with the control provided by deck construction, but there is enough
uncertainty that predicting the state of the game even a turn or two in the
future is futile. This isn’t Chess where every consequence of every action
can be thought through – randomness keeps players focused on the present.
Even beyond randomness, individual actions in Magic don’t have any
type of challenge or uncertainty to them. The rules are very clear and
consistent and tell you what is going to happen for each player action,
including all the crazy things that appear on specific cards. Most
uncertainty is at the goal and strategy level – will this have the desired
longer-term effect? How will the enemy respond? Any uncertainty on my
action depends on the enemy player responding to it with their own action –
instant speed cards like counterspells and protection spells mean my action
might not happen, but that’s enemy uncertainty, not uncertainty on the
action itself. The big exception to this is attacking – when I attack, the
opponent has the ability to block and I don’t know how they’re going to do
that. So, attacking is a key moment with more uncertainty and challenge
than other actions in Magic.
Magic focuses its uncertainty in two places: (1) other players and (2) the
future. The player can safely plan out a turn worth of actions, including an
attack, and the only unknowns are the other player. That doesn’t work over
many turns since drawing cards from the deck has a huge effect on those
plans. But immediate plans are a duel between two players with access to
the same resources and tools. This creates a highly competitive environment
where players feel that their skill is what determines success or failure.
Magic succeeds as a highly competitive space because the uncertainty is put
in the right place to create that environment. Other highly competitive
games like first-person shooters or sports might have different types of
uncertainty such as player mouse manipulation skill or physical strength,
but the principles are the same. Focus the competition on short-term player-
on-player uncertainty. Use long-term uncertainty to avoid over-planning
and keep things different each game.

Examples
So with Magic, we can build example moments to demonstrate some of the
categories we identified.

I’m playing a creature.

Verb: Summon creature
Goal: Build – Board Presence
Challenge: No direct challenge, but limited by the mana system
and can be countered or destroyed
I’m attacking with all my creatures.
Verb: Attack
Goal: Victory Condition (damage)
Challenge: Opponent can block or otherwise prevent my attack
from damaging them
I chump block your attacking creature, sacrificing one of my small
creatures to prevent the damage
Verb: Block
Goal: Build – in that you’re trying to retain your resources (life)
Challenge: Opponent and anything they might do
I block your attacking creature with a roughly equal creature, and they
both will die
Verb: Block
 Goal: Destroy – in that this removes their creature. But there also
might be some Build going on at the same time, by maintaining
your life total. Goals can overlap like that.
Challenge: Opponent and anything they might do
I’m casting a spell to destroy your creature.
Verb: Cast spell (that destroys a creature)
Goal: Reduce opponent’s power
Challenge: No direct challenge, but you might counter or
otherwise respond
I’m casting a spell that doesn’t directly add to my power or reduce
yours but sets up for something in the future (let’s say it’s an
enchantment that changes some of the rules of the game)
Verb: Cast spell
Goal: Build – but indirectly in the future, possibly overlapping
with directly moving toward a combo-based Victory Condition.
Challenge: No direct challenge

States
Similar to Centipede, we can identify some states to a game of Magic.
Players start with no resources, so every game has to start by gaining
some resources such as playing lands. During this building phase, players
are necessarily focused on Build actions. The length of this phase can vary
greatly. Some decks try to make this as quick as possible to overwhelm an
opponent before they’re ready, while other decks focus on long-term
Building and try to surpass their opponent in resources.
Once a player has creatures out, there are two options. The game can go
into an attack state where one player is trying to damage the other. Or it can
go into a stall state where both players have creatures out and neither has a
strong attack. In this case, both players revert to Build actions until one gets
enough of an edge to attack.
As this is Magic, there are plenty of exceptions to these general themes:

A player may have a way to damage the opponent without attacking,

so they can ping the opponent for damage during a stall state.
A player may have an evasive creature that can attack during stall,
such as a flying creature, so they can ping the opponent for damage
during a stall state.
A player may be trying to win through other means, so they prefer a
stall while they gather the resources needed for their combo.
A player with a combat trick may want the opponent to think that it’s a
good time to attack, then surprise them with an unexpected shift in
power during the combat.

If you want to understand how individual games play out, understanding the
states of the game and how they play out is key. These are important things
to understand, but I’m not going to dig super-deep into this, because there
are lots of people who can tell you a lot about these dynamics. Magic is a
game that has been highly analyzed, and most of that is at this level where
individual games are analyzed to help people become better players (I like
the podcast Limited Resources and videos from The Professor, but there are
plenty of other good examples). This is a great example of a type of game
dissection that is going strong already and can be found for most games
with a robust tournament play community.
Systems
Mana
The mana system is a great example of a way to make Build actions
interesting. The player has limited resources to use each turn. New
resources are available, but they’re heavily dependent on randomness and
are limited to one per turn. The game generally follows a standard pace of
one new mana per turn, but this is not guaranteed and stalls out at some
point. The player can influence these factors via deckbuilding but doesn’t
have complete control.
Not to say that the mana system is perfect. In many games of Magic, one
side loses quickly because they couldn’t get enough resources or got too
many. This leads to games where one side doesn’t feel like they even got a
fair chance to participate. This is not a feel-good moment, and many later
collectible card games change up the mana system to make it more
predictable.
The mana system touches every other system. There’s not much you can
do in a game of Magic without mana. This fundamental constraint creates a
pace for the game. Small things happen first, and big things might not
happen since the game could end before then.

Combat
Combat is a complex system with lots of moving parts of its own.
One reason Magic is fun is that each combat is like a new little puzzle
that both players see from a different angle. Each player is trying to figure
out how to optimize the damage dealt/taken and minimize the resources lost
(usually creatures, but sometimes also mana and other things). Most
combats have the right number of moving parts and options to be fun for
most people. When the number of creatures gets large, it becomes a
different type of fun – dealing with big numbers that get out of control.
Combat affects most other systems in Magic. Players can build a deck
that doesn’t use combat, but it’s pretty rare. And even that deck needs a plan
to deal with the enemy’s combat phases. Most of the Build actions players
take in a normal game of Magic are there to make combat better for the
player. Combat is the focus of the game, and the way that most games of
Magic determine a victor.
Spells
One system that has changed over the 30 years of Magic is the spell system
– how spells are cast and enter play. Old people like me remember the rules
changes of Sixth Edition (1999), which introduced the Stack. Before the
stack, part of the fun of playing Magic was arguing about the timing of
conflicting cards – many pre-Magic games like Wiz War or Cosmic
Encounter thrived on timing arguments. But that’s not fun for everyone, just
argumentative folk like me. Most people want clear rules that prevent
arguments. When each card has its own rules, the core rules need to be
extra-solid to handle all the conflicts. Pre-Sixth, the rules were manageable
but led to a good number of counter-intuitive exceptions, which led to
arguments. The Sixth Edition rules cleared most of this up, making timing
rules simpler and easier to manage, which is good for the game but bad for
the friends of mine who like to argue.

Deck Building
Deck building is a fascinating system that is part of how collectible card
games became hugely successful.
Deck building has a big effect on each game. Since players can make all
sorts of crazy decks, each session of Magic is very different from each other
one. A fast Red Deck Wins deck against a slow Green Stompy deck is going
to play very different from a controlling Blue Permission deck against a
White Weenie swarm. This provides a great amount of variety, which is a
goal for (almost) every game.
Deck building is also very important in that players can participate in the
game when they’re outside of the game. The hardest part of any board game
is finding other players and getting them all in the same room. With Magic,
even if you have no one else in the room you can still sort through cards
and think about what they do and build interesting new decks. The magic
circle of the game doesn’t end when the other player leaves – I can be part
of the game at any time. This keeps players thinking about and excited
about the game even when they’re not playing. This can also be seen in
digital games – maximizing a Diablo or World of Warcraft build can take
hours of planning, and in the modern era players can spend hours poring
over strategy guides and discussion forums to participate in the game even
when outside of the game. As a game designer, try to spot these
opportunities and encourage ways for players to be engaged even when
they’re not playing.
More Systems
Magic has lots of systems and a full systems analysis is worthy of its own
book. Alas, this is not that book. I have to settle for just scratching the
surface and showing some of the biggest examples, but hopefully this
provides a starting point for further explorations of this game and other
games.

Magic: The Gathering – Non-Gameplay

Dissection
History
Magic came out in 1993 and immediately rocked the board game world.
Stores literally couldn’t keep it in stock for years. Within a few months,
there were countless competing collectible card games. Most of these were
terrible games that didn’t understand the quite clever ways that Magic
worked, including the one I worked on when I was at Mayfair Games. A
few were great and had large followings, at least for a few years.
Somehow, miraculously, Magic has survived as a continually published
game with new content out regularly for the last 30 years. This is an
amazing accomplishment. Think about how many games have a continual
audience for that long. There are a few “classic” board games that can say
that – chess and Scrabble and arguably Monopoly. But these are all static
games with no new content (at least in a meaningful game design sense).
And there are a few video games that get close to that, but only if you count
some fairly radically different sequels as the same thing. With Magic, the
core rules today are very similar to the ones that came out in 1993. Every
card that came out then can still be played today.
Part of the reason that Magic has survived is that the team working on it
has had all that time to think about what works and what doesn’t in the
game. Mark Rosewater has been a lead designer on Magic for most of those
30 years, and he is very public about the ways that they design and how
they think about the game. His articles and blog and podcasts and other
discussions are a great insight into how to do a 30-year-long deep dive into
one game.
 Focus
I’m jealous of Mark Rosewater. Mark Rosewater is the head designer of Magic and has been for
many years. On one hand, that can be very restricting – sometimes you want to jump onto
something new. But on the other hand, that means he’s been able to think about how that game
works and how to make it better and keep deepening his understanding of the game for decades.
Most games don’t have that luxury. A game is pitched, it takes a year or two to make, and if it’s
lucky it has a few years of live support. If it’s a huge success, it might spawn a genre of similar
games. But with Magic, it’s been the same core rules for over 30 years, and Mark Rosewater has
been able to learn things about that rule set that wouldn’t normally be possible. He writes a
column on their website, so he sometimes shares some of the insights they’ve learned over the
decades, and I have a lot of respect for the depth of thought that has brought. There’s a great
piece on player types (“Timmy, Johnny, Spike”), my favorite starter piece on randomness
(“Kind Acts of Randomness”), and a thoughtful redesign of the whole game to keep rampant
difficulty creep in check (“New World Order”). Most games don’t even have time for rampant
difficulty creep to set it, much less time to identify it, and then come up with thoughtful
solutions that propel the game forward additional decades.

Money
Magic’s design is heavily based on a revolutionary monetization scheme.
Randomized cards that are also game objects were (mostly) unheard of until
1993. This was the immediate hook that made people pay attention to
Magic – it’s like baseball cards that are also a game. That was a compelling
hook back when baseball cards were a thing that everyone knew.
Magic revolutionized games in many ways, but this may be its most
lasting legacy. Magic created a new genre of trading card games, which is
still going strong in both print and digital. But the concept of randomized
game content that players can assemble on their own has created multiple
genres and re-shaped many more:

The entire free-to-play category of games is based on loot boxes that

are basically card packs.
 The deckbuilding genre of board games such as Dominion are deeply
inspired by the way decks are built in Magic.
The draft format is a popular way to play Magic that inspired board
games like Sushi Go! (2013).
Diablo was originally going to be a game like Magic, sold in booster
packs. While that didn’t happen, the idea of random rewards that
players assemble into a single thing has been the focus of many role-
playing games ever since. This style of RPG did exist before, but the
emphasis on high levels of randomization and character building
increases post-Magic and post-Diablo.

How does this affect game design?

It’s easy to see how monetization affects the design of Magic. The entire
game design is built around making sense of selling a randomized pack of
cards. Everything starts there. Deckbuilding is a clever way to allow players
to incorporate these new cards from purchased packs into their game. The
rules of the game are flexible enough to allow all the cards that come out of
the packs to work together.
It’s also easy to see how the outrageous success of this monetization plan
led to the history of Magic game design. The business plan of Wizards of
the Coast is all about releasing more packs with more new cards. This
requires constant exploration of new card mechanics and innovative new
ways to get as much value as possible out of every mechanic that exists.
And a continual near-impossible effort to keep all those cards balanced,
especially once the tournament scene takes off. It’s a great example of why
game design matters – the core ideas of Magic are strong, but we wouldn’t
be talking about it 30+ years later unless the content design managed to
keep that design fresh.
Technology
We don’t really think of card games as a technology, but Magic wouldn’t
have happened without access to advances in printing that came about
because of the non-game collectible card market. Carta Mundi in Belgium
already had large printing presses that were able to sort and distribute cards
into packs. In some ways, Magic was a logical extension of improvements
in the collectible card industry.
How does this affect you, the clever game designer?
Many of the details of early collectible card games were decided by
technology. Rarity levels only exist because the card sorters could handle
them. Cards were sold in randomized packs and randomized decks because
that’s what people knew how to do. Non-randomized decks were possible
and came along a few years later. The limitations and costs of printing have
a huge effect on decision-making in the board games world. Individual
game designs can change based on the availability of specific types of
pieces – either avoiding expensive or fragile pieces, or building a design
around a cool new technological option. And costs and limitations have a
huge effect on corporate decisions. Magic: The Gathering only exists
because Richard Garfield first pitched Robo Rally, but Peter Adkison
thought it would take too much effort to print all the parts and asked for
something simpler, like a card game.

Senses
When it first appeared, Magic did a lot of things right. Randomized packs
made every game different. The mana system balanced play. The different
colors gave players different factions to align with. One key idea that gets
less attention is the focus on art. Nearly half of every Magic card is pure art
with no effect on the game rules. And this is critical to the success of the
game – part of why Magic is fun is that the cards look so good.

Emotion
What does it feel like to play a game of Magic?
There’s frustration. Usually the satisfying motivating frustration. The
annoying discouraging frustration sometimes pops up, normally around
either the mana system or other players’ choices.
There’s fiero, the excitement of victory (terminology from Lazzaro’s 4
Keys 2 Fun). Winning a big game feels great.
There’s the feeling of being clever that is common to many strategy and
puzzle games.
But Magic takes this one even further by adding more layers to the
cleverness.

You can be clever in a particular choice at a particular moment. The

combat puzzles as discussed above.
You can be clever at the strategic level by maximizing the deck’s
strengths to pull out a win.
You can be clever before the game even starts by building a good deck
with the right cards to ensure victory.

That last one is especially important because Magic gives you all sorts of
ways to make victory your victory. When you’re playing with a certain
deck, it’s yours in a way that most other games can’t manage. Sure, I like
being the dog in Monopoly, but I don’t feel clever for being the dog. And I
don’t go online to discuss what it’s like to be the dog and how to do it
better. Building a deck in Magic is a statement of identity and ownership
and belonging that gives the player an emotional investment in the game.
 This can be seen in other trading card games that go even further. Legend
of the Five Rings (1995) was a collectible card game where different
factions battle in a magical world inspired by Asian stories and myths.
Legend of the Five Rings had a clever hook where the story of the world
was being written alongside the game and players could influence the
results. Winning a major tournament might win a certain reward for your
faction, which would appear in a story on the website a few weeks later, and
have narrative ramifications for years to come. This led to strong
identification with the factions, and players taking on that identity in a way
that drove strong loyalty to the game. There’s a story of two Naga players at
a major tournament at GenCon who refused to battle each other because
their faction wouldn’t do that (so they resolved the tournament win another
way). The story is told in the Imperial Herald v2 no3, and mentioned in the
Way of the Naga sourcebook. Legend of the Five Rings put this
factionalization front and center by putting the faction on the box of the
starter decks. So before you even played the game for the first time, you
were making a choice that had an emotional impact.

Narrative
Magic: The Gathering is an interesting example of how hard storytelling is
in interactive media. Magic works hard to build cohesive worlds and tell
compelling stories. But instead of the luxury of hundreds of pages of text
nicely arranged in a book, Magic has to tell a story one sentence at a time
on randomly distributed cards. The writer has no way to control the order in
which the reader/player experiences the story. As you might imagine, that
can make it difficult to tell a coherent plot. As such, Magic writers often
focus more on the worldbuilding side of storytelling, with key plot points
told outside of the game itself through website short stories, books, and
other supplemental material. It’s an interesting and difficult challenge, and
the extreme case of the problem of truly interactive storytelling.

Meaning
What is Magic about? What does Magic say about the world?
As is often the case, there are lots of possible answers to this.
To me, Magic is about Infinite Possibilities in Infinite Combinations
(phrasing borrowed from Star Trek). Deckbuilding is about trying to find
creative ways to combine different pieces together. Each piece may be weak
on its own, but together they can be stronger than each piece is alone. Every
card has value, it just has to find the right environment that embraces what
makes it special.

