Online Multiplayer Games

Copyright © 2010 by Morgan & Claypool

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in
any form or by any means—electronic, mechanical, photocopy, recording, or any other except for brief quotations in
printed reviews, without the prior permission of the publisher.

Online Multiplayer Games

William Sims Bainbridge
www.morganclaypool.com

ISBN: 9781608451425 paperback

ISBN: 9781608451432 ebook

DOI 10.2200/S00232ED1V01Y200912ICR013

A Publication in the Morgan & Claypool Publishers series

SYNTHESIS LECTURES ON INFORMATION CONCEPTS, RETRIEVAL, AND SERVICES

Lecture #13
Series Editor: Gary Marchionini, University North Carolina, Chapel Hill
Series ISSN
Synthesis Lectures on Information Concepts, Retrieval, and Services
Print 1947-945X Electronic 1947-9468
 Synthesis Lectures on
Information Concepts,
Retrieval, and Services
Editor
Gary Marchionini, University North Carolina, Chapel Hill

Online Multiplayer Games

William Sims Bainbridge
2010

Information Architecture: The Design and Integration of Information Spaces

Wei Ding, Xia Lin
2009

Reading and Writing the Electronic Book

Catherine C. Marshall
2009

Hypermedia Genes: An Evolutionary Perspective on Concepts, Models, and Architectures

Nuno M. Guimarães, Luís M. Carriço
2009

Understanding User-Web Interactions via Web Analytics

Bernard J. ( Jim) Jansen
2009

XML Retrieval
Mounia Lalmas
2009

Faceted Search
Daniel Tunkelang
2009

Introduction to Webometrics: Quantitative Web Research for the Social Sciences

Michael Thelwall
2009
iv

Exploratory Search: Beyond the Query-Response Paradigm

Ryen W. White, Resa A. Roth
2009

New Concepts in Digital Reference

R. David Lankes
2009

Automated Metadata in Multimedia Information Systems: Creation, Refinement, Use in

Surrogates, and Evaluation
Michael G. Christel
2009
Online Multiplayer Games

William Sims Bainbridge

SYNTHESIS LECTURES ON INFORMATION CONCEPTS, RETRIEVAL, AND

SERVICES #13

M
&C Morgan & cLaypool publishers
ABSTRACT
This lecture introduces fundamental principles of online multiplayer games, primarily massively
multiplayer online role-playing games (MMORPGs), suitable for students and faculty interested
both in designing games and in doing research on them. The general focus is human-centered com-
puting, which includes many human-computer interaction issues and emphasizes social computing,
but also, looks at how the design of socio-economic interactions extends our traditional notions of
computer programming to cover human beings as well as machines. In addition, it demonstrates a
range of social science research methodologies, both quantitative and qualitative, that could be used
by students for term papers, or by their professors for publications. In addition to drawing upon a
rich literature about these games, this lecture is based on thousands of hours of first-hand research
experience inside many classic examples, including World of Warcraft, The Matrix Online, Anarchy
Online, Tabula Rasa, Entropia Universe, Dark Age of Camelot, Age of Conan, Lord of the Rings Online,
Tale in the Desert, EVE Online, Star Wars Galaxies, Pirates of the Burning Sea, and the non-game
virtual world Second Life. Among the topics covered are historical-cultural origins of leading games,
technical constraints that shape the experience, rolecoding and social control, player personality
and motivation, relationships with avatars and characters, virtual professions and economies, social
relations inside games, and the implications for the external society.

KEYWORDS
online game, MMORPG, virtual world, roleplaying game, human-centered computing,
social computing
 vii

Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

1.1 Types of Online Multiplayer Games . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Preserving Game History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
1.3 Intellectual Approaches to Games . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Research Topic Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2 Historical-Cultural Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.1 A Tale in the Desert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

2.2 Dark Age of Camelot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
2.3 Age of Conan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 Lord of the Rings Online . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5 Star Wars Galaxies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3 Technical Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.1 Latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 Sharding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
3.3 Graphics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28

4 Rolecoding and Social Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

4.1 Systems of Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

4.2 Deviant Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.3 Game Masters and Mentors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.4 Legal Regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

5 Personality and Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

5.1 Psychological Theories and Typologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47

5.2 Game-Based Theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
viii CONTENTS
5.3 Theoretical Debates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.4 Non-Player Character Personality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

6 Avatars and Characters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

6.1 Building a Bond with an Avatar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

6.2 The Quality of Avatar Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
6.3 Secondary Avatars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
6.4 Facing the End . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

7 Virtual Professions and Economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

7.1 Work in Star Wars Galaxies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

7.2 Production in World of Warcraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
7.3 Division of Labor in Professions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

8 Social Relations Inside Games . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

8.1 Emergent Social Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

8.2 Examples of Guilds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
8.3 Quantitative Research on Guilds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

9 Implications for External Society . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

9.1 The Online Game Penumbra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

9.2 What People Learn in Online Games . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
9.3 Research Opportunities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

Author’s Biography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

 1

CHAPTER 1

Introduction
This lecture will emphasize the dominant genre of online games, so-called Massively-Multiplayer On-
line Games or MMORPGs, of which World of Warcraft is the most familiar example (Bainbridge, W.,
2010b,c; Nardi, B., 2010). Although the boundaries of the concept are vague, the core idea is a some-
what realistic computer-generated world in which the user is represented by an avatar or character,
interacting with other characters, under the constraints of rules and to achieve goals set by the
MMORPG’s creators and by the players. The truth is, these are far more than mere games, because
they provide virtual environments where people may socialize, explore, and to varying degrees, create
things, and well as undertake formal quests and engage in duels or battles.