Content
Magic provides a clear demonstration of another way to dissect games.
Games are made up of two things: rules and content.
Content is the stuff of the game – the enemies in a combat game, the
player’s powers in an RPG, the levels in a level-based game, the cards in a
card game. Rules are how these fit together and relate to each other to make
the game work.
An analogy I like here is cooking. When you’re cooking, you have a
bunch of ingredients: stuff, physical objects. And you have a recipe. A list
of rules that tell you how to use the ingredients. When you’re cooking, you
need both. A good recipe with bad ingredients does not lead to good results.
Neither does good ingredients with a bad recipe. The same is true in games
– a good set of rules alone doesn’t make a good game. Having the best
possible first-person shooter mechanics doesn’t matter much if you have
terrible level design and bad values on the weapons. And vice versa – great
card designs will never save a card game with bad or boring core
mechanics.
One way to see this is to look at expansion packs. In most cases,
expansions to existing games add content without adding rules. A new
update to Fortnite doesn’t change how Fortnite works or how weapons
work or the goals of the game, but it adds more stuff. More items. More
locations. More quests. Magic: The Gathering absolutely does this – for
Magic, a new set means a bunch of new cards, not a new way to play the
game. There may be a few new keywords or card types, but these are
(generally) just new text on cards that interacts with the existing systems of
the game, not new systems.
For a board or card game, this is the difference between the rulebook and
the pieces. The board, the cards, the tokens are all content. Not only
because they are physical things but also because those physical things
point to things in the fiction of the game. Within the magic circle, that token
isn’t a token, it’s an enemy out to thwart you. The game piece is content,
but so is the way the enemy works in the game. The enemy’ stats are
content. The rulebook explains what you do with those stats and powers.
For a digital game, the simplest distinction is the difference between code
and art. Code determines the rules. Art provides the equivalent to the
physical piece in a board game – the manifestation that the player
manipulates to get to the thing in the fiction of the game. But there are lots
of other types of content, too.
So some types of content are obvious – Magic cards, Halo levels, God of
War enemies. Some types of content can be harder to disentangle. When an
expansion pack adds a new enemy, it’s not just art. The enemy also has a
bunch of numbers to determine its damage and movement speed and so on.
These are also content. And maybe the enemy has a new way of moving
through the world – let’s say it can teleport short distances to reach the
player very quickly. This type of movement is a new feature that requires
new code and functionality and rules. The ability for enemies to move this
way is not content, it’s rules. The specific ways that this particular enemy
uses those rules is content. If another enemy in a later expansion uses the
same type of movement, it might change up all the specifics of the teleport
– the distance and speed and animations and visual effects would be new
content. But the way a teleport works – the functionality of the teleport
system – would be shared.
This can easily be seen in Magic cards. Most Magic cards are full of
content. Yargle is a 9/3 Frog Spirit creature that costs 4 and a black mana
(Figure 7.3). This is a new combination of numbers that hadn’t been done
before, but it just works. The current rules can handle this without any
changes. Some Magic cards require new rules to make them work. The
March of the Machine set added Battles, a new card type that needed new
rules to explain it (Figure 7.4). But within the set, there were 36 different
pieces of content that used these new rules.
Figure 7.3 Yargle.
 Figure 7.4 Battle card.

Repeating Core Loops

Rules are where the moments of the game come from. To understand a
game, you first have to understand the core gameplay loop and how that
creates an enjoyable experience. But that’s just the first step. Most games
last longer than a few loops. Creating a fun core loop is challenging, even
for experienced designers. But a commercially released game needs to
repeat that core loop hundreds or thousands or even millions of times and
remain compelling and enjoyable. Most games these days are hoping for at
least a few hours of fun, if not dozens or hundreds of hours of gameplay.
While this sometimes can be done with just a core loop, like in chess, it’s
usually far easier to vary the core loop by adjusting the content.
God of War has a fun core loop. It’s fun to destroy enemies. But if the
game had the same enemies with the same player abilities from start to
finish, it would get boring and repetitive very quickly. The bulk of God of
War is a continual stream of new enemies, new abilities, and new situations
that keep the player on their toes. Once the player gets used to how the core
loop works, the game throws something new at the player and asks them to
master it. The fun of games is challenge and content provides a stream of
new challenges.
Magic: The Gathering has a fun core loop. Build up resources and attack
with fun little battle puzzles. If there were only two decks with the same
creatures over and over again, most players would play this game a few
times, have a good time, and then move on. The influx of new cards and
new combinations make Magic a game that has survived 30 years and is
still going on strong.

Progression
Good game content is about providing that constant stream of challenge and
variety.
Variety is hard. If you have a good core loop, you don’t want to mess it
up. You want that core fun to persist through the entire game. But you also
want it to be different every time. How different is different enough? How
different is too different?
Are vehicle level in a first-person shooter like Halo too different? Some
people thought so, but others consider them to be the most fun in the game.
The secret is finding ways to make the core loop feel fresh and new
without changing how it works. Vehicle levels in Halo are still about
shooting enemies and moving through the level. You might be moving in a
new way, but the experience still feels the same.

Difficulty Curves
When we talk about difficulty, we often talk about it as a curve.
 The difficulty of the game should change over time. The first thing the
player does shouldn’t be as hard as the hardest boss. So, what should a good
difficulty curve look like?
A good curve makes sure that every player is challenged, but not too
challenged, at every moment. Which is literally impossible. Every player is
different. Different players find different things to be difficult. No difficulty
curve is perfect for all players. But a good curve keeps most players in the
engaged part of flow for the majority of the game.
Generally, it’s good if difficulty goes up over time. Start easy, let the
player get their bearings, and make things a little harder at each step. We
can graph this relationship, with time on one axis and difficulty on the
other, as seen in Figure 7.5.

Figure 7.5 Difficulty Curve A: Straight line.

But this probably isn’t ideal – you don’t want the game to get too hard.
So, we’ll level it off at some point once we reach our desired difficulty for
this game, as seen in Figure 7.6.
 Figure 7.6 Difficulty Curve B: Levels off after a while.

At the simplest level, this can describe many games. Just adjust the
starting point, slope, and leveling-off point and you have a useful tool to
talk about how different games manage difficulty. But this graph isn’t
perfect. Not everything is going to be a flat difficulty. In Magic, most turns
are fairly straightforward, but occasionally there’s a big battle that creates
extra challenge for each player. And what about bosses? Many games want
to have occasional moments when the difficulty peaks, when the player is
pushed extra hard and has to really demonstrate their maximum skill. We
can add some peaks to our diagram to show that., as in Figure 7.7.
 Figure 7.7 Difficulty Curve C: Add peaks.

And we don’t want to make this too shocking and sudden, so maybe we
should work our way up to those difficult peaks more gradually, like in
Figure 7.8.

Figure 7.8 Difficulty Curve D: Gradually rise to peaks.

This leads us to what is commonly referred to as the “Rollercoaster

Model” of difficulty (Figure 7.9). Like a rollercoaster, the game has peaks
and valleys and the player goes up and down in a fun way that keeps them
excited and engaged. After a big challenge, there is a resting period and
then a gradual increase to a new peak.

Figure 7.9 Difficulty Curve E: Shovel Knight is a rollercoaster.

Many games have curves that look a bit different than the default. This
isn’t a perfect goal for all games – you need to consider what you want your
difficulty curve to look like for your game’s needs. For example, some
games aren’t about maintaining a peak of challenge and adrenaline. The
recent growth of cozy games as a genre demonstrates that a game doesn’t
need to push difficulty curves to maintain player interest. Games like
Stardew Valley (2016) or Animal Crossing succeed with their audience
because they’re gentle and forgiving, not because they are aggressive and
challenging. These games provide uncertainty through exploration or dialog
rather than challenge, so their curves might be something more like Figure
7.10.
 Figure 7.10 Difficulty Curve G: Animal Crossing is low and flat.

Each game has a unique target curve. Some games start off hard and stay
hard. Some have bigger or smaller peaks. The difficulty curve for an open
world game like Grand Theft Auto 3 (2001) is hard to plot because it’s
really up to the player (Figure 7.11). If they want an easier challenge, they
will often choose to head to a different part of the game to get that. Giving
players more agency gives players the ability to adjust this curve on their
own, which is a powerful way to make sure the difficulty curve adjusts to
each player’s needs.
 Figure 7.11 Difficulty Curve E: Grand Theft Auto 3 has multiple
lines.

This form of difficulty tracking is mostly intended for video games where
there is a long sequence of single-player content for the player to consume
in a linear fashion. GTA3 breaks this rule by not following a linear
sequence. Board games break all of these rules. Mapping content over time
doesn’t really work for a game like Magic: The Gathering where content is
digested over many play sessions and the difficulty comes from opponents’
decks. There are ways to map the standard stages of a single session of
Magic, but they don’t use difficulty.
The real value in a difficulty curve when developing a game is
identifying your target so you can evaluate how your content matches up. If
you want a rollercoaster, say that up front so you can build levels to that
goal. Then when you’re reviewing levels in the game, you know what
you’re looking for. Similarly, if you’re trying to build a cozy game, that
gives you clear targets for the types of content you want to design. A
difficulty curve can be a great way to get the team focused on a shared
vision.

Fun Content
How do you make fun content?
This is a long topic that deserves its own book. And there are a few good
ones – I particularly like Level Up! by Scott Rogers. But in case that didn’t
spark you to immediately run out and buy it, let’s discuss.
Good content is a reflection of a good core loop.
What makes the core loop fun? How could it be done slightly differently?
Good content draws inspiration from the rules and features and systems
of the game. Is your game about moving and shooting? What if you could
move a little differently? What if you could shoot a little differently? What
if the thing you were shooting required you to do it in a different way?
The goal of content is to diversify the core loop while retaining the
qualities that make the core loop work.
Magic: The Gathering does this well.
The core loop of Magic is a combination of growth and combat. The
cards provide countless variations on these themes.

Growth is normally done through lands. There are lots of different

lands that do this in slightly different ways.
Growth is normally done through lands. What if it were done through
artifacts instead? Figure 7.12. is an example of this.
 Figure 7.12 An artifact that produces mana.
Growth is normally done through lands. What if it were done through
creatures instead? Figure 7.13 is an example of this.
 Figure 7.13 A creature that produces mana.
Combat normally shows you all the options ahead of time. What if
there were surprises? Figure 7.14. is the token used for one ability that
does this.
 Figure 7.14 A creature that is a secret.

Magic is a game that also will provide ways to completely break the normal
rules and patterns of play through card text. This works for Magic but needs
to be handled carefully. Some games require more focus on the standard
core loop to work. Magic has a core loop that is set up with lots of ways to
swap out different content. Try to design core loops like that. Scrabble is a
great game, but it doesn’t have lots of ways to vary up the game and change
it over time.
A great way to do this is to make content that makes other content shine
in new and interesting ways. If your game gives the player special powers
and abilities, then you often want to design the challenges to showcase
specific powers and abilities.
In Halo, the shotgun is a fine weapon. Until you get to the Flood, and
then it really shines. The flood is a very different style of enemies than
you’ve seen before, which forces you to shift up your strategies. The core
loop hasn’t changed, but changing the content has made the game fresh and
new again.
Good-level design and progression is built around this. Once the player
masters one skill, give them a new skill. Let them play around with it for a
bit, then introduce a new enemy or other challenge that requires skilled use
of the new skill. Then add variants of the new enemy that require using the
new skill in a slightly different way. And this isn’t just a combat thing.
Where’s My Water? wasn’t based on enemy progression, but it absolutely
used this form of progression – each level pack was built around a new
object in the game world, with a few levels to get used to it before
combining it with all the previous combat to create new and challenging
puzzles. Then the next pack introduced something new. But every new
thing followed the previously established rules.
The deepest form of content design is content that doesn’t just build on
the systems of the game, or the previously introduced content. The best
content builds on the themes of the game.
 Civilization is about history and power. Each civilization has a
different relationship with power, based on historical sources.
Bioshock (2007) is about choices and free will. Each area of the game
presents a different take on free will and how it affects the characters
in the setting.
Magic is about infinite diversity. Magic sets are constantly trying to
come up with new ways to twist the rules and lead the player down
new roads.

This is part of the reason to think about the meaning of your game. If you
know what the game is about, you can design around that and think of
interesting new content to show off the heart of the game.

Knobs
One reason Magic: The Gathering has lasted as long as it has is that it’s a very knobby game.
“Knobbiness” is a term that refers to how many “knobs” the game designer has to adjust
content. Games with lots of numbers and text that can change (like Magic) are “knobby”, which
is good. Games with few things that can change and a lot of restrictions on how those changes
can be made interesting are “not knobby”, which is bad. Look at Scrabble – the game is based
on the alphabet, which the game designer can’t really change, and a number value per letter,
which can change but needs to make sense for word usage. If I had to make new levels for
Scrabble, it would be hard to make them fun. I know – I tried. I worked on Scrabble Go (2017)
for a bit before it was released.
With Magic, each card has lots of numbers that can be easily adjusted. The numbers really
matter and change things. And then there’s that big text box that can include anything (including
more numbers that can be adjusted to make more cards). And more importantly, the vision of the
game is that the text box can literally do anything (within the bounds of the main rules). So,
there’s a lot of freedom to experiment and try new things there, which isn’t always the case for
other games. Having lots of knobs and regularly changing them has made Magic one of the few
(only?) games that has consistently released new content for over 30 years.
Content Flow
Good content is built to fit the meaning of the game, and the abilities of the
player. It’s also built to fit the rest of the content.
The difficulty curves discussed earlier are an example of this – every
enemy should provide a new challenge that fits a nice curve compared to
the challenge of the other content. But good content builds on other content
in other ways, too.
Magic: The Gathering releases content in sets. Each set has many cards
that are each a piece of content that needs to follow good design principles.
But the set itself is also a feat of game design. Players tend to experience
each set as a cohesive whole, especially in limited play formats like draft or
sealed. Each card needs to be well designed but also fit in well with the
other cards in the same set. A card that is interesting in one set, such as a
card that benefits from other enchantment cards, might not be very
interesting in a set without many enchantments. Content is like systems –
they don’t exist in a void and depend on the content around them.
I experienced a lot of this working on Where’s My Water? We designed
each level to be fun and interesting on its own but also to fit into a pack of
20 levels around a consistent theme, usually a new feature added in that
pack. Each level in the pack had a purpose. Most levels were demonstrating
a new quality of the new feature, showing the player how the new feature
interacts with the existing features. We spent a lot of time and effort making
sure the levels were in the right order so the player wasn’t overwhelmed
with too many new ideas at once. We often had to go back and build new
levels to fill in gaps of knowledge and build the player’s expertise to master
the final hard levels of each pack.
Good Ingredients
So while most of this book is about recipes – systems and other core game
design issues that make a game work – don’t neglect your content. The best
systems in the world don’t matter if there isn’t good content (levels, cards,
enemies, abilities) to show them off.
OceanofPDF.com
Chapter 8
Royale High / Audience
DOI: 10.1201/9781003346586-8

Royale High
Royale High is a great game.
Royale High (Figure 8.1) is one of the top games on the Roblox platform.
It has had over 9.7 billion visits (as of April 2024) and was one of the ten
most popular Roblox experiences in 2023.

Figure 8.1 Royale High screenshot, from March 2024.

If you’re not familiar with Roblox, it is a platform where anyone can

create games and easily publish them for anyone to play. It has grown over
the years to become one of the biggest places for games and other virtual
experiences, with over 70 million users a day (as reported in Roblox’s Jan
2024 financial reports). Roblox’s focus on user-generated content has
resulted in a different gaming ecosystem than console or PC games, with a
different mix of successful genres than anywhere else. There are genres in
Roblox with millions of daily users across hundreds of games that simply
don’t exist anywhere else. I like to compare Roblox to Australia – a self-
contained ecosystem that has resulted in fascinating divergent evolution.
See, Mom, I’m putting my biology degree to good use.
I work at Roblox now, mostly on educational games and assessment
games, but I’ve never worked on Royale High or directly worked with the
team that creates it. And everything you read here is all me – this book is
not sponsored or affiliated with Roblox other than indirectly due to the fact
I work there.

Royale High – Gameplay Dissection

Verbs
Royale High is a little different than most of the games we’ve talked about.
To start off, Royale High has a lot of verbs. And most of them have bonus
alternate forms.

Move
Players have an avatar and can walk around the world.
Players can teleport to various places to get around more quickly.
There are also vehicles and a few other modes of transportation.
Buy
Players can directly buy many things.
Most of these are decorations, either for the player’s avatar or
room or other things.
Some are powerups that make other things easier.
Customize
Each player has an avatar that is separate from their normal
Roblox avatar. Players can buy new items for their Royale High
avatar.
Similarly, players can dress their pets with cute little clothes.
Each player gets a house that they can decorate with furniture that
they purchase. This includes a number of sub-verbs to move,
rotate, and otherwise manipulate furniture.
Trade
Players can trade items with other players.
Action
Tidy
Adventure
Login
Wash
There are various things in the world or the player’s inventory
that players can select and activate. Most of these show a brief
passage of time and then the action is declared complete.
This is used mostly for quest completion. A quest tells the player
to check their locker. When the player clicks on the locker, they
get a reward. The player is then given a new quest, or shown a
timer and can do the quest again when the timer completes.
Mini-Games
“Classes” in the school are represented with mini-games. Mini-
games give the player a focused goal and (depending on the
game) new controls to achieve that goal. These are often drawn
from traditional classroom activities such as musical chairs but
can also be scavenger hunts or other types of mini-games.
 There are also some loose mini-games, such as the locker book-
stacking games.
Mini-games can be incorporated into the quest system similar to
actions.
Spin the Wheel
There is a reward wheel in the world that acts as a special type of
quest with different rewards.
Events
There are frequent events that add new things to do, but they
generally fall into new twists on the above options.

Goals
Royale High has lots of different verbs but organizes them into quests
(Figure 8.2). So while the above list may make it sound like there’s a lot
going on, it really boils down to a simple to-do list plus a few optional
bonus activities. While the verbs are all over the place, the goal is consistent
and easy to process.

Figure 8.2 Lots of verbs with some groupings.

 Completing quests levels up the player, which provides various rewards.
The primary reward is diamonds, which are used to buy things. Buying
things is the real motivating goal. There are all sorts of things to buy –
clothes and accessories for your avatar, furniture and other items for your
dorm room, and so on. There are special limited-time items to purchase,
some of which are given out through loot boxes where the rewards are
randomized and players are incentivized to buy many times to get the rarest
items.
This is a very standard motivational loop (Figure 8.3). Players do things
to get money, which they can use to customize their experience and unlock
more things to do for more money. It’s basically the same loop as Settlers of
Catan – get stuff to make stuff to get more stuff. This works partially
because it’s a loop humans are accustomed to. Go to work. Get money.
Spend money on things. Go to work to get more things.

Figure 8.3 Core game loop with lots of verbs leading to

diamonds, which lead to bigger goals.

This gameplay loop is especially successful in social settings. People like

buying things and customizing their experiences. They really like buying
things and showing off those things to other people. Royale High has
players navigating through public spaces where they can show off their
costuming choices, and players will notice and chat about what everyone is
wearing. Roblox is an inherently social platform, so this sort of socially
motivated display of stuff makes perfect sense.
This motivation loop works well on Roblox, where many of the players
are teens or kids. Many of the top games on Roblox are aspirational,
allowing teens to try out adult-style behaviors – owning homes in
Brookhaven, having jobs in Work at a Pizza Place, doing scary things in
Piggy. And nothing is more adult than showing off your wealth. But this
motivation is not unique to kids. Social displays were a big motivator for
gameplay during the “social games” phase on Facebook and other social
media platforms. Farmville was all about building and maintaining an
impressive space to show off to your friends (or at least not embarrass
yourself). Royale High is just the latest evolution of this trend.

The Past
While the current (2024) version of Royale High has a successful game
loop that ties back to classic social games, it has not always been this way.
The school theme was stronger – most quests were tied to classes, to the
point of requiring students to bring the right books to each class. Most of
the systems and loops discussed above existed but in much simpler forms.
Many games today, especially on online platforms like Roblox, are not a
single unchanging thing. “Game as service” is a common model these days,
where a game is expected to constantly be adding features and content and
events and constantly be changing. This is very different from when old
people like me started making games. When making games for older
console systems, whatever you put on the “gold master” disk was the only
game that anyone outside of the development team was ever going to play.
As a small side note, this makes certain types of game dissection
essentially impossible these days. You can’t dissect every version of
something that’s constantly changing. You can only describe the current
state at a slice of time. When you’re using dissection as a tool to understand
the market or design your own games, that’s fine – you don’t care about the
past or the future, just what does or doesn’t work today. But when you’re
trying to write something for posterity (like, let’s say, a book) that can be
more problematic.
So Royale High used to be much simpler. The player was still about
customizing and building a student in a magical school, so the emotional
core of the gameplay was the same. But the daily loop was different, and
many things that make the game huge today didn’t exist. The skeleton of
the game was the same, but lots of smaller things were radically different.

Challenges
Royale High is not a difficult game.
The quests generally don’t have any enemies or randomization or player
skill moments or difficult choices. The player doesn’t have to manage
multiple resources and balance between multiple quest choices to optimize
carefully. The only real challenge in the quest system is time management –
the player can choose to focus their time on the basic quests or doing other
activities for other rewards. But this only provides minor optimization over
just doing each quest as it comes up.
Similarly, the customization and decorating features are much more
about personal expression than they are about personal challenge. There are
no stats to balance or competitions to win based on how you customize.
Games like Covet Fashion (2013) or Dress to Impress (2023) build very
strong gameplay loops out of nothing but customization. (If you haven’t
tried these games, I recommend taking a look. There are lots of good
inspiration to be had in these games.) These features are rewarding in and of
themselves, with the only challenge coming from personal goals the player
sets for themselves. Loot boxes add some randomness, but this is not
randomness that the player has any control over, so it doesn’t really enter
into the decision-making other than requiring more resources to get all the
rewards. The only place where challenges really exists is in trading with
other players, which isn’t something players get into until they’ve been
playing a bit.
This may seem odd at first. We’ve talked about how challenge is
important for games. How uncertainty is necessary for a game to matter.
You could argue that a game like Royale High that doesn’t have a lot of
challenges isn’t really a game; it’s more of a toy or an interactive
experience. Which is fine. A movie doesn’t have a lot of challenge, and we
still value them for their artistic merit and entertainment value.
There’s a growing appreciation of low-challenge games such as The Sims
or Animal Crossing where the difficulty curve is more of a difficulty slant.
These “cozy games” don’t push players to make tough decisions or
demonstrate difficult skills. Players have choices that affect the space and
make progress through their actions, but the actions and choices aren’t
difficult or stressful. They still have Sid Meier’s “interesting choices”, but
they’re interesting in other ways. Customization is interesting as a form of
personal expression. Completing simple quests is interesting as a way to
feel completion. And while a traditional flow state requires some challenge
to keep people engaged, I posit that different people have different targets.
Some people require high levels of challenge to maintain their flow, while
others only need a little push. A Royale High player working through their
list of quests to level up is just as in the zone as the top Halo player, even if
their experiences are very different.
Cozy games and other low-challenge low-stress games attract a different
audience than traditional gamers. For most of the history of video games,
competition has been king. High score lists and multiplayer deathmatch
modes attract people who like to dominate and be the best at something.
But not everyone is motivated by that desire. Some people just want the
satisfaction of a job well done. Some people just want to find cool things
and show them off. Some people just want to express themselves creatively.
The artist who creates interesting spaces in Royale High (or The Sims or
Animal Crossing) is no less a success than the deathmatch dominator. And
no less a gamer. This has always been true to a certain extent with games
like SimCity and Mario Paint, but it’s becoming a bigger and bigger part of
gaming. The games industry has been learning this gradually over time as
social games and mobile games bring new types of players and require new
types of game designs to appeal to these new motivations (Figure 8.4).

Figure 8.4 Royale High Flow 1: Simple diagram of Actions,

Challenges, and Goals for Royale High.
Dynamics
Royale High’s core gameplay flow starts with two main states.
The player completes quests by navigating around the environment and
selecting the right objects or characters to meet their current needs. The
player then switches into customization mode, buying things and using
them to decorate themselves, their pet, and their house.
Royale High also has a few states that exist outside of these core loops.
These are worth dissecting as they are key to the success of this game.
Players will often spend time in the big public spaces without completing
quests or customizing. They will engage with other players through
Roblox’s chat or voice system. Some of this ties back to the customization
and rewards systems, as players will chat with other players to discuss the
best outfits or trade items. Some of this is just using other players as a
resource for information, asking questions about how to acquire special pets
or equipment. But there are other uses for socialization that really matter to
players.
Some socialization is showing off. Walking around with a rare item gives
players a chance to be noticed. And many will. Players will compliment
each other’s items or pets, or offer trades for things they see. This is a key
motivator for the social loop. Having something that other people want
feels good. And this acts as good advertising for the items in the game – if
rare items impress people, players are more likely to want them. This works
best in a gaming space with strong social hooks built in, like Facebook used
to be or like Roblox is now. If it feels natural to see other people’s stuff,
then people will show off the stuff they have.
The really interesting socialization in Royale High, and in many Roblox
games, is the role-playing. Not like Ultima (1981) or Wizardry (1981) or
Skyrim (2011) or other traditional computer games where RPG means stats
and equipment and monsters. This is actual role-playing like playing a role.
Pretending to be something and acting out improvised interactions is a key
part of tabletop role-playing like Dungeons & Dragons, and is also a key to
the success of Roblox.
Players in Royale High often enjoy playing pretend. But not as an elven
warrior or vampire mage. Royale High role-play might be a wealthy
socialite with the most expensive items. Or just a happy student who is
hanging out with their friend and navigating classes. This isn’t as
complicated or elaborate as playing pretend in Dungeons & Dragons. Most
Royale High players don’t write out elaborate backstories to explain how
their character relates to the history of the realm. But they do engage in the
same storytelling and fantasy play that motivates tabletop role-players.
They just do it in short stints without the long-term progression that stats
and equipment provide. A Royale High player might try on a role for a
single statement, or a quick exchange with a friend. Deeper role-play in
public with strangers is rarer, but it does happen.
Royale High is built to maximize this style of play. In addition to the
basic social and customization experiences that are built into Roblox,
Royale High adds features specifically intended to heighten the role-play
experience. The deep customization system allows players to have a unique
persona on Royale High separate from their normal Roblox avatar. And
customization extends into the player’s space – players can create their own
dorm rooms and show them off, too.
Royale High is not the only experience on Roblox where this happens.
Many games on Roblox encourage and support this style of play, and some
are entirely built for it. Games like Miraculous RP: Ladybug & Cat Noir
(2021) actively encourage players to role-play existing characters by giving
them costumes and locations built for it. And games like Brookhaven RP
(2020) (RP for Role-play) or Roblox High School actively encourage kids to
role-play aspirational roles based on being older and more successful than
their reality. Games like Untamed Animals (2020) or Holocene (2019) let
you role-play as things you can’t possibly be in real life.
As an old-school tabletop role-player and a long-time video gamer, it’s
great to see video games embracing actual imagination play. I love some of
the great digital RPGs, and many of the best of them go to great lengths to
provide players the flexibility of choices and deep narrative that a good
game of Dungeons & Dragons can provide. But I see great potential in
digital games that give players complete freedom to pretend whatever they
want to pretend. Imagination is the heart of all games, so seeing how new
types of imagination flourish on new platforms like Roblox makes me
excited to see the future of games.

Systems
Royale High starts the player off with an experience progression system, a
character customization system, and a related room decorating system.
These set up the player for long-term goals and long-term systems. Start by
making the player know what they should want. Royale High is a good
example of a key idea when setting up Goals – don’t tell the player that they
want something, show them something that they naturally will want. Royale
High doesn’t tell the player that they need a better dorm room. Royale High
shows the player a tiny room, then lets you walk down a hallway where you
see better rooms. Wanting a better room is a natural human impulse in this
situation, and the game doesn’t feel like it’s forcing the player to want
anything.
Wanting better things motivates the player to engage in the core
economic systems of the game. If you want to customize, you need things.
And if you need things, you need diamonds. This guides players very
naturally to the quest system. Quests are the best and most obvious way to
make money in the game, and they do a good job of tying into the core
school fiction to make quests fun.
Royale High has lots of systems, but they all feed back into the core
economic systems (Figure 8.5).

Figure 8.5 A lot of different systems all leading to the economic

system.

Royale High – Non-Gameplay Dissection

Technology
Royale High is built on Roblox.
Roblox Studio is a powerful set of tools that makes it easy for anyone to
generate, publish, and maintain experiences, which are often games. “Easy”
is of course a relative term – building anything with rules or goals or
interaction requires setting up a bit of code. But that’s part of the fun –
beginning creators can create empty spaces that players can walk around
and enjoy, creators who want to learn code can start with simple-to-create
games like Obbies (short for obstacle courses, but many of the players who
use the term “Obby” all the time probably don’t even know that), and
advanced creators can build nearly anything they can imagine.
The platform has a big effect on the design of any game. Roblox is
inherently multiplayer – you have to work a bit to make an experience with
only one player. So the experiences that succeed on the platform are the
ones that craft a fun multiplayer experience. Royale High is a good example
of this – the core loop and core systems of the game don’t require the player
to interact with other players, but they are all enhanced by it. Showing off
your possessions – your room and your locker – is a powerful motivator in
any game where other players can see your stuff (see: Farmville and many
other Facebook games).

Senses
Royale High feels like you’re living in a fancy castle. The visuals enhance
the aspirational nature of the core loop. Players want fancy things because
they see fancy things. But there are a lot of different ways to be fancy,
allowing for personal expression. As a UGC (user-generated content) game,
Royale High has a lot of respect for the individual creators who make these
items – many items list the creator so players can track down artists they
like (Figure 8.6). Players like to see Roblox experiences as individual
creations of people like them, rather than faceless corporate creations. At
this point, a game like Royale High is actually being built by a larger team,
but that personal style still influences what the game looks like.
 Figure 8.6 Listing the creators of individual items.

Narrative
Royale High has a simple but effective narrative (Figure 8.7).

Figure 8.7 Story in Royale High.

There is a full tutorial that introduces the concepts of the fantasy world
and the player’s role in it. But most players aren’t motivated by the story.
 The real role of the story is to provide a compelling setting. The space is
interesting and has lots of fun little moments of personality. These provide
great hooks for players to build their own role-play experiences around. Or
just pretty scenery for players who aren’t motivated by role-play. There is
enough narrative to explain what’s going on (magic high school) and the
motivations for the player’s in-world goals (you are a student). That’s all
most games need.
For the players who do role-play, even a little, having a strong backdrop
is a great help. But they go beyond those starting points and build their own
story. The game doesn’t tell them which friend should play which game
with whom, or what fashions to wear. The players make up their own
motivations, often mixing real-world and in-world motivations. The real
stories are the stories the players tell about themselves and their actions and
(most importantly) their relationships with each other. And the game
embraces this and gives them the tools to make this easy.

Emotion
The core loop of the game depends on player’s desire to be a good student
and part of the community. Social play hits on a few different emotions. As
discussed, the desire to have nice things to show off to others is a strong
motivator. This taps into deep emotions based on our social desires as a
species – pride in accomplishments, shame at not meeting society’s
standards, and coveting that which one does not have.
Royale High does also have some narrative pulls at emotion. Characters
in the world talk to the player and ask them for help and encourage them to
do things. These are important to driving the player through the content and
they work, but they are much less powerful than the social drivers. When
walking around Royale High, there are always more people talking about
who has what halo rather than what the NPC Guards are up to.

Money
How does Royale High make money?
Roblox as a whole is based on Robux. On the platform, players can buy
Robux which they can spend on cosmetics or in any experience. Robux
spent in an experience are shared with the developer. Developers also earn
Robux when a player who is paying for a subscription spends time in their
game, even if they don’t spend anything.
The developer can then cash out their earned Robux as real-world
currency. Roblox takes a cut of the currency, similar to how distribution-
only services like Apple or Steam take a cut. The percentages that
developers get on Roblox are different than the percentages distributed on
distribution-only systems, but the services provided are also different.
Roblox also provides a lot more – the tools to build games, servers to host
the data of the game, and moderation and safety controls to ensure your
game is in a safe space for players of all ages. And when Roblox appears on
other platforms like Apple or Google, they get their normal cut on top of
Roblox’s. So Roblox takes a larger cut than other platforms but also
provides more services.
How to monetize within this system is determined by each Roblox
developer. Royale High creates its own internal economy that players can
earn and spend for customizations and other things. The real driver of the
in-game economy is the customization system.

History
Royale High was launched in 2017. Roblox was a smaller platform at the
time, and Royale High did well in that smaller ecosystem. At first, Royale
High was very focused on the magical school role-play. The first versions
of Royale High were strongly based on the TV show Winx, but after some
discussions with Winx’s lawyers those elements were removed. But the
role-playing focus didn’t change. Royale High didn’t invent the Roblox
role-play genre – other games like Robloxian High School (2009) or
Welcome to the Town of Robloxia (2010) have similar themes and some of
the gameplay. But Royale High built a strong game from those ideas, and
there’s enough room in Roblox for multiple success stories.