1.1 TYPES OF ONLINE MULTIPLAYER GAMES

Refining categories proposed by Roger Caillois, Gonzalo Frasca (2003), suggest that electronic games
can be described as either ludus or paidia. Ludic virtual worlds are ruled-based games involving
direct competition between players, whereas paidic worlds emphasize free play and creativity with
less emphasis on rule-constrained competition.
At the ludic end of the spectrum stands EVE Online, in which the chief activity of advanced
players is warfare where one corporation of players seeks to seize or defend solar systems of a vast
galaxy against other corporations of players. At the paidic end of the spectrum stands Tale in the Desert,
a virtual ancient Egypt in which people mainly do work of the kind real people might have done
in the land of the Nile thousands of years ago, and direct aggression by one player against another
is outlawed. Most popular multiplayer games fall between these extremes, and they give players
considerable freedom whether they personally want to emphasize the ludic or paidic qualities of the
virtual environment. However, a great variety of multiplayer online games exists or are emerging,
with some but not all the characteristics of MMORPGs, in such rough categories as the following.
Strategy games differ from MMORPGs chiefly in limiting the number of players interacting
with each other, currently from two to sixteen, depending on the game, and by representing the
user by something like an army or an economic system, rather than by a single avatar. The historical
model for strategy games is the highest-prestige board game, chess; although, the online strategy
games that emphasize economic exchange can count Monopoly as their ancestor. A good example is
Warcraft: Orcs versus Humans, the ancestor of World of Warcraft but limited to two players who control
a number of soldiers who not only battle each other directly but also seize territory and resources to
build up their relative strength. The analogy in chess is that a skilled player seeks to achieve a good
position for his or her pieces, quite apart from capturing the opponent’s pieces. The space-oriented
younger sister of Warcraft, StarCraft, innovated by having three competing forces, and Warcraft III
had four. Computerized strategy games often let the user play solo, competing with simple artificial
2 CHAPTER 1. INTRODUCTION
intelligence built into the game, thereby building up skills, before going online to play against other
people.
Non-game virtual worlds are in some respects the opposite of strategy games because they
use the online environment primarily for cooperative rather than competitive purposes. Second Life
is a well-known example (Rymaszewski et al., 2007), but others include There, Active Worlds, and
Entropia Universe. Most of the content of Second Life is user-created, and the chief source of revenue
for the company that hosts it is payment for virtual land on which companies, educational institutions,
government agencies, and dedicated individuals build their own virtual architecture. A script-based
programming language and built-in three-dimensional graphic tools allow people to build their
own objects that actually perform complex functions, and some individuals try to make a living
selling virtual goods. Group meetings are often held, including formal classes, business and scientific
conferences, and even birthday parties, at which each user is represented by an avatar.
In contrast to Second Life, Entropia Universe gives users little power to manufacture things,
but they can own property and conduct business while there are also gamelike elements such as
hunting virtual animals and escaped robots. Both Second Life and Entropia Universe have special
in-world currencies but allow conversion to and from dollars or other conventional forms of money.
Over the years, Entropia Universe has experimented with various schemes in which people would
treat the environment as their real world, for example, selling a space station resort for $100,000 to
an enthusiast for this virtual world who called it Club Neverdie and thinking he could make a profit
renting space in it. That hope was apparently not fulfilled, and when Entropia Universe went through
a huge transformation in August 2009, switching to a more advanced graphics engine and completely
recreating all the virtual environments, Club Neverdie became inaccessible. Similarly, most parts of
the New Oxford cultural center of this virtual planet vanished, including the art museum depicted
in Figure 1.1 at the time when it was displaying works uploaded by residents. Entropia Universe is an
interesting and valiant attempt to explore a range of possibilities for virtual worlds, and it illustrates
the challenging possibilities for evolution along the ludic-paidic spectrum, and between play and
reality altogether.
Microgames is a placeholder term for a range of small multiplayer online games that may
come to have a variety of names as they diversify and consolidate in the coming years. Some have
qualities of the other types, but they are distinguished simply by the fact that they are modest in
their demands for time and in the environment they offer. For example, Travian is a browser-based
Medieval society strategy game with very simple two-dimensional graphics, which the user can
have running simultaneously with other software, hopping over to it every few minutes or hours to
make a couple of moves. Many small games run on cellphones, and others run off social networking
sites such as FaceBook and MySpace. It is technically possible to integrate microgames with an
MMORPG in the near future. For example, a fantasy gamefan could complete some minor activity
via cellphone while riding home on the bus, such as mining virtual metal and using blacksmithing
skills to make a sword out of it, then wield that sword in combat later on a high-quality desktop
machine.
 1.1. TYPES OF ONLINE MULTIPLAYER GAMES 3

Figure 1.1: The New Oxford Art Museum in Entropia Universe.

Live action role-playing games (LARPs). These are like MMORPGs, except conducted in the
“real world” rather than over Internet, but many experiments have been carried out in recent years
that augmented real-world with online elements (Walther, B., 2005; Jonsson et al., 2006). One often
hears that this genre emerged in the 1970s, influenced by Dungeons and Dragons, but arguably, it
can be traced back much farther, to Capture the Flag and even to the first military training mock
battles that must have originated in ancient times. With ubiquitous Internet connectivity, and mobile
access through cellphones or PDAs, it is possible this genre could grow into a major industry, possibly
integrated with traditional activities such as history-oriented tourism, nature trail hiking, or stock
market trading. The main point to keep in mind here is that the possibility of accessing Internet
anywhere at any time opens vast territory which future game innovators will explore, integrating
online games with other more mundane activities.
4 CHAPTER 1. INTRODUCTION

1.2 PRESERVING GAME HISTORY

It is crucial to note that online gaming has a past as well as a future, and the current popularity of a
game does not precisely measure its value for students and researchers. Consider the analogy with
literature. Professors rightly do not assign their students to read the top ten novels on the current
best-seller list, but select works that were published across many decades and even centuries because
qualities of style, content, or historical significance outweigh their contemporary lack of popularity.
Ideally, this would be true for multiplayer online games as well, but there exists no library of out-of-
print worlds, nor academic publishing houses that reissue old games. Yet already, some of the very
most interesting examples have ceased to exist, notably Uru, Matrix Online, and Tabula Rasa.
Uru was the multi-player online version of the highly innovative solo-player computer game,
Myst, and one version of it was called Myst Online. The original game in the series, Myst, was
released in 1993, and Uru was the fourth game in the series, coming at a time when technological
possibilities had expanded considerably and multiplayer online games seemed to be the logical next
step. A multiplayer beta-test version of Uru was opened in late 2003, and about 10,000 people entered
it, only to become refugees when this beta was closed down early in 2004, and the game was sold as
a solo-player game like the earlier ones in the series. A series of developments gave the passionate
Uru refugees temporary opportunities to experience online play as the publishers experimented with
multiplayer versions of the game. Furthermore, Pearce, C. (2010) studied this complex case for five
years, calling it “The Uru Diaspora.” Using ethnographic research methods, she observed refugees
attempt to recreate the Uru world in two non-game virtual worlds, There.com and Second Life. In her
writings about this case, Pearce argues convincingly that researchers should not become confined
within a single popular game, but study relationships across multiple games and recognize that some
of the unpopular ones may be especially worthy targets for research.
Matrix Online (MxO) was a more-or-less successful online game that lasted from March 2005
through July 2009 (McCubbin, C., 2005; Bainbridge, W., 2010a). Based on the popular trilogy of
movies, which started with the widely acclaimed The Matrix in 1999, MxO was set two centuries
in the future within a computer simulation of a 1999 city. Partly inspired by the book Simulacra
and Simulation by social philosopher Baudrillard, J. (1994), the Matrix series engaged issues of false
consciousness, the illusory nature of social conventions, exploitation of workers by capitalists, and
existential doubts about the reality of the physical world. Unfortunately, the logical result of this
train of thought made the city a rather disturbing and confining place, from which the player could
never escape. Yet, the city in Matrix Online was a very impressive accomplishment, visually and
intellectually – vast, realistic, and filled with interesting architectural ideas – in fifty-two substantial
neighborhoods from the filthy slums to the awe-inspiring business district. Hosted by Sony, but
not created using Sony’s technology and thus incompatible with its other online multiplayer games,
MxO may have been a burden to maintain. At the time MxO closed, Sony offered a subscription
plan in which a player got access to an entire suite of games for the price of subscribing to any two,
and after a sufficient number of other games were available, Sony may have felt that maintaining
MxO was of no commercial advantage.
 1.3. INTELLECTUAL APPROACHES TO GAMES 5
Figure 1.2 shows a scene from Tabula Rasa that illustrates its fundamental premise and com-
plexity. Created by a team led by Richard Garriott, a leader in the history of these games and the
son of an astronaut, TR begins with the premise that an ancient extraterrestrial species has split into
two factions, the possibly benevolent Eloh and the power-hungry Neph. Using advanced technol-
ogy, the Neph have launched an army against many solar systems, already defeating the Earth and
now fighting the remnants of our forces on two other planets, Foreas and Arieki, where humans
have gained alliances with the local intelligent species. This picture illustrates the conflict because
a battle is taking place in the background, fought with ray guns against force shields, and the two
aircraft are delivering Neph soldiers to the fray. But it also illustrates two more intellectual themes,
the search for knowledge and the development of technology. The two levitating pylons are Eloh
artifacts, inscribed with glowing hieroglyphics in the Eloh language called Logos, which the player
must acquire through exploration and which confer advanced powers. The figure at the right is the
main character’s clone, a secondary avatar that can be created only once when certain Logos elements
and other resources have been acquired.
Many games have aspects of conflict, questing, and crafting valuable new items or even creating
secondary avatars. What makes Tabula Rasa almost unique was that it couched them in terms of
an idealistic, real-world mission to seek wisdom by exploring the universe. In October 2008, just
before the closure of Tabula Rasa was announced against his wishes, Richard Garriott actually
rocketed to the orbiting International Space Station, delivering a memory unit containing all the
TR avatars, mine included. Given how interesting and even fun I found Tabula Rasa to be, it is hard
for me to avoid the conclusions that many current gameplayers lack the intellectual sophistication to
appreciate the cultural depth some of the best games achieve and that popularity does not correlate
with philosophical significance.

1.3 INTELLECTUAL APPROACHES TO GAMES

Potentially, each of the social sciences and many fields of the humanities could be applied to online
gaming, but here we will seek an integrated approach that also relies upon computer and information
science. While some traditional social science has been written, most of the extant scholarly literature
is in one or both of two traditions: (1) game studies and (2) human-centered computing. Specialists
in this area also need to draw upon two kinds of mostly non-scholarly literature: (1) game instruc-
tion manuals whether produced by professionals or amateurs (e.g., Lummis and Kern (2006)), and
(2) trade books in some way connected to game design (e.g., Prensky, M. (2001)).
As the website of the National Science Foundation explains, “Human Centered Computing
(HCC) research explores creative ideas, novel theories, and innovative technologies that advance our
understanding of the complex and increasingly coupled relationships between people and comput-
ing”1 . A report of an NSF-funded workshop about HCC observes (Sears et al., 2008, p. 8):

1 http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=503302&org=IIS
6 CHAPTER 1. INTRODUCTION

Figure 1.2: A Tabula Rasa exobiologist and his clone.