Meaning
There are two ways to look at Royale High.
Royale High is about stuff. Wanting stuff, showing off stuff, and
acquiring new stuff. Commercialism for the sake of commercialism is
generally frowned upon, but Royale High isn’t just commercialism.
Royale High is also about aspirational role-play. Being someone else.
Exploring who you want to be. Playful imagination with friends.
Just like in real life, these two somewhat contradictory meanings can live
together and both be true. I would argue that the role-play goals elevate the
commercialism, at least somewhat. Wanting things can be dangerous, but
wanting things with friends is a better place to be in life.

Audience
Interactivity
Games are about interactivity.
Games are about wanting things. Striving toward goals.
Game designers work hard to craft goal systems that make sense and
guide a player through a positive experience.
 What happens when the player doesn’t want that experience?
Players are diverse and crazy and want weird things. If you doubt this,
run a few sessions of a tabletop role-playing game. You will quickly find
that no matter how much you construct a beautiful and elegant plot, the
players will choose to focus on something else. Why battle the Ancient Evil
Necromancer when there’s a peasant girl who could use a new doll? Do you
know where the best doll maker in the land is? Oh, and the doll maker made
a joke to me? Now I need to spend a full game session on finding the right
gift to woo him. These sorts of distractions always happen in tabletop role-
playing games, and good players and leaders know how to roll with it and
craft a fun experience around whatever crazy thing the players may want to
do.
This isn’t really an option in a digital game. In a digital game, the player
is interacting with the code and if the code doesn’t know how to woo a
toymaker, it’s not going to happen. Many games just accept this – there is
one thing to do and if you don’t want to do that, you should be playing a
different game. If I want to make friends with the goombas in Super Mario
Bros, there’s not a button to do that.
The early exception to this was Infocom. Infocom was a company that
created weird games entirely out of text, such as Zork (1977). The player
would type a command into the game, and the game would respond with
text saying what happened as a result. These are still around and are known
as interactive fiction. Text is a lot easier to generate than images, so these
games could get away with giving the player a lot of freedom. Most of
Infocom’s games had a single plot that you’d eventually be guided toward,
but if you wanted to try silly things for a while, the game would let you
jump up and down or play with the leaves and would respond with cute
little snarky commentary while you did.
 This sense of freedom was lacking in 3D worlds until Grand Theft Auto
3. GTA gets a lot of attention for its adult themes and urban style, but it
really grabbed gamers because it gave them freedom. GTA has a plot you’re
supposed to follow, but the vast majority of time in GTA is spent wandering
the city and doing goofy things. If you see a ramp, you can jump off it in a
motorcycle. If you see a long stretch of road, you can race down it. If you
think “what will happen if I do that?”, the game generally has an answer,
and that answer is generally fun to do. GTA was the first game that gave the
player a wide range of agency in a 3D space and had the 3D space respond
to anything the player wanted to do. This was huge and led to the whole
genre of wildly popular open world games, including hits like Skyrim and
Legend of Zelda: Breath of the Wild (2017).
These games build freedom by having flexible systems that allow varied
responses to player input, allowing for emergent gameplay. Royale High
reaches the same goal in a different way. In many ways, Royale High is
more like a tabletop role-playing game or a MUD (multi-user dungeon)
than a digital open world game. MUDs were early online games where
players could move through digital spaces and interact with other players.
Most MUDs use text interfaces similar to interactive fiction. But you can
type things to other people and engage socially. Some MUDs had strict
rules that mirrored Dungeons and Dragons or other tabletop systems, but
many of them were more free-form, allowing players to type anything.
Other players could choose to run with that idea, or just ignore it.
Royale High gives players freedom by opening the experience up to true
role-playing. The game doesn’t respond to players’ agency, but other
players do. Players can make up their own stories and live them in the 3D
world that the game provides. Other players can join into the make-believe
and play along. The game doesn’t control this or moderate it, but it
encourages it and (as previously discussed) has the right setting and features
to allow this to work. If you want to pretend to be a princess or a
businessman or a duck, the game doesn’t stop you. The game doesn’t insist
that you run to the right and jump, it just lets you hang out and say whatever
you want (within the limits of the trust and safety filters Roblox uses).

Dissecting Audience
To truly dissect a game, you need to dissect its audience. Not literally, of
course.
Games are built from actions and goals and challenges, but none of this
matters if the player doesn’t want to interact with those systems. Games are
interactive, so the player has a lot of input into the experience. The game
offers the player a lot of choices, but the player decides what to put into that
choice. A good game gives the player the freedom to make those choices
meaningful. Meaningful to that player in that moment and what they want.
Understanding the players of a game is hard. A popular game will attract
many different people who enjoy the game for many different reasons. Even
a single feature may be seen very differently by different players. The
players themselves may not really understand why they like something, so
even if you bring players into a room and ask them questions there may be
no clear answer. And players are weird. As mentioned, players often want to
do strange things that don’t make any sense.
If you’re dissecting an existing game, it’s good to talk to the existing
players. They may not always know what they want, but sometimes they
do. And if not, you can find out what they do. If players are having
unexpected interactions with a system, it’s often a sign that the players are
experiencing unexpected goals. And those goals are driven by emotional
desires that the game is not fulfilling. As a game designer, if you can find a
way to tap into that desire, you can make your game better.

Player Creativity
In the class I teach, I’ve been running the same D&D module (Fishing for Gods In Strade’s
Gallows) for groups of college kids for years now. It’s a simple classic starting point – the
players’ characters are hanging out in a tavern, and someone offers them money to deliver a
package to a dangerous location. But I’d estimate only about half of the time do the students
actually make any progress on the actual quest.
One time they decided to kidnap the person who gave them the quest, since they must know
useful information. The city guard didn’t like that.
One time they met a random kid on the road and made it their mission to improve the life of
the kid.
One time they got into a fight with other people in the tavern, who I’d previously decided
were mobsters. That didn’t go well.
Multiple times, the group decided to fight among themselves, stealing the package back and
forth between players even though they don’t get any reward unless the package makes it there
safely.
After we play, I always stress to the students that all the strange desires and goals that they
saw at the tabletop are also things that players of digital games want. The digital players just
don’t have a good way to tell the designer what they want.

Understanding Audiences
Game designers over the years have attempted to define their audiences.
The best-known example of this is Bartle’s taxonomy of MMO players
(Massively Multiplayer Online). He breaks players down into four groups
(Socializer, Explorer, Achiever, and Killer) based on what he’d seen in
talking to many players of early MMOs. This particular taxonomy has
become popular, to the point where people try to apply Bartle’s categories
to every type of game. To me, this is missing the point. Bartle did a great
job of breaking down the types of players who enjoy a particular genre at a
particular time. The secret sauce there is not his final results, but his
methodology. Don’t think about how many Killers you have in your game,
think about what your players want to do and what they get from the game
emotionally.
Mark Rosewater, the lead designer of Magic: The Gathering, created a
similar taxonomy of Magic: The Gathering players. Tammy/Timmy players
want big experiences, Johnny/Jenny players want to express themselves
through the game, and Spike players want to excel. These are great
categorizations as they get at not just what the players do differently in the
game, but the root motivations that drive them to play this way.
These are good examples of how to think about an existing player base.
What do they want? What do they enjoy? How can the game design make
them happy? How do you get to this level of understanding for your game?
Especially if you’re not the target audience?

Understanding Other People

Understanding audiences is harder when you’re not part of that audience.
Many game designers start by building a game that they want to play.
Which is a great source of inspiration and often leads to a great game. And
in the end the game is guaranteed at least one fan.
But what about when the target audience is someone different from the
game designer?
I made games for kids for many years. And while I consider myself to
have a healthy inner child, I haven’t been a kid for many, many years. I can
remember what it was like and think back on the joy I had playing Atari
2600 games on holiday breaks. But I don’t know what kids today want. I
don’t know how kids today interact with technology. I don’t know how
today’s games interact with the complex and ever-changing lives of kids.
 This can be especially hard in a new space like Roblox. I did not have
Roblox around when I was a kid. There are whole systems of interaction
that I will never directly experience in the same way. Roblox is an
inherently social experience, so kids tend to travel in packs. They pop into a
game, devour as much fun as they can as quickly as they can, then pop to
another experience. Sometimes these roving bands of kids only spend a few
minutes in a game, even a game that they like and plan to come back to
later.
As a game designer, this can be difficult to design around. Tutorials need
to be non-existent or built into the feature only when encountered. Players
like some permanence and enjoy building things, but those things need to
be quick and easy to access (and show off) at any time. Players decide
whether or not to stick with a game in a few seconds, so those first few
seconds are key. If the game is trying to make the roving bands of kids read
text boxes in that time, the game may have trouble finding an audience.
As a game designer, it’s even harder to build around these play patterns if
you don’t know they’re there. None of this is obvious just looking at
Roblox or the games (or non-game experiences) on the platform. Even
talking to a few random players inside popular games doesn’t provide a
complete picture. Most players in most games aren’t looking to have a
random discussion with a stranger. Roblox is social but it’s also safe, and
many kids today are savvy enough on internet safety to avoid conversations
with people they don’t know. Much of the Roblox audience doesn’t want to
talk to you because they’re fully engaged with their friends.
Talking to People
So, as a game designer, how do you figure this out? How do you understand
groups of players who are motivated by different goals and different
experiences and different relationships?
 The most important thing is to talk to the actual audience. If you’re
interested in an existing audience that likes to gather online and discuss
games they like on Reddit or Discord, this is easy. Go there, engage with
people, and you’ll learn a lot.
But many games have audiences that don’t want to talk about gaming
online. Roblox is popular with kids and teens who might be talking online
but not in public forums. Many mobile games are played largely by adults
who don’t have the time or inclination to talk about their gaming lives
online. Candy Crush Saga does not have a wildly popular Reddit subforum
or a big presence on TikTok (yet). How do you talk to these people?
The rise of influencers can be a boon for understanding certain
audiences. I don’t often get to talk to Roblox players directly, but I can learn
about the games they’re playing indirectly through YouTube and other
social media. This is where many players on Roblox and other platforms get
their information about what games are interesting, so it can be a valuable
place for indirect insights.
If you’re working for a larger studio, you can pay to talk to the players
directly. Focus groups and playtest sessions are great ways to learn about
your audience or related audiences.
As a game designer without access to these resources, you need to be
scrappy. Talk to everyone you know who fits your target audience. I have a
nephew who loves Roblox and has taught me a lot about his play styles. But
they’re very different than the play styles of other kids I know. I talk to
every family member who plays and try to understand how they play and
why they play. Information like this is gold and insights from the actual
audience are often what makes games in these spaces succeed or fail.
 No Demographics
When talking about audiences, people often discuss demographics – the age, gender, race, and
other basic physical properties of their audience.
I don’t think demographics are very useful. Defining an audience by these traits usually only
hits on the simplest, most superficial qualities of the audience. Pick a specific age, gender, and
race and you still haven’t really narrowed it down. My kids are Americans aged 15–25 but
between them and their friends there are many different stories and desires and purchasing
patterns.
I don’t want to know how old they are, I want to know what they want. I want to understand
their desires better than they do so I can make something that delights them.
Demographics are a very sloppy shorthand to talk about audiences. I admit I’ll use the terms
sometimes, but only as a starting point. Dig deeper. The understanding you need is not on the
surface, it’s deeper.

Understanding Non-Existent Audiences

Understanding a game’s audience is even harder when the game doesn’t
exist yet.
Understanding the audience for a game that only exists in your head can
be done in two main ways. First, pretend like your game is a different game.
Second, pretend like there are players.
Every game is built on the shoulders of giants. A new pitch generally fits
into an established genre, or at least has strong connections to games that
already exist. This helps for many reasons. First, companies want their new
games to sell to the audience of those old games. To understand how to
design for a new game, look at how previous games have done it. Find the
closest analogues to your idea and see what their audience wants. If you can
find something that they want but current games don’t provide, that’s a
great place to start your ideation.
Second, you can define what your theoretical audience would look like.
This is especially useful if you’re trying to create a new audience or
combine multiple existing audiences around something new. In the world of
UI/UX, this is known as creating a user profile. Don’t just say “my game
will appeal to kids”, but break down exactly what sort of kid your game will
appeal to. Think about specific examples of the sorts of kids you want to
play your game. What is their experience in the game likely to be? How
will it be different from the experience of another kid? Your target audience
should include a diversity of experiences and desires and fulfill them all in
ways that might be different but are all based on the same game. Creating
user profiles is fairly common in UI/UX and fits games well, but the
practice hasn’t become common yet. It will.

Understanding Goals
For a game designer, the important things to know about the players include
the following:

What do players want?

What excites players?
What turns players against the game?

If all you know about Royale High players is that they like to complete
quests, that doesn’t tell you enough to design around them. Just adding
more quests wouldn’t satisfy the players’ goal. You need to understand that
players are really doing quests in order to get money to buy clothes and
other customization options. And even then, there’s a big difference
between role-players, and players who are customizing to express
themselves artistically, and players who are customizing to impress in social
situations. Role-players might like more narrative hooks and special events
and ways to take on meaningful items temporarily. Artists might want more
freedom to customize things themselves. Social players might want
exclusive items that they can earn based on things other than just in-game
currency to make their wardrobe more elite. The more you know about your
audience, the better you can design things that delight them and keep them
excited about your game.

OceanofPDF.com
Chapter 9
Papers, Please / Meaning
DOI: 10.1201/9781003346586-9

Papers, Please
Papers, Please is a great game.
Papers, Please (Figure 9.1) is a unique game about bureaucracy and
balancing difficult choices, which came out in 2013 for PCs and mobile in
2014. Papers, Please was hailed as a critical success immediately and
(along with other games like Spelunky) helped grow the indie scene and
respect for games as an art form.

Figure 9.1 Papers, Please screen.

Papers, Please – Gameplay Dissection
Verbs
Papers, Please is a less active game than many of the others discussed in
this book. There is no variant of the “Move” verb. The player never directly
controls an avatar. The player’s direct actions are very simple:

Stamp: Pass or Deny each applicant. In some cases, the game asks you
to click on a supporting piece of evidence to show why you’re
rejecting.
Allocate: Choose where to spend money at the end of the day.

There are a few other special actions that are used for specific story points
later in the game, but these aren’t core to the game, so they will be avoided
to avoid spoilers.
While playing the game, it’s clear that there’s more going on here than
just that. This is a good example of why it’s important to play the game
yourself when dissecting, not just watch other people play it. YouTube is a
valuable resource, but nothing compares to the feeling of actually pressing
buttons yourself. When you play the game, you see that “Stamp” is not a
complete statement of what the player is doing. The actual experience is
much more than that. There are a lot of player actions that are important but
don’t result in their own final decision or physical action, similar to Candy
Crush Saga where we called these “secret verbs”. Many of these use
separate clicks through separate User Interface (UI) systems but are
building toward the “Stamp” decision rather than leading to their own
separate actions.

Examine: Read over the papers that are presented.

 Research: Flip through the rulebook to check the current rules and the
other supplied data.
Link: Click on the error and cross-reference against the broken rule.
Question: Discuss the discrepancy with the person.
Weigh: Make a moral judgment comparing the infraction (or lack
thereof) against the moral qualities of the person, as expressed in their
words.

At minimum, “Review” can be used as a general verb to define all of the

analysis going on behind the “Stamp” action.

Goals
The game starts with a simple goal to review the incoming people and
stamp them. This is very directly tied to the player actions in the core
gameplay loop, so it’s easy to pick up and start doing and feel good about
completing. People like the satisfaction of a job well done, especially when
there’s a sense of a checkbox being ticked at the end of it.
Then the end of each day makes another goal clear: each time the player
stamps a person correctly, they get money. Money is necessary to keep the
player’s family alive and healthy. The player needs to stamp as many people
as possible each day to earn money to keep their family alive. There is a big
emotional reward to rapid (but accurate) stamping.
As the game continues, the people who come up to the border crossing
start to present interesting little self-contained goals. The conversation
pieces make it clear that the player’s actions have large consequences on the
NPC citizens of this world. The player can help the people they like and
hurt the people they don’t. There is no direct material or economic benefit
to these actions, but goals don’t have to be economic. These interactions
provide another layer of goal for the player to consider.
 This is an interesting set of goals, as the players need to feed their family
conflicts with the core narrative goal of helping out the people coming
through the border crossing. Analyzing and understanding their needs
sometimes takes time. And some later scenarios sometimes require
unsuccessful stamping in order to perform the morally correct action. This
creates a nice tension where the player has to consider the moral
implications to their family versus the moral implications to strangers. This
friction is what makes this an interesting game – too many games that claim
to be about morality don’t make the moral questions difficult. Morality in
games is only interesting if the player has to struggle a little with the
choices.
Finally, as the game goes on for a few in-game days, the player starts to
become aware of the larger story. There are narrative threads that drive the
player to larger goals. As with the smaller narrative goals, these often
require the player to toss out previously important goal structures in order
to succeed. The core gameplay loop rewards the player for taking part in the
system. The narrative encourages the player to ask if the system should
even exist.

Challenges
The challenges of Papers, Please are different than many other games.
There are no enemies. Nothing to jump on or destroy. The player is
presented with a series of people trying to get through a border crossing.
Each person presents a self-contained little puzzle for the player to solve –
is their paperwork correct? Is anything showing the wrong date or wrong
gender or wrong haircut? This is a different type of challenge than jumping
from platform to platform, in that this requires attention to detail and rapid
visual processing of information. It stretches the same muscles as a hidden
object game or those old “find six problems in this picture” puzzles in
newspapers.
The game keeps this simple premise fresh by changing things every day.
There’s always a new form or new restriction that the player has to
consider. And each new piece of bureaucracy is layered on top of all the
previous ones. So, the player is always pushed a little harder to stretch their
memory and attention. The difficulty curve keeps going up.
These core gameplay challenges aren’t the only challenges, though. The
game also offers moral challenges through the narrative framing each
petitioning person offers. Do you help the husband and wife who appear to
have a simple typo on their form? Do you deny the evil jerk who endangers
others even though his paperwork is all in order? Even if you successfully
complete the cognitive challenge, the game adds another interesting
decision on top of that.
The flow diagram for Papers, Please can start off fairly simple, as seen
in Figure 9.2.

Figure 9.2 Papers, Please flow: Simple diagram of the Actions,

Challenges, and Goals of Papers, Please.

But there’s really more depth to the goals than this captures (Figure 9.3).
Everyone wants to get to the end of the day, but some players are motivated
by the narrative goal of feeding their family while others are motivated by
the moral goal of doing the right thing. The game doesn’t really
differentiate – there’s not a different choice or outcome for those two
players, but they will end up having very different experiences. This is a
great example of why it’s important to not just analyze the systems and
rules of the game but to consider how the player experiences them. Papers,
Please is a fairly simple game in its features and rules but uses those simple
combinations to lead to complex results in the player’s perception.

Figure 9.3 Papers, Please flow: Same diagram, but with

expanded goals.

Systems
Papers, Please has relatively simple core gameplay but a number of
supporting systems. This is a good example of how systems depend on your
perspective. Let’s look at it in detail first.
A close examination of the core gameplay involves a few different
systems (Figure 9.4):

The Book: The player has a book that contains the current rules and
some supporting information. The player has to understand the rules in
order to perform their actions successfully.
The Documents: The player needs to analyze the documents given to
them by each person. This is where most of the gameplay actually
happens.
The Interface: The player needs to know how to manipulate things on
the screen, including stamping passports, riffling through documents,
and flipping through the book. The game incentivizes optimizing
within this system, and players often develop heuristics such as placing
things in certain locations and navigating the book via the table of
contents for quick flipping.
The Person: There is a 2D representation of the person that is part of
some forms of analysis, to confirm that this is the right person.
The Chat: The person chats with the player, providing more
information that needs review. The player can prompt this with the
speaker to gather more information, which is encouraged when there is
a discrepancy. The player can also pull up a transcript of the full text to
double-check things and refer to them.
The Clock: The UI includes a display that gives the current time and
date, which is cross-referenced for some documents. This is simple and
static information, so this might be considered more of a piece of data
rather than a full system.
 Figure 9.4 Grouping verbs.

These systems are all elements of the core gameplay loop and inform the
actions the player takes. As noted, some of these might be considered data
points or specific features rather than full systems. When taking a close
look at gameplay, this line can get blurry. And at some level, a strict
definition doesn’t really matter here. The important thing for our current
goal is to note all the factors that affect the player’s actions. A closer
examination of each of those factors will reveal differences and
relationships between different factors. Sometimes labeling these things
into different buckets is a useful way to understand what’s happening.
Sometimes it’s not. It depends on the goals of the investigation.

If the goal is to help with a new game concept with some similar ideas,
then the focus is really on player choices. That suggests breaking
things down by a strongly player-based perspective – what does the
player see and where can they change it? This is basically the
perspective taken in the example up above.
 If the goal is to understand the narrative and how the story is
constructed, then the focus becomes the unique qualities of each
person that appears. Breaking out these differences is more relevant
and deserves more expansion and thought. Thinking about every factor
in the player’s decision is less important and things like economics can
be hand-waved away.
If the goal is to understand difficulty and progression, then the daily
rules changes need more analysis. It might be worth considering each
new rule as a node in the system and see how they layer on each other.

All of this is focused on the core gameplay loop of Papers, Please. This is
just one part of the overall game. There are bigger systems that provide
another possible avenue for dissection.

Gameplay. Perspectives on papers to process people.

Economy. Based on the player’s actions in gameplay, they receive
money. In the daily post-gameplay UI, the player has to allocate this
money to satisfy the needs of their family. This creates a separate
interaction that puts the player’s gameplay choices in context.
Progression. Each day has different gameplay. New rules, new
characters, new difficulty.
Narrative. There is a story that evolves across the progression. Story
events generally occur in a simple pre-set linear progression, but the
story can branch and change based on player choices. The story has
many possible endings depending on these branches.

The Economy is important to Gameplay because it drives the player’s

longer-term goals and incentivizes rapid processing of people for a strong
narrative reward (your family lives). The Progression is important to the
Narrative as they generally flow along together, with the narrative stakes
increasing as the gameplay challenge increases (Figure 9.5).

Figure 9.5 Papers, Please system flow with systems instead of

challenge, showing connections.

Papers, Please works well because each NPC that comes along forces
the player to make interesting decisions about each of the systems (Figure
9.6).
 Figure 9.6 Papers, Please system diagram plus an NPC’s face.

The player needs to evaluate purely along gameplay lines – what are
the facts of the case and how do they line up with the bureaucracy?
Each choice has an economic impact on the player, and the nature of
this impact encourages fast play.
The narrative and progression are closely linked, and determine the
flow of difficulty for the gameplay.
Some NPC choices have larger impacts on narrative, which can then
affect long-term progression.
Some NPC choices don’t have large story or narrative impacts but still
present interesting moral choices.

So even in a relatively focused and clear game like Papers, Please, there
are many ways to dissect. Do you focus purely on the core loop? Do you
focus on narrative or economy or decisions or the player or something else?
These are all valid dissections. It really does just depend on your goals and
needs. Even as we near the end of the book, there is no one answer to how
to do this. You need to understand your own goals and create your own
personal dissection.

Papers, Please – Non-Gameplay Dissection

History
Papers, Please came out in 2013. It was created by Lucas Pope as a solo
developer. Papers, Please is a very successful example of an indie game.
“Indie” is not a precise term but generally applies to smaller games built by
smaller teams with more focus on artistic expression than on commercial
success. Papers, Please had moderate commercial success, but nothing on
the scale of the large AAA games that sell millions and millions of copies.
Papers, Please was a huge critical success, winning multiple major awards
and quickly becoming a standard example of what artistic games can
accomplish.
Indie games as a separate defined category didn’t really exist until the
conditions for small successful games existed.

The rise of multiple digital distribution channels. When the only way
to sell a game is by getting a box into a big retail store, it’s hard to
make a living by making artistic games with a smaller audience. Most
indie game studios don’t have a marketing team.
The rise of game engines like Unity or Unreal made the task of
programming much easier, allowing artists to focus more on
expressing an idea and less on getting pixels to animate.
The growth of the game industry and the expansion of gaming into
more audiences opened up indie games to appeal to people who don’t
play first-person shooters or whatever else is popular these days.
People who like personal quirky games have always existed, but it
took a certain critical mass to sell more than a few artistic games.

Papers, Please came out after this Indie revolution had already become a
thing. Games like Braid and World of Goo and Super Meat Boy (2010) had
shown that smaller games could succeed and could explore innovative
creative ideas that bigger commercial releases generally avoided. Papers,
Please is a great example of how this can be done.

Senses
Papers, Please has a strong and pronounced visual style that permeates
everything in the game. Colors are gray and faded. Fonts evoke Eastern
European themes. The player’s view of everything is a little detached and
distant and separated by glass or other barriers. The player’s avatar isn’t
immersed in the world even when the player’s UI is.

Emotion
Papers, Please is a game that works hard on the player’s heartstrings. The
player doesn’t just look at paperwork; they talk to the people. At first, this is
just adding a little spice – someone will complain about all the paperwork
in a way that resonates with the player. An NPC will be nervous,
emphasizing the stakes for their virtual life. But later NPC interactions
build complex and deep moral dilemmas out of a few lines of text.
How does this work? Papers, Please has a strong emotional foundation.
Players are familiar with this sort of interaction, and understand the tensions
and frustrations and relief that people are likely to experience in this
situation. Normally, players will have seen this from the other side, so
thinking about how it makes the border guard feel might be a new
experience. But the game’s ability to put the player in those shoes is one of
its greatest strengths (Figure 9.7).
 Figure 9.7 Screenshot of the character’s family, as presented in
the daily results screen.

Similarly, bureaucracy and paperwork are things people have a

familiarity with and a strong emotional connection to. Ask people about
their last trip to the DMV and many people will have opinions. This is a
game about working at the worst possible DMV at the worst possible time.
Games that resonate with existing feelings have an easier time building
emotional strength.
Papers, Please also builds emotion out of repetition by establishing
baseline emotions. All of the encounters the player has are in the same
context, so once an emotion like fear or frustration is established by an
NPC, it permeates to all future NPC interactions. The game doesn’t have to
emphasize fear and confusion every time, as that’s already been established.

Money
How does Papers, Please make money?
Papers, Please is generally sold as a complete game in a digital
download. The rise of digital download is one of the factors that allows the
indie scene to exist. Steam is a much easier (and cheaper) platform to
manage than consoles or even PC games in boxes. By cutting out the
middleman and skipping the store experiences, indie developers can get
smaller games to customers in ways that never existed before.
Papers, Please did release on tablet and mobile devices, but not until
many years later.
Because Papers, Please is meant to be downloaded as a complete game,
the game can focus on telling a single complete story. Because players
expect to get hours of gameplay out of a complete game like this, the game
can take time and build on themes without worrying about players leaving
at the drop of a hat. Players are invested, literally. That wouldn’t work in an
arcade or even in a world of free games.

Meaning
Papers, Please is different from many of the games discussed in this book
in that it clearly has a meaning that it wants the player to experience. The
game is about bureaucracy and empathy and the challenges of working
within a flawed system.
Let’s talk about that more.

Games Are about Something

Papers, Please is a game that is about something. Every game is about
something, even if it’s not trying to be, but Papers, Please is trying to say
something. Just by playing Papers, Please, you can see that there is a
driving vision that guides many of the decisions about how this game was
made. Everything – the core gameplay, the visuals, the story, the long-term
resource goals, the music, the little character moments – is all about
creating this empathic connection between the player and a fictional
bureaucrat in a fictional 1980s Cold War dystopia.

How Does Papers, Please Do This?

Papers, Please Works because There Is a Message That Makes
Sense
Papers, Please has a strong setting that supports its vision. A 1980s-era
Soviet bloc country is not a common location for a game, but it’s familiar to
many people. Some people lived through it, while others have seen it in
history books or the popular media of its time. The choice of setting sets the
tone for the rest of the vision – dark, bureaucratic, soul-crushing, but with
moments for individuals to shine.
Most people these days don’t live in a 1980s Soviet bloc country, but the
themes that the setting evoke resonate with modern problems. Random
pointless bureaucracy that makes even simple tasks needlessly complex.
Soul-crushing jobs that ask you to deny both your basic humanity and the
basic humanity of others. A societal system that asks everyday people to
turn against each other in order to keep their families safe and fed. These
are all disturbingly familiar topics for many people. These topics can be
seen in other media (the movie Parasite, Terry Gilliam’s dystopia Brazil,
the show “Severance”, etc.).
Papers, Please takes the themes that are found in this setting and stays
true to them throughout the game. The message remains the same whether
the player is immersed in gameplay, setting up in the UI, watching a bit of
story, or plotting their long-term plan. Every piece of the game aligns with
the same bureaucracy and hopelessness and personal tragedy. Different
moments emphasize different aspects of the theme:

Music establishes the setting.

Resource management is (strangely) more personal in how it related to
the family needs.
Narrative moments with individual petitioners provide discrete moral
choices.

But all the moments tie together into a coherent whole that builds on the
focus of each one.
Papers, Please Works because There Is Consistent Attention to
Detail
Papers, Please is an endless stream of well-crafted little touches that
emphasize its vision. The theme is never forgotten and used to enhance
even small moments.

Music
Color
Font
Character
Dialog
Story Moments
Endings
Short-Term Goals
Long-Term Goals
Player Actions
Rule Structures

Papers, Please Works because It Has Sub-themes

Papers, Please is about bureaucracy. But it’s also about managing and
surviving in desperate times. And it’s also about the little guy stepping up.
And it’s about dealing with a terrible job because you need the money. And
it’s about family. And fascism. And a number of other modern social issues.
These themes all weave together nicely and play off each other.
A theme shouldn’t be a monolithic single statement. Part of the value of a
theme is that it can be explored in different ways from slightly different
angles.
Papers, Please Works because Its Theme Evokes Emotion
A good theme ties into the emotions of the game. As discussed above,
Papers, Please has a strong emotional foundation. Because the themes
resonate with people, there are clear emotional responses that players can
easily tap into. Everyone knows what it’s like to wait in a long pointless line
only to be turned away for some stupid meaningless reason. That has a
certain emotional feel to it that players can easily summon up in
themselves.
Papers, Please doesn’t present the player with much of a window into
the world of Arstotzka. The player doesn’t see anything outside of their
border station. The player never even sees the family that provides the
emotional bedrock of the game. But the game does show the player the
NPCs. The player sees their faces. The player hears what they say, whether
it’s important exposition or nervous fumblings. The game chooses to focus
the player’s attention on the NPCs so that the player builds rapport with
them. Then when the bureaucracy tries to crush them, the player feels
something.
Papers, Please Works because the Message Is Well-suited to the
Medium
Games are about choices. Games give the player a chance to experience
something not as an observer, but as a participant. Games can immerse the
player in a role and then prompt the player to see how they would respond
in that role. This is a great tool for empathy. Watching someone experience
a problem or challenge in a movie or television program provides a
valuable perspective on that problem. But playing a game about the
problem forces the player to consider the problem on their own, inserting
their own values into the perspective that the game provides.
 Papers, Please does this well. The setting is clearly established from the
moment the game starts. The player is given a clear role with a personal
perspective. The player is given a simple repetitive task that fits
expectations from other types of games. The structure of games primes the
player to accept arbitrary goals and rewards as natural. Bureaucracy is a
system, and games are systems, so it’s easy to present the ridiculous rules of
bureaucracy as just another part of the game. Players are used to be told to
abide by various arbitrary rules in games, so it’s a natural fit here.
But as the game progresses, the player sees how these systems are set up
for him to fail. The player is going to have to make difficult personal
choices, both in the resource management side of things and also in the
narrative side of things. The game sets up a simple structure, then
undermines it through narrative moments. The player knows that the game
is telling them to do one thing (process people as quickly and accurately as
possible) but the moral dilemmas presented through the petitioners pull the
player in another way (are you going to separate the husband and wife over
a minor infraction?). Papers, Please does a great job of balancing these two
needs – the player is given clear benefits and costs and has to weigh them
based on traditional game values but also has to decide personally how they
want to weigh larger moral values. The moral choices are clear and
interesting without being preachy. Many games fall into the trap of using
moral choices to make the player feel praised, but Papers, Please instead
provides actual moral challenges that are much more interesting.
Watching a movie about a bureaucrat who has to choose between feeding
his family and helping strangers would be compelling and interesting (see:
To Kill a Mockingbird). But having the player make that decision forces an
even deeper understanding. Games are about interesting decisions. By
making the moral decisions interesting, it forces the player to evaluate those
moral decisions in more detail. The player has to pause and think about how
much they value helping a certain stranger, and quantify that decision. Is it
worth saving someone’s life if it means my child might have to go without
medicine and thus might die? What if there’s enough ambiguity that I can
look the other way and hope that they won’t actually die? What if the
stranger whose life I can save is kind of a jerk about it? How much does
that change my calculation?
Papers, Please can be seen as a long string of interesting moral
dilemmas. With the added depth provided by a connective resource
management game and associated narrative. Not every game goes this far
with their moral themes, but many games dip their toe in this water. Many
games force the player to decide how to balance moral narrative choices
against clearly defined resource gains. This is what draws many players to
games like Life is Strange, Bioshock, or even Civilization. There’s a
philosophy book I like called Games: Agency as Art that discusses this –
using player decisions as the canvas to create artistic expression is
something interesting that artists haven’t really focused on before. But in a
game like Papers, Please it can lead to fascinating revelations.