Multiplayer online gaming produces new cultures, social norms, and communication
mechanisms. Understanding how these social aspects of games develop, as well as what
causes individuals to become engaged in these virtual communities, could prove useful
in other social computing contexts. For example, it would be useful to understand how
certain social norms are established, what kind of interaction the members of the com-
munity are engaged in, and how specific communication mechanisms can be encouraged
as new online communication mechanisms and online communities are developed to
support work-related activities, distance education, and various other applications.
Klastrup, L. (2009) has argued that multiplayer online games must be considered as multi-
dimensional worlds, for which no single analytical approach will suffice:
For instance, we can successfully analyse an “old world” like EverQuest as a game and
learn a lot about the functions of this world by looking, for instance, at the game mechan-
 1.3. INTELLECTUAL APPROACHES TO GAMES 7
ics and the type of goals and challenges the players are presented with. But in doing so
we might not understand why players choose to engage themselves exactly in EverQuest
and not any other gameworld. One might also look at a gameworld like EverQuest pri-
marily as a social culture by looking, for instance, at how people in EverQuest socially
interact with each other and for which reasons, but then we might not be able to explain
why some players enjoy playing a world mostly by themselves.
In her own analysis, Klastrup identifies three distinct ways of understanding gameworlds that
would need to be combined to give a full picture:
1. Fictions: a form of make-believe in which users willingly suspend their disbelief in the reality
of the gameworld.
2. Narratives: meaningful presentations of a series of events within a symbolic frame of reference.
3. Interpretations: representations of “what a world is or should look like.”
I suggest that traditional fictions differed from today’s gameworlds in two ways that could be
described as grammatic: they were told in the past tense and the third person. Once upon a time,
Robin Hood dwelled in the Sherwood forest with his merry men, yet at this very moment, Orastes
is preparing to have his picture taken sitting on his horse in front of the temple of Mitra in the
Aquilonian hamlet called Tesso. Orastes is a character from the Conan stories written in the 1930s
by Robert E. Howard, which tell of a land long lost in the mists of time, yet he lives today as one of
my characters in Age of Conan. One might argue that much science fiction is set in future tense, yet
the stories are almost always written in the past tense – future perfect tense if you insist – as if they
had already happened. At experience level 50, Orastes has completed the research tasks I have set
for him, but until I cancel my game subscription, he can live again in present tense, allowing me to
experience his adventures in realtime, moment to moment.
A few traditional fiction stories are written in first person, like autobiographies, but, typically,
they are written in the third person, about “him” or “her.” Consider this sentence: “Hither came
Conan, the Cimmerian, black-haired, sullen-eyed, sword in hand, a thief, a reaver, a slayer, with
gigantic melancholies and gigantic mirth, to tread the jeweled thrones of the Earth under his sandaled
feet”2 . This is the last sentence of the first paragraph of Howard’s first Conan story, “The Phoenix on
the Sword,” but it pretends to be a quotation from an ancient book called The Nemedian Chronicles.
It introduces Conan to the reader, a powerful barbarian with only the most primitive sense of honor,
quite different from the reader however much the reader may come to identify with him. In contrast,
when I was a small child, my father used to invent stories about a fox who lived in the woods near
our house, named Peters Foxy. But instead of telling them in a simple third person, about how
Peters Foxy saved the other foxes when the floods came, my father narrated in the second person,
describing what I did to help my fox friend. In this case, my father told me I had instructed the foxes
to take the mattresses from their beds and float around on them as if they were rafts. Online games
2 http://gutenberg.net.au/ebooks06/0600811.txt
8 CHAPTER 1. INTRODUCTION
are, to a significant degree, fictions in the second person because they are about your experiences
and accomplishments; although, this is one of the area’s most intense, scholarly debates, how much
the players really identify with their characters and take the initiative that would render the games
first-person narratives.
Traditional narratives tell stories in a linear fashion through time, from a set of starting
conditions to a satisfying conclusion (Abbott, H., 2003, 2008). Typically, a hero seeks to achieve a
goal, against opposition but usually with allies, through a series of episodes or chapters (each of which
is a brief narrative). The human mind likes goal-directed narrative structures, presumably because
this style of thought evolved to serve our daily needs for food and shelter, and then transferred
over into literature. This linear structure often applies to single-player videogames and computer
games, for instance, in Chronicles of Riddick in which the protagonist’s sole goal is to escape from
prison. The nice irony of that example is that whenever Riddick does temporarily escape, he is
always recaptured, and most of the time he has no choices whatsoever concerning what he must do
next. Many fiction authors have experimented with alternatives to the traditional linear narrative,
yet it prevails. Samuel R. Delany’s novel Dhalgren is circular, concluding with the beginning of the
sentence that ends on the first page. The protagonist in Samuel Beckett’s Malone Dies attempts to
cast away all his memories, so he can cease to exist before death claims him, yet his emotions become
aroused and he dies unwilling to let go of his past. Numerous novels employ flashbacks, and some
experiment with moving backward in time, yet is hard to name any successful work of traditional
fiction literature that does not rely heavily upon linear narrative from a start to a finish, perhaps
partly, because books are bound from first page to last.
Much has been made of the non-linear character of role-playing games (Rouse, R., 2001),
but this does not mean they avoid narrative, merely that the player has a considerable range of
choices across multiple parallel narratives. Every gamelike virtual world has a backstory, a narrative
that describes how the fictional world came to be. In the case of World of Warcraft, it is provided by
the three previous Warcraft strategy games, and in the case of Star Wars Galaxies, it is provided by
the four movies that are set in time prior to when the game events take place in the larger Star Wars
narrative. The games tend to emphasize quests or missions, each of which is a brief narrative the
player must act out correctly in order to get credit, although there is great choice which quest to do
next and exactly how to complete it. The larger narrative structure in most such games is actually
highly linear, the progress up the ladder of experience levels which is required to gain access to
many features of the virtual world. Even Entropia Universe, which lacks missions and has no single
measure of experience, requires players to advance along several pre-defined experience tracks in
order to exploit the virtual world successfully, such as strength of armor, weapons, and hunting skill
needed to defend against the ubiquitous nasty animals.
Klastrup defines the interpretation dimension of gamelike virtual worlds somewhat broadly,
suggesting that, at the very least, each of them needs to harmonize with what human beings expect
a world to be. However, many of the games, arguably even all of them, also serve as interpretations
of the particular “real world” we actually inhabit.
 1.4. RESEARCH TOPIC AREAS 9

1.4 RESEARCH TOPIC AREAS

In their 2007 survey of massively multiplayer online games, Achterbosch et al. (2007) reported that
the literature in this field chiefly covered four topic areas:
(1) The social interactions between players.

(2) The different architectures to build MMORPGs.

(3) The effects of latency. (Internet delay time.)

(4) Problems that plague MMORPGs.

Except for the third of these, latency, these categories could be defined very broadly to cover
a variety of things, but the authors seem to construe them fairly narrowly. For example, under
problems, they list specifically these security issues: “... cheating within the game to gain an unfair
advantage; hacking an account and using it illegally; and counterfeiting documents to obtain an
account illegally.” Some of my own research falls into the first category, social interactions between
players, but I also do research in two areas not covered by any of these four where I believe there is
already considerable literature:
(5) Relations between a player and his or her avatars or characters.

(6) The cultural content of the game.