How Can You Do This?

Know Your Themes
Making a game with a theme only works if you know the theme. Themes
come together from many places and take some time to pin down. When
you’re first thinking about what games you want to make, you may not be
able to articulate why a certain combination of rules and backstory and
player actions appeal to you. It takes some time and introspection to
connect the dots that make up a theme. The reason these things “feel right”
together is often because they say the same thing about the world. Look for
that and embrace it.
Once you start to connect the dots, it may take a little massaging to turn
“feels right” into a coherent theme. The act of defining the theme often
requires some changes to make the direction work for the game as a whole.

A theme like “choices” is very broad and applies to most games – it

may need to be refined to add more detail and specificity, turning it
into something like “even little choices matter” or “the obvious
choices aren’t always the right ones” or “even when you think you
have a choice, you might not”.
A theme like “love” may be great for the story you want to tell but has
nothing to do with the gameplay you want. Is Super Mario Bros about
the love between a plumber and a princess? Maybe, but it’s not
something the player ever has to think about. You can make a game
where the story and the gameplay are focused on different things, but
it’s never going to be as strong as one where they align. Clint Hocking
even came up with a name for this – ludonarrative dissonance –
because it happens so often (Hocking 2007). If you find yourself
starting down this road, consider ways to align the two. Usually
changing story is easier than changing gameplay, but sometimes
exciting new gameplay ideas come from trying to fit to a story theme.
Games that are actually about love are rare but interesting (see:
Florence (2018) or Consentacle (2014)).
A theme like “it’s fun to be powerful” may fit your gameplay but not
be very interesting or unique. What can you do to tweak this into
something new? “Protecting others makes me feel powerful” could
lead to interesting gameplay and story elements. Or play with other
directions – “power corrupts, absolute power corrupts absolutely” or
 “there’s always a bigger fish”. Marvel comics became the world leader
in superheroes by tweaking the traditional power fantasy into “With
great power there must also come – great responsibility” (Lee & Ditko,
1962).

Once you identify and massage your theme, it’s worth spending a little time
to find the right phrasing. A good theme is a key part of a game’s vision,
and as such it’s worth communication throughout the production process
and in the marketing. Writing it up in a short pithy easy-to-remember phrase
makes it easier to transmit and discuss and share.

Theme Everything
Once you identify and name your theme, make it a lens that you consider
with everything you do. This can be applied in all sorts of cases to make
decision-making easier. If you’re debating between two UI color palettes,
think of your theme. If you’re writing up notes to the composer who is
going to create background music for you, tell them about your theme. If
you’re not sure which feature to prioritize, think about how they support
your theme. A theme won’t make all your decisions for you (what color
palette says “even little choices matter”?), but it will help.
This becomes more important the more people working on the game. A
clear well-phrased theme can be easily shared with the whole team. If
everyone knows the theme, everyone can use the theme. This is especially
valuable on larger teams where team members are not all in constant
communication. Having a unified theme is hard on these teams. Every team
member makes choices every day about how to build things or tweak
things. Having a shared understanding of the theme helps those decisions
line up to point at the same goal. Sure, theme is going to have the most
impact on the big fundamental decisions – core gameplay loop, story, visual
style – but if you want every little thing to align in the same way then every
person working on the project needs to be thinking about the same things.
The person making the street signs that the player will walk past while on
their quest will make better street signs that better fit your game if they
understand the theme.
This may sound arduous and oppressive and restrictive, but it’s actually
the opposite. The random junior sign-maker may want to put their creative
stamp on their work, but it’s actually easier to do that when they know that
their creative stamp aligns with everyone else’s. Having a shared vision is a
great and powerful thing. This is what bonds teams together and makes
them stronger. When a team aligns and agrees on a theme, it becomes a goal
that drives them together in shared and glorious purpose.
Vision
What if you don’t want to make a statement?
What if you just want to make a fun game and maybe make some
money?
There’s nothing wrong with that. Games can be used for many things.
Not all games are about getting a message across.
Even if your game is not about getting a message across, you’re still
going to be getting a message across. Every game is about something.
Every game says something about the world. If you’re not thinking about it
when you start work on the game, it’s still there but you have no control
over it. I don’t think the creators of Pac-Man were trying to make a perfect
summary of the 1980s, but they chose to include a core gameplay loop of
endless consumption, overvalued power fantasies, and bright neon colors. It
bubbled up even without being intended.
The same principles that apply to a game’s message can also be applied
to other qualities of the game. In addition to thinking about what the game
is trying to say, it’s good to start off thinking about other “why” questions
about the game:

Why do these features fit together?

Why would someone want to play this?
Why do I want to make this?
Why is this game great?

Understanding why this game is interesting and deserves to exist is an

important first step in making any game. This book starts off saying that the
player’s actions are the first thing you need to understand a game. And this
is true. But when you’re making a game, it’s often better to start off with an
understanding of why you want to make it.
Similar to theme, it’s good to crystalize these why thoughts into
something that can be easily shared amongst the team. A team with a shared
vision of why their game exists is going to produce a much better final
product than a team with multiple or conflicting visions. Writing this down
and discussing it early allows for adjustments and corrections before a lot of
time and money are invested. Defining a shared vision for the game is
worth the effort.
There are many ways to communicate vision. Find the methods that work
for the needs of your game.

Vision Statements (aka Pillars). Write out a carefully worded statement

that guides development. Keep it short and memorable so everyone on
the team knows it and repeats it.
The Where’s My Water? team often said that our goal was to
“Make the player feel smart”
 Vision Points. Write out a few Vision Statements to emphasize
multiple important why statements. Prioritize them so everyone on the
team can make meaningful decisions based on them.
We did this consciously on games like Hunter: the Reckoning,
and it helped. For that game, our specific list was something like:
It’s fun to kill things
Enemies show off player abilities
The world is dark and Gothic
Mood Boards. Select a few different examples from existing media to
show what you want the game to look and feel like.
This is often better for defining visual style than for gameplay but
can be used for all sorts of things.
Prototypes. Build out the things that you think are important and then
anyone who plays that sample has something to point to.

A clear vision is great for the game as a whole but can also work for
individual features and content. Knowing why each piece of the game is
there helps everyone align on how to implement it.

Why can the character spin around?

The spin move allows for 360 attacks, which are useful to clear
out large numbers of enemies and give the player a quick breather
in big fights.
Spinning is also used during long traversal areas. Breaking small
objects for small rewards gives the player something to do when
getting from point A to point B.
Why should a player care about this enemy?
The big dude with a shield is here to guide the player to use their
heavy attack instead of just spamming the light attack.
 Why is this level in the game?
This level introduces the player’s gliding ability and gives them
lots of ways to play around with it and learn how to use it.
This level is also where the character first meets an annoying rival
who later turns out to be their long-lost sibling, so that character
should be a keynote throughout the visual design of the level.

This sort of information is often as important as functional specifications in

design documentation. When a programmer is implementing small scenery
objects in the game, they may need to make decisions. Knowing how those
objects fit into the game helps guide those decisions. Even if the design
documents already say that they need to break and sometimes have rewards
in them, the specific details about how to do that are different when you
want them as small traversal rewards versus mid-boss-fight health
replenishes or post-battle randomized rewards. Knowing the why for each
piece of functionality helps understand the what and the how.

Games Matter
Papers, Please is an excellent example of how games can matter. Games
can have something to say. Games can change how people view the world
and think about problems. I know I have a little bit more empathy for every
airport security guard I meet these days because of the game.
I work with the Education team at Roblox (but, as previously mentioned,
I don’t speak for them or the larger Roblox corporate entity). We’re trying
to use the capabilities of the Roblox platform to build new exciting
interactive tools for teachers. I believe that the future of education is going
to be shaped by games. Game developers have spent a long time teaching
people how to do silly things like jump on turtles and help alligators take
showers. But those same principles can be used to teach people important
things about the real world. It’s not easy, but it can be done.
When you’re making games, think about how to make them fun. Think
about how to keep people engaged and excited. But also think about what
you’re doing to the world. Can you make the world a better place, even in
small ways? Adding joy to people’s lives is a good. I’m proud of the games
I’ve made that are simple escapist fun. But games that add more than just
joy are even better.
OceanofPDF.com
Chapter 10
Post Mortem
DOI: 10.1201/9781003346586-10

1. Every game is unique.

When trying to understand games, each game requires its own separate
analysis. The same tools can be used, but they need to be recalibrated and
re-examined. A feature that works in one game won’t necessarily work for
another. A good dissection will help you see why that is the case. Games are
complicated systems, so a small change in one place can have large
unintended ripples throughout the rest of the design.

2. Every game is the same.

Every game shares many of the same elements. Every game has things the
player can do. Without interactivity, it’s not a game. Every game has things
the player wants. Without goals, it’s a toy. Every game has challenges.
Without obstacles, a game is no fun. Identifying this shared connective
tissue is the first step toward understanding games. Further examination of
these factors usually leads to interesting insights. Does the game have lots
of multifaceted goals but one simple action the player repeats? Does the
game give the player lots of tools but keep them focused on the same goal
throughout? Or does the game present lots of everything for the player to
explore? There’s no right or wrong way to set these things up.
 3. Know your goal when dissecting.

There are lots of ways to understand games. There are lots of different ways
to think about how to break down a game, and how to view the parts. If you
try to have a complete unlimited understanding of a game, you will never
reach that goal. When you start your dissection, understand what you’re
trying to get out of the experience. Are you building something similar?
Trying to understand the genre? Want to know why one game succeeded
and another failed? Just trying to get at the psychology of the players? Just
want to understand why you like it? There are many reasons to dissect, and
they’re all valid. But they require different tools and lead to different
results.

4. Games are systems.

I love systems. Systems are interesting and complicated and fun to look at.
Dynamics are a great example of this – combining simple rules to create
something complex. Game design requires not only understanding how
each feature works but also understanding how each feature relates to each
other. As mentioned, this quickly gets infinitely complicated. That’s why no
game designer fully 100% understands even their own games.

5. Game development is iteration.

Game development is done iteratively because no one can predict every

consequence of every decision in a complex system. A good game designer
can develop some understanding and some rules or even just intuition about
which changes are better than others. But there’s no better way to
understand a game than to play it.
 6. Systems are about relationships.

The best games are the games that have tightly connected systems. Players
have interesting decisions to make because every decision affects so many
things across the game. Each action the player takes ripples across multiple
systems, changing the actions, challenges, and goals that the player
experiences next. Building relationships that work this way is hard, but
that’s the heart of game design.

7. Always think of the player.

Game design is about creating a player experience. Every game design

decision needs to be evaluated in the context of how it affects the player. If
you have a great game design idea but can’t translate that into something
that delights the player, it’s not worth adding to your game. The first test of
any game design idea is “how does this make the player’s experience
better?”.
Game design is hard. Games are complex systems that are impossible to
predict. But it gets worse. In addition to designing interesting complex
systems that relate to every other system, game designers need to make
each of those systems and relationships make sense to the player. The
player needs to be able to jump in and experience the game with only the
information the game gives them. And still have fun every moment of that
experience. Every game needs to go through an extensive teaching process
to get the player up to speed on their abilities and goals and obstacles. And
if that learning process isn’t fun for every moment of the player’s
experience, that player is gone.

8. Game design decisions are often due to non-gameplay factors.

Games are more than just the gameplay. Games are visuals. Games are
sounds. Games are part of a larger history. Games are stories. Games are
products. The team developing the game may not be focusing on each of
these things, but they are affecting how decisions are made about the game.
Video games are constrained by technology and how many polygons can be
displayed at one time. Video games are equally constrained by the games
that came before them. Genres exist because people build on what worked
before.
Most games are commercial products. Most development teams want to
make something fun that players will enjoy, but they also want to pay their
bills. Monetization features are added to just about every game, and this
changes the game designs. This has always been true. The classic arcade
games that are honored today are designed to suck quarters from players.
Classic console and PC games are designed to sell boxes and make players
so excited that they tell their friends to buy new hardware to play. Modern
games are designed for streaming and social media. Understanding these
influences is critical to understanding why those games are the way they
are.

9. Games have meaning.

Some games are clearly “About Something”. Papers, Please wants you to
think about bureaucracy and empathy and spend some time viewing the
world from someone else’s perspective. Papers, Please was designed to
emphasize this meaning through every decision – gameplay, visuals, story,
progression, etc.
Many games are not designed with a clear goal like this. Candy Crush
Saga did not start with the idea of “Unexpected Consequences” and build
from there. I don’t think the creators of Centipede sat down to have
discussions about how to bring out the “frantic nature of the times” in their
visuals, story, or progression. But regardless of the creators’ intent, these
things are present in the games.
Games are art. Art has meaning. It’s not always the meaning the author
intended. The meaning may change over time. But just like a great book or
movie or painting, games say something about the world and the human
condition. When you finish playing a game, you are a changed person. Your
perspective on the world has changed. At least a little.

10. Everyone should analyze games.

Everyone plays games. Self-proclaimed “gamers” make it a key part of

their identity, but they’re not the only ones playing games. Plenty of self-
proclaimed “non-gamers” spend an hour or two every few days on Candy
Crush Saga or similar games. Lots of people pull out a board game when
family or friends come over to visit. Almost everyone played a variety of
board, card, sport, playground, or other games growing up. Games are an
integral part of the human condition.
Many parts of the book are intended for game designers, game
developers, and people who view games as part of their identity. But just
like any art, critical analysis of the things that influence us as humans is
inherently valuable. I took a wide range of classes in college on reading
works of literature or analyzing pieces of art. I believe those experiences
made me a better person. Or at least a more thoughtful person. I look at
books and movies and paintings differently now than I did before those
classes. I like to think that I get more out of them now – more pleasure or
value or depth (whatever those words might mean). The same truth can
apply to games. Everyone plays games, at least a little. Understanding how
they work and how they affect us humans can only enrich our lives.
Games are fun.
Games are interesting.
Pay attention to how they work.

OceanofPDF.com
Bibliography – Games
Adopt Me!, (2017) [Roblox game] Bethink, Uplift Games
Agricola, (2007) [Board game] Uwe Rosenberg, Lookout Games
Angry Birds, (2009) [iOS, Nokia N900 game] Rovio, Chillingo
Animal Crossing, (2001) [GameCube game] Katsuya Eguchi, Hisashi
Nogami, Nintendo
Axis & Allies, (1984) [Board game] Larry Harris, Milton Bradley
Battle Cards, (1993) [Card game] Merlin Publishing
Bejeweled Blitz, (2010) [Facebook game] Jason Kapalka, PopCap Games
Bejeweled, (2001) [PC game] Jason Kapalka, PopCap Games, Electronic
Arts
Bioshock, (2007) [PC, Xbox 360 game] Ken Levine, 2K Boston
BIRD, (2019) [Roblox game] Roblox BIRD Community
Braid, (2008) [Xbox 360 game] Jonathan Blow, Number None, Microsoft
Game Studios
Brookhaven RP, (2020) [Roblox game] Wolfpaq, Roblox
Bubble Witch Saga, (2012) [iOS, Android game] Sebastian Knutsson, King
Candy Crush Saga, (2012) [Facebook, iOS, Android, Windows Phone
game] King/Microsoft
Candy Land, (1949) [Board game] Elanor Abbott, Milton Bradley
Carcassonne, (2000) [Board game] Klaus-Jürgen Wrede, Hans im Glück
Castle Crashers, (2008) [Xbox Arcade game] Dan Paladin, The Behemoth,
Microsoft Game Studios
Castlevania, (1986) [Famicom game] Hitoshi Akamatsu, Konami
Catan, (1995) [Board game] Klaus Teuber, Kosmos
Cave Story, (2004) [PC game] Daisuke Amaya, Studio Pixel
Centipede Recharged, (2021). [PlayStation5, Nintendo Switch, Xbox One,
PlayStation4, PC, Mac, Atari VCS game] Adamvision Studios,
SneakyBox, Atari
Centipede, (1981). [Arcade game] Dona Bailey, Ed Logg, Atari
Commander Keen in Invasion of the Vorticons, (1990) [PC game] Tom Hall,
Ideas from the Deep, Apogee Software
Consentacle, (2014) [board game] Naomi Clark, (self-published)
Cosmic Encounter, (1977) [Board game] Eon Games
Crazy Cake Swap, (2016) [iOS, Android game] Tom Smith, Zindagi Studio,
Zynga
Diablo, (1997) [PC game] Blizzard Entertainment
Dominion, (2008) [Board game] Donald X. Vaccarino, Rio Grande Games
Duelmasters, (2002) [Card game] Takara Tomy, Wizards of the Coast
Dungeons & Dragons, (1974) [Table-top role-playing game] Gary Gygax,
Dave Arneson, TSR
El Grande, (1996) [Board game] Wolfgang Kramer, Richard Ulrich, Hans
im Glück
Elder Scrolls V: Skyrim, (2011) [PC, PlayStation 3, Xbox 360 game] Todd
Howard, Bethesda Game Studios, Bethesda Softworks
Farmville, (2009) [Facebook game] Zynga
Flight Control, (2009) [iOS game] Firemint, Namco
Florence, (2018) [mobile game] Ken Wong, Mountains, Annapurna
Interactive
Fluxx, (1996) [Card game] Andrew Looney, Looney Labs
FTL: Faster than Light, (2012) [PC, Mac, Linux game] Subset Games
God of War, (2005) [PlayStation 2 game] David Jaffe, Sony Santa Monica
Studio, Sony Computer Entertainment
Grand Theft Auto 3, (2001) [PlayStation2 game] DMA Design, Rockstar
Games
Grand Theft Auto: Chinatown Wars, (2010) [iOS port of a DS/PSP game
from 2009] Rockstar Games
Hades, (2020) [PC, Mac, Nintendo Switch game] Supergiant Games
Halo: Combat Evolved (2001) [Xbox game] Jason Jones, Bungie, Microsoft
Game Studios
Hearthstone, (2014) [PC, Mac, iOS, Android game] Blizzard Entertainment
Hunter: the Reckoning (2002) [Xbox game] Dave Rodriguez, High Voltage
Software, Interplay Entertainment
Infinity Blade, (2010) [iOS game] Donald Mustard, Chair Entertainment,
Epic Games
Kaijudo (see Duelmasters)
Legend of the Five Rings, (1995) [Card game] Alderac Entertainment Group
Life is Strange, (2017) [PlayStation 3, PlayStation 4, Windows, Xbox 360,
and Xbox One game] Raoul Barbet, Michel Koch, Dontnod
Entertainment, Square Enix
Lode Runner, (1983) [Apple II, Atari 800, VIC-20, Commodore 64, PC
game] Doug Smith, Broderbund
Magic: The Gathering, (1993) [Card game] Richard Garfield, Wizards of
the Coast
Manhattan, (1994) [Board game] Andreas Seyfarth, Hans im Glück
Mario Bros, (1983) [Arcade game] Shigeru Miyamoto, Gunpei Yokoi,
Nintendo
Mario Paint, (1992) [SNES game] Hirofumi Matsuoka, Nintendo
Mario Party, (1998) [Nintendo 64 game] Kenji Kikuchi, Hudson Soft,
Nintendo
Marvel Snap, (2023) [iOS, Android game] Second Dinner, Nuverse
Medici, (1995) [Board game] Reiner Knizia, AMIGO
Metroid, (1986) [NES game] Nintendo
Minecraft, (2009) [PC game] Markus Persson, Mojang Studios
Monopoly, (1935) [Board game] Lizzie Magie, Waddingtons, Parker
Brothers
Munchkin, (2001) [Card game] Steve Jackson, Steve Jackson Games
Museum of Mechanics: Lockpicking, (2022) [PC game] Johnnemann
Nordhagen, Dim Bulb Games
Nethack, (1987) [PC game] NetHack DevTeam
Nintendogs, (2005) [Nintendo DS game] Kiyoshi Mizui, Nintendo
Pac-Man, (1980) [Arcade game] Toru Iwatani, Namco
Pandemic, (2008) [Board game] Matt Leacock, Z-Man Games
Papers, Please, (2013) [PC, Mac game] Lucas Pope, 3909 LLC
Pet Simulator 99, (2023) [Roblox game] Preston “BuildIntoGames”
Arsement, Big Games
Pocket God, (2009) [iOS game] Bolt Creative
RoboRally, (1994) [Board game] Richard Garfield, Wizards of the Coast
Rogue, (1980) [Unix game] Michel Toy, Glenn Wichman, Ken Arnold
Roller Coaster Tycoon, (1999) [PC game] Chris Sawyer, Hasbro Interactive
Royale High, (2017) [Roblox game] callmehbob, Roblox
Scrabble, (1948) [Board game] Alfred Mosher Butts, Hasbro
Settlers of Catan (see Catan)
Shariki, (1994) [PC game] Eugene “Zhenya” Alemzhin
Shovel Knight: Shovel of Hope, (2014). [3DS, WiiU, PC, Mac game] Sean
Velasco, Yacht Club Games
Sid Meier’s Civilization, (1991) [PC game] Sid Meier, MicroProse
SimCity, (1989) [Amiga, Mac, PC, Commodore 64 game] Will Wright,
Maxis, Electronic Arts
Slay the Spire, (2019) [PC, Mac, Linux game] Clark Aboud, Mega Crit,
Humble Bundle
Space Panic, (1980) [Arcade game] Universal Coleco
Spelunky, (2012) [Xbox 360 game] Derek Yu, Mossmouth, LLC, Microsoft
Studios
Stardew Valley, (2016) [PC game] Barone, Eric
Street Fighter 2, (1991) [Arcade game] Yoshiki Okamoto, Capcom
Super Mario Bros, (1985) [NES game] Shigeru Miyamoto, Nintendo
Super Meat Boy, (2010) [Xbox 360 game] Edmund McMillen, Tommy
Refenes, Team Meat
Sushi Go!, (2013) [Card game] Phil Walker-Harding, Gamewright Games
Tamigochi, (1996) [Physical game] Aki Maita, Bandai
The Elder Scrolls IV: Oblivion, (2006) [PC, Xbox 360, PlayStation 3 game]
Ken Rolston, Bethesda Game Studios, 2K Games
The Game of Life, (1960) [Board game] Bill Markham, Reuben Klamer,
Milton Bradley
The Legend of Zelda: Breath of the Wild, (2017) [Nintendo Switch game]
Hidemaro Fujibayashi, Nintendo
The Secret of Monkey Island, (1990) [Amiga, Atari ST, PC, Mac game] Ron
Gilbert, Lucasfilm Games
The Secret of Monkey Island: Special Edition, (2009) [iOS, PC, Xbox 360
game] Ron Gilbert, Lucasfilm Games
The Sims, (2000) [PC, Mac game] Will Wright, Maxis, Electronic Arts
Ticket to Ride, (2004) [Board game] Alan R. Moon, Days of Wonder
Tigris and Euphrates, (1997) [Board game] Reiner Knizia, Hans im Glück
Tomeko’s Great Adventure: Mystery Dungeon, (1993) [Super Famicom
game] Tadashi Fukuzawa, Chunsoft
Track and Field, (1983) [Arcade game] Konami
Triple Town, (2010) [Amazon Kindle, Facebook game] Dan Cook, Spry
Fox
Ultima 1: The First Age of Darkness, (1981) [Apple II game] Richard
Garriott, Kenneth W. Arnold, Origin Systems, California Pacific
Where’s My Water?, (2011) [iOS, Android game] Tim FitzRandolph,
Creature Feep, Disney Mobile
Wiz War, (1985) [Board game] Tom Jolly, Jolly Games
Wizardry: Proving Grounds of the Mad Overlord, (1981) [Apple II game]
Andrew C. Greenberg, Robert Woodhead, Sir-Tech Software
Work at a Pizza Place, (2008) [Roblox game] Dued1, Roblox
World of Goo, (2008) [PC and Wii game] Kyle Gabler, 2D Boy
World of Warcraft, (2004) [PC, Mac game] Blizzard Entertainment
XCom: Enemy Unknown, (2013) [iOS port of a PC game from 2012] Firaxis
Games, 2k Games

OceanofPDF.com
Bibliography – Non-Game
(2024). (n.d.) Roblox Reports Fourth Quarter and Full Year 2023 Financial
Results. Roblox.com. https://ir.roblox.com/news/news-
details/2024/Roblox-Reports-Fourth-Quarter-and-Full-Year-2023-
Financial-Results/default.aspx
Bartle, R., “Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs”
(1996).
Cabera, P., & Soesbee, R. (1999). The Way of the Naga (p. 2). Alderac
Entertainment Group.
Curry, D. (2024). “Candy Crush Revenue and Usage Statistics.”
https://www.businessofapps.com/data/candy-crush-statistics/
Csikszentmihalyi, M. (1996). Creativity: flow and the psychology of
discovery and invention. HarperCollinsPublishers.
England, L. (21 April 2014). “The Door Problem” – “Liz England”.
lizengland.com
Faidutti, B. (2017). “Postcolonial Catan.” Boardgame Design by Bruno
Faidutti. February 6, 2017.
https://faidutti.com/blog/blog/2017/06/02/postcolonial-catan/
Friedman, T. (1999). “Civilization and Its Discontents: Simulation,
Subjectivity, and Space” in On a Silver Platter, ed. Greg M. Smith, New
York University Press, 1999.
Gamma, E., Helm, R., Johnson, R., & Vlissides, J. (1994). Design Patterns:
Elements of Reusable Object-Oriented Software (1st edition). Pearson
Education, Limited.
Griesemer, J. (2010). “Changing the Time Between Shots for the Sniper
Rifle from 0.5 to 0.7 Seconds for Halo 3” GDC Talk.
Guide, C., & Gary, K. M. (1982). How to win video games.
Hocking, Clint (October 7, 2007). “Ludonarrative Dissonance in
Bioshock”. Click Nothing. Archived from the original on January 14,
2020).
Huizinga, J. (2014). Homo Ludens: A Study of the Play-Element of Culture.
Mansfield Centre, CT: Martino Publishing ISBN 978-1-61427-706-4.
L, D. (2018). Fishing for Gods in Strade’s Gallows. [Downloadable game
module] Dungeon Masters Guild.
Lazzaro, N. (2012). Why We Play: Affect and the Fun of Games—
Designing Emotions for Games, Entertainment Interfaces, and Interactive
Products. In Human Factors and Ergonomics. CRC Press.
https://doi.org/10.1201/b11963-36
Loring-Albright, G. (2015). “The First Nations of Catan: Practices in
Critical Modification” Analog Game Studies 2.
Lee, S. and Ditko, S. (1962). Amazing Fantasy #15, Marvel Comics, NYC.
Meadows, D. (2008). Thinking In Systems: A Primer. Chelsea Green
Publishing. ISBN 978-1844077250.
Meier, S., & Noonan, J. L. (2021). Sid Meier’s memoir!: a life in computer
games. W. W. Norton & Company, Inc.
Miller, G. A. (1956). The magical number seven, plus or minus two: Some
limits on our capacity for processing information. Psychological Review,
63(2), 81–97.
Nguyen, C. Thi (2020). Games: Agency as Art. New York: Oxford
University Press.
Raphel, A. (2017). “The Man Who Built Catan.” The New Yorker,
https://www.newyorker.com/business/currency/the-man-who-built-catan
Rogers, S. (2014). Level up!: the guide to great video game design (2nd
ed.). Wiley.
Rosewater, M. (2002). Timmy, Johnny, and Spike. Wizards of the Coast.
https://magic.wizards.com/en/news/making-magic/timmy-johnny-and-
spike
Rosewater, M. (2009). Kind Acts of Randomness. Wizards of the Coast.
https://magic.wizards.com/en/news/making-magic/kind-acts-
randomness-2009-12-14
Rosewater, M. (2011). New World Order. Wizards of the Coast.
https://magic.wizards.com/en/news/making-magic/new-world-order-
2011-12-05
Rosewater, M. (2014). Lenticular Design. Wizards of the Coast.
https://magic.wizards.com/en/articles/archive/making-magic/lenticular-
design-2014-12-15
Schell, J. (2015). The art of game design: a book of lenses (2nd ed.). Taylor
& Francis.
Shakespeare, W., 1564–1616 author. (1954). The tragedy of Hamlet, Prince
of Denmark. [London]: The Folio Society.
Suits, B. (1978). The grasshopper: games, life, and Utopia. University of
Toronto Press.
Tekinbaş, K. Salen, & Zimmerman, E. (2003). Rules of play: game design
fundamentals. MIT Press.
Winkie, L. (2021). “The Board Games That Ask You to Reenact
Colonialism” The Atlantic
https://www.theatlantic.com/culture/archive/2021/07/board-games-have-
colonialism-problem/619518/
Wolf-Meyer, M. (2019). “Everything I needed to Unlearn I learned from
Sid Meier’s Civilization”
https://matthewwolfmeyer.com/2019/05/22/everything-i-needed-to-
unlearn-i-learned-from-sid-meiers-civilization/
Raphel, A. (2014). “The Man Who Built Catan.” The New Yorker, February
12, 2014. https://www.newyorker.com/business/currency/the-man-who-
built-catan
Wright, W. (2011). The Replay Interviews: Will Wright. Game Developer.
https://www.gamedeveloper.com/business/the-replay-interviews-will-
wright
Yu, D. (2016). Spelunky. Boss Fight Books. ISBN 9781940535111.

OceanofPDF.com
Index
Pages in italics refer to figures.

A
actual audience, 221
Ancient Evil Necromancer, 216
Angry Birds, 57
Animal Crossing, 11, 194, 195, 208
animation, 28, 52, 76, 189
Apple, 56–57, 215
arcade games, 5, 42, 74–75, 87, 89, 92–95, 247
ASCII graphics, 150
avatar
character position, 20–21, 28
normal Roblox, 203, 210
player’s, 25, 94, 101, 159–160, 203, 224, 233

B
Bailey, Dona, 93
Baldur’s Gate, 35
Bartle’s taxonomy of MMO players, 219
baseball, 4
cards, 184
Battle card, 189, 190
being all-powerful is boring, 7, 9
Bejeweled, 51
Bioshock, 200, 239
board games, 2, 5, 33, 49, 85, 107, 111–146, 170, 182–186, 189, 195,
248; see also Catan/Settlers of Catan
rulebooks, 5, 107
Braid, 41, 233
Brookhaven RP, 211
Bubble Witch Saga, 57
build actions, 117, 118, 175–176, 180–181
bureaucracy, 224, 227, 231, 234–238, 247