In the pages that follow, we will see much research that does not fit comfortably even in these
six categories, and it would require us to construct a longer list or start the categorization process
over again with broader concepts. For example, many researchers have been looking at the different
personalities and motivations of players. Others are studying the economics both internal to the
games, and external where games compete in the marketplace, and these researchers might not want
to classify their topic as “social interactions.” In addition, there is much literature concerning the
computer graphics required to display the gameworlds and characters on the computer screen, and
some of this connects to human beings, either because it relies upon how the human brain reacts
to different visualization techniques or because the graphics need to depict the human body as it
engages in various actions and gestures.
In computer science, the term architecture tends to be applied to large-scale design decisions,
but there are also many smaller-scale design decisions that are worthy of study. I am particularly
interested in the techniques for operating secondary avatars, such as the hunting animals used by
hunter characters in World of Warcraft, and many other researchers have focused on the (admittedly
simple) artificial intelligence techniques developed to operate non-player characters (Maher et al.,
2005; Merrick and Maher, 2006; Moriarty and Gonzalez, 2009).
Similarly, problems come in all sizes, and many of them relate to social interactions and
architecture, as well as to details of game design. One of the more challenging sets of examples
concerns the auction systems in which players buy and sell virtual items, including products of
10 CHAPTER 1. INTRODUCTION
gathering and crafting professions. Every game I have tried that has one of these uses a somewhat
different user interface, distinctive search tools, unique pricing and bidding rules, and, in some cases,
restrictions on which characters are allowed to use which part of the system. A problem that might be
described as social afflicts games that are not very popular, namely a lack of items for sale that might
be worth buying, simply because there are too few players posting things for sale. I found this in both
Matrix Online and Tabula Rasa, both of which were shut down soon after I finished my research on
them. Matrix Online tried to compensate by placing some items in the auction system automatically,
and the very good auction system in EVE Online seemed to have little to do with what individual
players were selling. Automatic operation of an economic system requires very careful design, and
it, indeed, may require constant adjustment to prevent the system from spiraling out of control.
 11

CHAPTER 2

Historical-Cultural Origins
Today’s games cannot be fully understood without a knowledge of the history of online gaming.
Castronova, E. (2003) offered an early sketch of this history, for example, noting the importance of
Dungeons and Dragons as a pre-Internet influence and citing the early examples of Ultima Online in
1997 and Lineage in 1998, but much research will need to be done before we can confidently chart the
most important influences. For example, the notion that the fantasy stories of J. R. R. Tolkien were
very influential reflects widespread awareness of Tolkien’s work among the general public, while
Tolkien was probably a very minor influence within this specialized field because creators of the
games were deeply familiar with a much older and deeper tradition of heroic fantasy stories by many
authors.
Thus, it is also important to understand what might be called the pre-history of online games:
the deeply rooted subcultures and artistic genres on which the entire field is based. A big part of the
cultural content of any particular game is what players call lore, the mythology and legends of the
game, which often are created by the game designers afresh, although influenced indirectly by the
earlier work of other people. A coherent and compelling mythic story is required to make the world
come alive and draw the player into it. In a study of how this has been done successfully in World
of Warcraft, Krzywinska, T. (2006, p. 383) argues that “the mythic plays a primary role in making a
consistent fantasy world in terms of game play, morality, culture, times, and environment. It provides
a rationale for players’ actions, as well as the logic that underpins the stylistic profile of the game, its
objects, tasks, and characters.”
A Tale in the Desert is a good example of a gamelike virtual world based on actual human
history because it seeks to replicate ancient Egypt, with an emphasis on the technology and socio-
cultural structure. Dark Age of Camelot draws, to some extent, upon real European history, but it
draws even more intensely upon the folklore of Dark Ages societies. Some games recycle or build
upon stories, characters, and concepts developed for another medium, notably novels and movies.
Examples include Age of Conan, based upon Robert E. Howard’s Conan stories, Star Wars Galaxies
based on the movies by George Lucas, and Lord of the Rings Online based on J. R. R. Tolkien’s
novels and on the movies based on the novels. This dimension of culture can be studied from an
anthropological perspective or from the perspective of literature and dramatic arts. But it also has
profound implications for how we understand the game and what it offers to players.