C
Candy Crush Saga, 221, 225, 248
counting options, 69–70
flow, 68, 68
frame of, 52–53
gameplay dissection
challenge, 50–51, 52
goals, 50, 58–73
verbs, 47–50
inputs, 49
match-3 genre, 50–51, 55, 63
mobile game by King, 47
non-gameplay dissection
emotion and narrative, 54–55
events, 66–68
 meaning, 58
mobile games history, 56–58
money, 55–56
player-generated goals, 65–66
secret verbs, 68–71
senses, 53–54
technology, 52–53
verbs and goals, 71–72
outputs, 49
rules, 49
Saga map, 67
sense of progression, 51, 64
swap/swapping, 49–50
variety of different objects, 50–51
Candy Frogs, 50
Candyland, 4
card games, 2, 69, 135, 144, 172, 181–182, 184–185, 187–188; see also
Magic: The Gathering/Magic
cartoony and childish, 53–54
Castle Crashers, 151
Castlevania, 15, 41
Catan/Settlers of Catan, 2, 12, 177, 205
challenges, 115
core loop, 116–118
goals, 113–115
non-gameplay dissection
history, 118–119
meaning, 122
 monetization, 120
narrative, 121–122
senses, 120–121
not verbs
Roll, 113
Settle, 113
randomness
allows mixed groups to have fun together, 140
benefits, 139
building in Catan, 143
catch-up mechanic, 140–141
development deck, 144–145
downsides, 141
initial board setup, 143
make a game easier, 139
make a game more widely appealing to large audience, 140
overwhelm choices, 142
punishes best players, 141–142
resource production, 143–144
rolling a seven, 145–146
takes away earned rewards, 142
uncertainty, 122–123
benefits of, 135–138
certainty about, 146
downsides, 143
game state, 131–135, 131–134
hidden information, 124–125
mental skill, 124, 126
 people, 123–125
physical skill, 124–126
randomness, 125–126
works, 126–131, 126–129
US edition of, 120
verbs, 111–113
Cave Story, 41
Centipede, 2, 74–110, 248
Centipede Recharged (2021), 86
challenge
fleas, 81
to gardening, 82
scorpions, 81–82
spiders, 81
dynamics
combination of rules, 96–99
complicated by other rules, 96
fleas cause through set of, 96
game designer’s target/goal, 97–98
logical outcome of series of other rules, 95–96
Mushroom System, 103–104
phases, 99–100
priorities, 104
feature ideas, 87–91
flow, 82–84, 83
breakdown, 83
decision process, 83
decision process ++, 84
gardening dynamics
mushrooms interact with centipedes, 77–78
Mushroom Bounce, 78
Mushroom Tunnel, 79
goals
action, 79–80
anti-goal of “Don’t Die”, 79–80
collect points/gold, 76–77
destroy enemies, 77
don’t die, 76
gardening, see gardening dynamics
high-level
actions, 101
challenge, 102–103
goals, 102
non-gameplay dissection
golden age of arcade games, 92–93
meaning, 94–95
money, 93–94
narrative and emotion, 94
senses, 92
Bug Blaster, 74, 88, 94
progression, 92
rules, 84–86
breaking down a feature, 85
to identify changes, 86–91
simple AI rules, 96–97
systems, 100–101
 dissecting, 108–109
mapping non-existent, 109–110
verbs
move, 74–75
shoot, 75–76
visualization
Centipede as flowchart, 106, 106–107
Centipede as rulebook, 107, 107
Venn diagrams, 105, 105–106
visual map, 104
challenge, 122–146
Candy Crush Saga, 50–52
Centipede, 81–82
Magic: The Gathering, 177–180
Papers, Please, 227–228
Royale High, 207–209
Settlers of Catan, 115
Spelunky, 148
chess, 70, 72, 126, 133, 136, 141, 177, 183, 190
Civilization, 4, 63, 200, 239
classic arcade games, 75, 93, 247
Commander Keen (1990), 41
commercialism, 216
console games, 42, 56
constant rewards, 54
content
difficulty curves, 191–195
add peaks, 193, 193
 gradually rise to peaks, 193, 193
Grand Theft Auto, 3, 194–195, 196
levels off after a while, 192, 192
Rollercoaster Model, Shovel Knight, 193, 194
Stardew Valley/Animal Crossing, 194, 195
straight line, 192
flow, 201
fun, 196–200
good ingredients, 201
good-level design and progression, 199
knobs, 200
Magic cards, 189
progression, 191
repeating core loops, 189–190
God of War, fun core loop, 190–191
rules and, 188
themes of game, 200
types of, 189
Covet Fashion, 207
cozy games, 11, 194–195, 208
Cozy Grove, 11
Crazy Cake Swap, 47
Crazy Kitchen, 47
Csikszentmihalyi, Mihaly, 11, 123
customization, 161, 163, 207–211, 215, 223
D
deckbuilding, 177, 181–185, 188
design process, 14, 106
Diablo, 150
digital game, 172, 182, 189, 211, 216, 219
Discord, 221
dissection
game design is thievery, 14
know your goal when dissecting, 245–246
meta-dissection, 14
pre-dissection, 13
valuable skill, 13
The Door Problem, 157
downloadable content (DLC) dynamics, 43, 77–79, 95–100
Dress to Impress, 207
Duelmasters, 169
Dungeons & Dragons, 5, 119, 138, 150, 210–211, 217

E
emotion
Candy Crush Saga, 54–55
Centipede, 94
Magic: The Gathering, 186–187
Papers, Please, 233–234, 237–238
Royale High, 214–215
Shovel Knight, 43–45
 Spelunky, 151–152
empathy, 235, 238, 244, 247
Enemy AI System, 96–97, 102–103
England, Liz, 157

F
Facebook, 57, 206, 210
Faidutti, Bruno, 121
Flow, 11, 123, 208
Fluxx, 137
Fortnite, 57, 188
free-to-play games, 55–58, 62, 66
frustration, 44–45, 186, 234
FTL: Faster than Light, 151

G
game(s)
concepts of, 1
definition
challenges, 3–4, 10–12, 19, 25, 36–37, 37, 45, 50–51, 52, 58,
60, 81–84, 83, 102–103, 108–110, 115–117, 117, 122–
125, 127, 132, 134–135, 141, 148, 149, 156, 158, 161–
163, 177, 177, 178–180, 187, 191, 193–194, 198–199,
201, 207–209, 218, 227–228, 227, 231, 231, 235, 238,
245–246
choices, 3–5, 7, 9–10, 15, 17, 33, 50, 70, 84, 91, 112, 116,
127, 127, 128, 130–132, 132, 133, 135, 139–140, 142,
 144–145, 157, 177, 186–187, 200, 206–208, 211, 218,
224, 226, 228, 230–232, 236, 238–241
goals, 3, 5–7, 10–13, 16, 18, 24, 32, 35–36–37, 37, 42, 50–
51, 52, 58–65, 60–63, 65–68, 71–72, 76–80, 83–84, 85,
91, 102, 105, 108–110, 113–115, 114, 117, 121–123, 127,
132, 146–148, 149, 162–163, 171, 174–176, 177, 179,
188, 205–206, 206, 208, 209, 211–212, 214, 216, 218–
219, 221, 223, 226, 227–228, 228, 231–232, 235, 237–
238, 245–247
interactivity, 3–4, 6–7, 9, 16, 19–20, 33, 111, 122–123, 157,
216–218, 245
pretend, 3, 12–13, 210–211, 217, 222
skill, 3–5, 11, 13, 33, 45–46, 58, 63, 67, 69–70, 72, 75, 87,
93–94, 109, 124–126, 134, 137, 139–142, 144, 146, 178,
193, 199, 207–208
uncertainty, 3, 9–11, 15, 56, 115–116, 122–126, 132–133–
139, 141–143, 146, 155, 177–178, 194, 208
development, 8, 13, 91, 159, 164, 246
gamers/non-gamers, 248
history
Candy Crush Saga, 57–58
Catan, 118–119
Centipede, 92–93
Colonialism, 121–122
Magic: The Gathering, 183
Mobile games, 56–57
Papers, Please, 232–233
Royale High, 215
 Shovel Knight, 40–41
Spelunky, 150–151
meaning, 244, 247–248
Candy Crush Saga, 58
Centipede, 94–95
Magic: The Gathering, 188
Papers, Please, 235
Royale High, 216
Settlers of Catan, 122
Shovel Knight, 45–46
Spelunky, 152
monetization, 42
rules, 7–9
systems, 246
terminology, 2
game design, 1, 52
decisions due to non-gameplay factors, 38, 247
is thievery, 14, 40–41, 109
player’s experience, 246–247
Game Design Document, 86, 107, 244
game designer(s), 4, 11, 13–14, 61, 71–72, 97, 110, 121, 124, 152, 159,
169–170, 216, 219–220, 247–248
Game Design Patterns, 14
gameplay, 16, 43, 45
deep analysis of related game, 21
influenced by look and feel of game, 39–40
games industry, 55–56, 119, 209
Gardening Moments, 81–82
Garfield, Richard, 186
Go, 97
goals
in Candy Crush Sag
design, 72–73
goals merging, 63
into bigger goal, 60
into a game, 64
layers of analysis, 59–60
many goals inside another goal, 62
one goal inside another goal, 61
other goals, 65
player-generated, 65–66
swap/swapping, 49–50
verbs and, 71–72
God of War, 32, 102, 189–190
Grand Theft Auto 3 (GTA3), 194–195, 196, 217
graphics, 39, 120, 149–150

H
Hades, 151
Halo, 32, 36–37, 97, 102, 189, 191, 198, 208, 215
Hamlet, 15
Hearthstone, 172
Hocking, Clint, 240
Holocene, 211
Homo Ludens: A Study of the Play-Element in Culture, 12
 Huizinga, Johan, 12
human psyche, 58
Hunter: the Reckoning, 96, 243

I
Indie games, 15, 41, 43, 147, 150–151, 232–233
Infocom, 45, 216–217
input/output system
electrons flow, 28, 28
game designer controls, 25, 26
inputs through a device, 25
jumping/jump state, 28–29, 29, 31
pixels change, 29, 30
pixels on screen, 25
player notices change, 30, 30
player’s brain to the computer, 26, 27
player wants new thing, 30, 31
press the button, 26–28, 27

J
Jackson, Steve, 119
jumping, 16, 28–29, 29, 36, 41, 150, 227

K
Kaijudo, 169
Klondike, 2
Knizia, Reiner, 119
knobbiness, 200

L
Legend of the Five Rings, 187
Way of the Naga sourcebook, 187
Legend of Zelda: Breath of the Wild, 217
Life Is Strange, 35
Lode Runner, 41
loot boxes, 184, 205, 208

M
Magic Circle, 12–13, 45, 182, 189
Magic: The Gathering/Magic, 130, 219
challenges
uncertainty, 178
combo decks, 175
creature, 175–176, 180–181, 189, 191, 197
attacking, 179
casting a spell, 179–180
playing, 178–179
produces mana, 198
secret, 199
examples, 178–180
flow, 177
genre of collectible card games, 172
goals
 attack, 175
build, 174–175
Magic actions, 175–176
Magic cards, 186, 189
non-gameplay dissection
content and rules, see content
emotion, 186–187
history, 183
meaning, 188
money, 184–185
narrative, 187
senses, 186
technology, 185–186
states of game, 180–181
systems
combat, 181–182
deck building, 182–183
mana system, 181
more, 183
spells, 182
verbs
activate abilities, 174
attack, 174
play cards, 173–174
mana system, 177, 179, 181, 186
Super Mario Bros, 41–42
Marvel comics, 240
Marvel Snap, 130, 172
Massively Multiplayer Online game (MMO), 6, 219
Medici, 119
Meier, Sid, 4, 9
meta-dissection, 14
Minecraft, 6, 45, 56, 113
mini-games, 18, 24, 99, 204
Miraculous RP: Ladybug & Cat Noir, 211
monetization, 42–43, 247
Candy Crush Saga, 55–58, 62
Centipede, 93–94
Magic: The Gathering, 185
Settlers of Catan, 120
Shovel Knight, 42–43
Monopoly, 141, 183, 187
pass GO, 33
move and attack, 18, 22, 36, 96
core verbs of, 33
God of War, 32
Halo as, 32, 36
Super Mario Bros, 32
multi-user dungeons (MUDs), 217
Munchkin, 137, 143
Museum of Mechanics: Lockpicking, 41

N
Non-Player Character (NPC), 23–24, 43, 231, 237–238
O
Obbies, 212
Oblivion, 108
One Piece Card Game, 172

P
Pac-Man, 92, 241
Papers, Please, 224, 225
bureaucracy, 231, 236–237
challenges, 227–228
flow diagram, 227–228
flow, 227, 227
with goals, 228
games are about something
choices, 238–240
games matter, 244
know your themes, 239–240
theme everything, 240–241
themes, 236–238
vision, 235, 241–244
goal for the player, 226
non-gameplay dissection
emotion, 233–234, 234
history, 232–233
meaning, 235
money, 235
 senses, 233
for PCs and mobile, 224
systems
core gameplay loop, 229–230
goals of investigation, 230
simple core gameplay, 228–229
verbs, 224–226
PC games, 41–42, 56, 150–151, 202, 235, 247
philosophy
Grasshopper, 11
player-generated goals, 65–66
players, 3
avatar, 25, 94, 101, 159–160, 203, 224, 233
character, 8, 97, 218
creativity, 218–219
freedom, 217
goals, 7, 91, 174
individual skill of, 4
interactivity, 6, 145
Johnny/Jenny, 219
mental capacity, 70
on Roblox, 221
Spike, 219
Tammy/Timmy, 219
Pocket God, 56
Pokémon, 171
Pokémon Go, 53
poker, 2, 72
pre-dissection, 13
progression, 191–195, 199
Candy Crush Saga, 51, 61, 63–65
Centipede, 92–93
Magic: The Gathering, 195
Papers, Please, 230–232
Royale High, 211
Spelunky, 149, 151, 161, 163, 170
puzzle game, 152, 186

R
Reddit, 221
Roblox, 6, 57, 220–221, 244
audience, 220
to Australia, 202
experiences, 213
on other platforms, 215
platform, 202, 206
Studio, 212
Roblox High School, 211
Robux, 215
Rogue, 150
role-playing games (RPGs), 63, 73, 185, 188
Rosewater, Mark, 130, 174, 183–184, 219
Royale High, 202, 203
audience
dissecting, 218
 interactivity, 216–218
no demographics, 221–222
avatar, 203
challenges
customization and decorating features, 207
dynamics, 209–211
flow, 208, 209
goals, 205–206, 206
non-gameplay dissection
emotion, 214–215
history, 215
meaning, 216
money, 215
narrative, 213–214, 214
senses, 213, 213
technology, 212
past
Game as service, 207
game loop, 206–207
systems, 211, 212
understanding
audience, 219, 222
goals, 223
non-existent audiences, 222
other people, 220–221
players of a game, 218
verbs, 205
action, 204
 buy, 203
customize, 203–204
events, 205
mini-games, 204
move, 203
spin the wheel, 204
trade, 204
Rules of Play, 12
running, 16–17, 20, 36, 124, 150, 159, 218

S
score points, 76–77, 79, 113
Scrabble Go, 200
Shariki, 51
Shovel Knight, 15–46, 92, 124, 157
2D platforms, 15, 41
aspect of, 41
game flow level, 37, 37–39
game screen, 26
non-gameplay dissection, 38–39
emotions, 44–45
history and influences, 40–41
meaning, 45–46
monetization, 42–43
narrative, 43–44
senses, 39–40
original game in series, 15, 16
 position, 34
verbed input/output, 31–32
input verbs and output goals, 32
verbs, 15–25
Shovel Knight: Shovel of Hope, 15
SimCity, 6, 37
The Sims, 208
Skyrim, 217
Slay the Spire, 151
Smurf Village, 57
social games, 206
socialization, 209–210
solo games, 2
Space Panic (1980), 41
spells, 70, 172, 175–176, 178, 182
Spelunky, 41, 147–171, 148
2D side-scrolling platformer, 147
flow, 149
gameplay dissection
challenges, 148
goals, 148
progression, 149
verbs, 147
non-gameplay dissection
emotions, 151–152
history, 150–151
meaning, 152
money, 151
 senses, 149–150
systems
add/remove Gold, 167, 167
deep systems, 170–171
dissecting, 159–164
friendships/friend groups, 166
fun, 171
link, 164–167, 165–167
map, 162
map groups, 164
procedural fun, 156–157
randomness, 153
relationships, 165
thinking, 157–159
thinking–feedback loops, 167–170, 168
understanding fun, 153–156
Spiel des Jahres, 118–119, 119
spinning, 243
sports, 3, 5, 124–125, 141, 178, 203, 229, 248
Stardew Valley, 194, 195
Star Wars Unlimited, 172
Street Fighter, 2, 108
Super Mario Bros, 15, 41–42, 240
Super Meat Boy, 233
Super Smash Bros, 141
Sushi Go!, 185
T
target audience, 220–221
Teuber, Klaus, 122, 145
theoretical audience, 222
Tigris and Euphrates, 119
Track and Field, 5
Triple Town, 37

U
UI/UX design, 29, 222
uncertainty, see challenge
Untamed Animals, 211
user-generated content (UGC), 202, 213
User Interface (UI), 25, 164, 225, 230

V
Venn diagrams, 105, 105–106
verbs, 15–25
advanced verb-ing, 16–19
big, 36
Candy Crush Saga, 47–50
Centipede, 74–76
direct action, 19–24
as genre, 36–37
to the goal, 36
interface, 24–25
 Magic: The Gathering, 173–174
overlapping with different goals, 35–36
overlapping with different rules, 33, 33–34
Papers, Please, 224–226
pre-verbs, 17
Royale High, 203–205
Settlers of Catan, 111–113
Shovel Knight, 15–25
Spelunky, 147
understanding, 15–16, 32–33
victory point, 12, 111, 113, 115–118, 125, 127, 131, 145
video game, 2, 8, 25, 27, 29, 33, 49, 58, 74, 85, 93, 95, 107, 116, 120,
124, 177, 183, 194, 208, 211, 247
vision
clear, 243
Mood Boards, 243
points, 242–243
prototypes, 243
statement, 241–242
“why” questions about game, 242–243
visuals, 39, 85, 120, 155, 213, 235, 247–248

W
Where’s My Water?, 56, 106, 157, 199, 201, 242
world economy, 100
World of Goo, 151, 233
World of Warcraft, 2
Y
Yargle, 9/3 Frog Spirit creature, 189, 190
Yu, Derek, 153, 156

Z
Zindagi, 47
Zynga, 47

OceanofPDF.com