2.1 A TALE IN THE DESERT

The little-known treasure among online games is Tale in the Desert, a charming and educationally
valuable depiction of ancient Egypt that has far fewer players than it deserves. This is not to say
12 CHAPTER 2. HISTORICAL-CULTURAL ORIGINS
that it depicts ancient life with perfect accuracy, but it does capture the spirit of the past in a way
that communicates some real facts. The world represents a large but, admittedly, somewhat reduced
model of the Nile valley where the focus is on building a virtual society and gaining knowledge of
the environment, rather than fighting. Indeed, although players may compete to accomplish various
feats, there is absolutely no violent combat. Much of the action involves solving puzzles, building
things that are beyond the capacity of any one individual, and conducting rituals that require two or
more people.
My research was done during the fourth telling of this tale, which is restarted with new features
every year or two. I built a home on the east bank of the Nile in the center of the country, joined three
guilds because multiple guild membership is permitted, began raising sheep, and started constructing
tools for manufacturing. Frankly, a good deal of the technology looks like the nineteenth century
AD, rather than BC, but it offers an excellent metaphor of technology in general, including the
necessity of establishing a complex system of resource supply and product manufacturing. Many
players may find this game boring because, for example, building a house to complete one of the
earliest missions requires to first make a thousand bricks, but that may realistically reflect the fact
that it was hard work to survive in the ancient world.
The game’s own newbie-zone introduction provides the best example for present purposes.
My character, Ren Hotep, began on an isolated island, with the task of gaining the skills and materials
needed to build a boat to sail to the Egyptian mainland. He started with practically nothing. Walking
along the shore, he occasionally found a piece of slate rock, and banging two of them together often
fractured one in just the right way to create a cutting blade. One stone blade and four additional
pieces of slate can make a wood plane to turn wood gathered from the nearby trees into boards,
although the blade breaks occasionally and needs to be replaced.
It sounds simple to make a boat because the materials needed are 20 pieces of wood, 20 boards,
4 ropes, 4 units of tar, and a sail. The tar can be collected easily from a small tar pit, although the
player needs to hunt all over the island to find it. The sail is graciously provided, so that leaves only
the rope. This is where the plot thickens because making rope turns out to require a huge number
of gathering and crafting steps. The rope must be made from flax, and the player receives a gift
consisting of flax seeds. So far, so good. Planting the seeds, weeding the garden, and waiting until
the flax can be harvested takes a little patience, but the real problem is that one unit of seeds produces
one unit of flax, consuming all the seeds. To get enough seeds, the player needed to plant some but
avoid weeding, and then wait until the flax has gone to seed, which yields a profit in seeds. With
practice, Ren Hotep could manage four little gardens at once, but had to do some over and over
again, building up his stock of seeds, then investing them in growing flax.
The next step is to release the fibers from the flax plant. The technical term for the method
used is retting, but the game calls it rotting, which is also factually correct and more familiar. The flax
is placed in water, where microorganisms go to work on it. Ren Hotep would drop a bundle of newly
harvested flax into a nearby stream, and he would wait. The next step, separating the retted flax
into its components, requires a flax comb. Unfortunately, the flax combs in the game are substantial
 2.2. DARK AGE OF CAMELOT 13
devices, requiring 18 boards, 60 thorns, and 36 bricks to build. Ren made more boards with his
plane, and he gathered thorns for nearby bushes, but what about the bricks?
Bricks can be made from mud, sand, and straw. Mud and sand can be gathered from different
kinds of soil, but straw must be made from grass. The process was simple: gather grass, drop it in
bundles on the ground, and they will dry into straw under the blazing Egyptian sun. The need for
straw was clearly stated in the Bible, Exodus 5:7, where Pharaoh punishes the Israelites by saying,
“Ye shall no more give the people straw to make brick, as heretofore: let them go and gather straw
for themselves.” These components must be assembled in a brick rack, made from four boards and
liable to break after several uses. One of the components of flax, tow, can be spun into twine, and
twine can be spun into rope. This requires a small distaff, which is made from 10 pieces of wood, 12
boards, and fully 100 bricks. Figure 2.1 shows the three major pieces of equipment: the wood plane
(center foreground), flax comb (left), and small distaff (background).
All three of these machines are far larger and more complex than the average ancient Egyptian
would have used. Well-cut boards were probably rare in ancient times, and they were formed by saws
rather than planes. A flax comb, often technically called a heckling comb, was also worked by hand,
whereas the one in the game operates by itself. Even in recent centuries a distaff was a small stick
used for manual fiber working, and here it seems to have been combined with a rather formidable
spinning wheel. Probably, it would have been difficult for the game designers to work out the detailed
graphics needed to show skilled hand work with comb and distaff, and the machines they came up
with were marvelous to watch in operation as historically incorrect as they may be.
The fundamental lessons of this introduction to ancient Egypt are actually rather profound.
Starting with nothing at the dawn of history, the human species needed to bootstrap its technology
to a complex level, in order to support a populous and complex society. To do so, humans needed
to invent tools, including tools to make tools, investing in the physical capital needed for economic
production. Many processes needed to be carried out in a precise series of steps, following plans
that modern computer scientists would call algorithms. First, plant the seeds, then harvest the flax,
then rot it, then run it through a comb, then spin the resultant tow into twine, then spine the twine
into rope. Everything that happens in a computer game depends on how humans interact with
algorithms, but perhaps that is also true for “real life.”

2.2 DARK AGE OF CAMELOT

Some games take history very seriously without attempting to duplicate past eras accurately. A
good example is Dark Age of Camelot, which is set in a period shortly after the death of King
Arthur during the western European Dark Ages (Mylonas, E., 2005). The world is divided into
three separate cultures which are at war with each other. Albion is the English society of Arthur’s
legacy, and Camelot is its capital city. Midgard is the Viking society, representing Scandinavian and
Teutonic traditions, and it has a capital named Jordheim that does not seem to correspond to any
historical place. Hibernia represents Celtic folklore, named after but not limited to Ireland, and its
capital is Tir na Nog, a mythical place in Celtic legends about an Otherworld that exists over the
14 CHAPTER 2. HISTORICAL-CULTURAL ORIGINS

Figure 2.1: Ren Hotep Making Rope in Tale in the Desert.

sea, under the ground, or on a supernatural plane of existence. Figure 2.2 shows a Viking visiting
Stonehenge in Albion, a member of one of these warring societies acting as a peaceful tourist in one
of the others. This was possible for her only because she exists on the one “cooperative” server in
which the three realms are at peace with each other.
The game’s Wikipedia article suggests that Albion is the most historically accurate. Hibernia
is the least accurate, and Midgard falls between them:

Albion is based on Arthurian legend, with such notable real-world places as Hadrian’s
Wall, Stonehenge, and other locations in Great Britain. The races and classes of Albion,
in the original game and early expansions, tended to be more professions and peoples of
European history and mythology than inventions of the gamemakers. Hibernia is based
 2.2. DARK AGE OF CAMELOT 15

Figure 2.2: A Viking Visiting Stonehenge in Dark Age of Camelot.

on Celtic folklore and the landscape includes lush green rolling hills typical of Ireland.
Although the quest storylines, place names, and numerous game elements are firmly
fixed in Celtic mythology, Hibernian races and classes are typically the creation of the
gamemakers. Midgard is based on Norse mythology and its landscape includes misty
fjords and pine forests... The adherence of the architecture to the forms and design of
medieval and pre-medieval Norse architecture is notable1 .
Both Camelot and Stonehenge are the centers of elaborate legends, mythologies, and occult
speculations, yet clearly Stonehenge is real, and Camelot may have been as well. As it happens, I
visited Stonehenge twice, and the first time (in 1965), it was still permissible to touch the stones
to verify their reality. It is possible I may have visited Camelot as well, although I cannot test this
hypothesis empirically. Archaeologists, historians, and folklore enthusiasts have speculated that a
1 http://en.wikipedia.org/wiki/Dark_Age_of_Camelot; retrieved September 19, 2009.
16 CHAPTER 2. HISTORICAL-CULTURAL ORIGINS
small number of sites in England might have been Camelot, and one of them is Cadbury Castle
in Somerset. In 1984, I tramped around this completely ruined hill fort, recognizable only by its
surviving perimeter wall, wondering where I was in history. The story of King Arthur is especially
interesting because there really may have been such a person, although living in an illiterate society
that did not record the facts of his life, and his current Wikipedia page nicely summarizes the debates
about him2 . If Arthur lived in the Dark Ages, written stories of his exploits do not date until much
later, in the Middle Ages or even early Renaissance. The lack of contemporary written records does
not prove he and his city were fables, but it leaves us wondering. Camelot is a lot like Troy, a city that
was remembered in Homeric Greek legends but not confirmed until Heinrich Schliemann excavated
the site in the 1870s, except we may never have proof in the case of Camelot.
The standard image of Camelot today is the one that writers hundreds of years ago believed
in, and Dark Age of Camelot depicts it as a late Medieval castle town, more than a Dark Ages citadel.
This kind of anachronism is quite common in popular histories, seeing the distant past from the
point of view of the more recent past. This phenomenon deserves a technical name. I rather like the
term vector, because in computer science, it can refer to a dynamic data array, and in mathematics,
it often refers to representations that have both length and direction. Given all the many meanings
already attached to vector, introducing one more usage cannot increase the confusion. A vector history
is one that looks at a particular point in past time from another, later point in past time, or from a
different culture. Note that in doing a vector history, we select a point of view (in time and cultural
space) that is not our own, and we then look at a third point in human history from that perspective.
This kind of relativity is illustrated in several ways by Figure 2.2 The version of Stonehenge
depicted in the game is not very different from the real one, perhaps a little wider but having many of
the same features. At the time of King Arthur – assuming he lived in the real Dark Ages – Stonehenge
was already perhaps three thousand years old. Some of the legends say it was built by Merlin using
magic. Supernatural power was not required to raise the stones, but it certainly would have been
required to do so in the distant past, long before the magician was born! Modern adherents of Druidic
religions like to think it was a temple of their ancient ancestors, and many quests in the game require
fighting Druids, but realistically even the Druidism encountered by the ancient Romans may have
arisen when Stonehenge was two thousand years old. Dark Age of Camelot depicts Stonehenge in
nearly mint condition, and none of its stones have yet fallen down. The lone stone on the horizon
in the picture, possibly the famous “heel stone” and defended by Druid magicians, currently leans at
an angle, as I confirmed in 1965 by leaning on it, but the game shows it still standing fully upright.
The Zone of the game where Stonehenge is found carries the correct name, Salisbury Plain, and it
possesses a number of other ancient circular trenches and standing stones, although I have not yet
located the virtual equivalents of two other spectacular nearby sites, Avebury and Silbury Hill. Well,
one of the circles of standing stones look very much like Avebury, but without a label, how can I tell
if it really “is?”

2 http://en.wikipedia.org/wiki/King_Arthur; retrieved September 19, 2009.

 2.3. AGE OF CONAN 17
The Viking in the picture is my valkyrie, Reitsche, and her vector perspective is cross-cultural
more than anachronistic. To their credit, the creators of Dark Age of Camelot have used a large
number of Scandinavian names and words for the Midgard faction, as well as folklore elements, but
the dominant modern image of valkyries is German, notably the familiar music to the Ride of the
Valkyries in Richard Wagner’s opera, Die Walküre.The opera actually contains nine valkyries, so when
I was creating this character, I methodically entered the name of each one, discovering somebody else
had already taken it for their valkyrie character. With Wagner’s music running through my mind, I
tried to think of yet another valkyrie name, and the best I could do was remember the actual heroic
German woman test pilot, Hannah Reitsch, adding a final “e” to make it like the other valkyrie names.
This seemed a logical choice, given that she was the first woman to pilot a helicopter, a jet fighter
plane, and an honest-to-goodness rocket ship. Notice again the vector history. Many players selected
names from a nineteenth-century German opera, and I selected the name of a twentieth-century
German, to name a Scandinavian character from perhaps the eighth century.

2.3 AGE OF CONAN

The Conan stories are based on the clever conceit that many aspects of ancient European and
Egyptian culture were remnants of an even earlier age that has been entirely forgotten, that itself
was built upon the ruins of the even earlier period in which Atlantis flourished. Like Dark Age of
Camelot, Age of Conan depicts three separate cultures, although one difference is that they are not
at war with each other. Stygia represents ancient Egypt, Aquilonia represents the Roman Empire,
and Cimmeria represents Celtic traditions but also barbarism more generally. This scheme allowed
Robert E. Howard to draw upon ancient history without being a slave to it and to imbue his stories
with intense emotionality despite their scholarly framework.
In recent years, both social science and computer science have rediscovered emotion. In his
presidential address to the American Sociological Association, Massey, D. (2002) explained that
emotion is an integral part of social interaction and of individual cognition, and thus it cannot be
ignored in any serious analysis of human behavior. The term affective computing is widely applied to
methods that endow computers with the ability to recognize and to respond realistically to human
emotions (Picard, R., 1997; Peter and Beale, 2008).
The emotions expressed in the Conan stories are chiefly rage, lust, and terror, although Conan
himself is incapable of feeling fear. In the technical meaning of the term, these are immoral stories,
because Conan is allowed to kill other people merely to serve his own self-interest or express his
momentary impulses. It is worth contemplating the extent to which this immoral ethic exists in many
MMORPGs because the player is often required to kill non-player characters or even characters
operated by other players, merely to advance in the game. Morality aside, the Conan stories are
psychiatric and can suggest principles from depth psychology because they sprang from the arguably
pathological personality of the author.
At the age of thirty, in 1936, when told his mother was about to die, Robert E. Howard
killed himself. Although he exercised physically to gain at least a fraction of Conan’s brute strength,
18 CHAPTER 2. HISTORICAL-CULTURAL ORIGINS
emotionally he was a child tied to the apron strings of his mother, and he had promised her as early
as age six that the two of them would always live and die together (De Camp et al., 1983). Although
he wrote stories about ancient history and exotic nations, he had little higher education and hardly
ever left his native Texas. Thus, the Conan stories are the extreme in vicarious compensation for
real-world deficiencies, the ultimate in escapism. A popular slogan in the science-fiction subculture
says, “Science fiction is escape... into reality.” We may well ask to what extent today’s multiplayer
online games are also escapist, versus legitimate exploration of an emerging new reality.
There is an even deeper level of analysis suggested by the case of Conan, a kind of historical
psychoanalysis. It is remarkable how many of Sigmund Freud’s writings are historical in focus, and,
arguably, psychoanalysis is about two intertwined pasts: the cultural history of humanity and the
psychological history of the individual. Indeed, psychoanalysis itself may already be an historical
artifact because its theories are based heavily upon the family experience of a certain class of people
in Austria a century ago, in which fathers were powerful, whereas in Howard’s case, the mother was
powerful. The key psychoanalytic publication that brings together history and personality is Freud’s
1930 book Civilization and its Discontents, arguing that the wide sweep of history has required
individuals to subordinate their personal impulses more and more to society’s demands, leading to
frustration, inhibition, and widespread neurosis. Conan smashes civilization with his sledgehammer,
achieving the triumph of desire over discretion. Ironically, the violence in multiplayer online games
may represent the liberation of rage that has been suppressed by civilization, using the most civilized
of technologies to unleash primitive forces.

2.4 LORD OF THE RINGS ONLINE

Everybody knows that Lord of the Rings Online is based on a novel by J. R. R.Tolkien,The Hobbit, and
on his Lord of the Rings trilogy, but people may not be widely aware of what that means for the values
embedded in the game. On the one hand, Tolkien was a professor of historical linguistics, philology,
and literature, so his creative works of fantasy are based on far more scholarly erudition than is usually
the case. On the other hand, his works are in a deep sense religious allegories. A Roman Catholic,
he was also conversant with the Protestant and Anglican traditions, and approached Christianity
in the manner that highly educated but devout members of English culture often do, with a poetic
detachment and sense of aesthetic grandeur.
Conventional religion is underrepresented in videogames and online virtual
worlds (Bainbridge and Bainbridge, 2007a). Given the fact that two of the three popular
videogame system companies are Japanese (Playstation and Nintendo), diluted versions of Asian
religions are found in many games. Occasionally, sectarian or speculative versions of conventional
western religions creep in, for example, in the videogame version of The Da Vinci Code. However,
most religious references in multiplayer online games concern pagan religions of the past, religious
cults, or fantasy religions similar in many respects to ancient polytheistic religions. A striking
exception is the videogame version of The Chronicles of Narnia: the Lion, the Witch, and the Wardrobe,
based on a novel by Tolkien’s close friend C. S. Lewis, which explicitly teaches love and respect
 2.4. LORD OF THE RINGS ONLINE 19
by making the player operate four characters who are children of the same family and who must
cooperate to succeed.
The trilogy of novels, Lord of the Rings, concerns a battle between good and evil, focused on
assembling and using a ring of immense magical power. Even the best and most moral heroes become
seduced by the possibility of gaining personal power with the ring, but in the end it is destroyed and
balance is restored to the Earth. Players in Lord of the Rings Online gain many of the abilities of players
in other games, such as World of Warcraft, but they are not couched in magical terms. For example,
Rumilisoun, my elf lore-master, does not cast spells at enemies, but hurls hot coals at them, which has
exactly the same effect but does not violate conservative Christian prohibitions against practicing
magic. Over the centuries, Christianity has struggled with the issue that belief in supernatural
powers implies the possibility of monopolizing some of them for one’s selfish benefit, and its general
response has been to reject magic and rely upon God’s grace instead (Stark and Bainbridge, 1985,
1987). Figure 2.3 shows Rumilisoun visiting her friends the Hobbits in The Shire.

Figure 2.3: An Elf Visits the Hobbits in Lord of the Rings Online.

Elves, in Tolkien’s world, are essentially immortal, although they tend to behave more like
detached philosophers than like angels. Hobbits, like the one walking just behind Rumilisoun, are
modest people, living modest and limited lives, but who have long dwelled peacefully in The Shire,
20 CHAPTER 2. HISTORICAL-CULTURAL ORIGINS
which expresses one traditional view of “Merrie Olde England.” However, The Shire may represent
another mythical land, in that, it lacks any church or temple. Where in all of human history was there
never a church, temple, synagogue, or mosque? Answer: the Garden of Eden. If one lives perfectly
in harmony with fellow beings and with God, there is no need of religion. Thus, unlike many of the
other virtual worlds described here, Lord of the Rings Online largely lack religious institutions, not
because it is secular, but precisely because it is in its entirety a religious allegory.
The Lord of the Rings is often described as a prominent example of high fantasy, whereas the
Conan stories are low fantasy. The distinction does not concern quality or cultural sophistication
but the extent to which the mythos is connected to actual history and the folklore of real societies.
High fantasy, like Tolkien’s stories, is set in an entirely fictional world which is quite distinct from
the real world, whereas low fantasy is on balance less imaginative because it includes many elements
taken from the real world. Obviously, this distinction is a matter of degree. Orcs and elves were long
part of European folklore, and Tolkien’s hobbits seem to speak the Anglo-Saxon language and thus
to be English. Another category, sometimes called science fantasy but more commonly called science
fiction, gives fantasy ideas a veneer of plausibility by couching them in the metaphors of science and
technology.

2.5 STAR WARS GALAXIES

This online multiplayer game is obviously based on the Star Wars movies, and it depicts fa-
miliar fictional planets such as Tatoonine and Naboo, where players have created vibrant eco-
nomic and social systems that allow them to experience living in one of their favorite fan-
tasies (Ducheneaut and Moore, 2004; Ducheneaut et al., 2007). Specifically, Star Wars Galaxies is
set just after the action of the original 1977 movie, but three prequel movies were later made, so it
falls between the fourth and fifth of the six main movies in terms of the story. Indeed, on its fifth
anniversary, it gave players the opportunity to travel to a remote location and congratulate Luke
Skywalker for having just destroyed the Death Star and even to inspect his rocket- propelled fighter
craft. Some locales from the chronologically last movie are included, such as Jabba’s fortress and the
Ewok villages, but presumably these existed for years prior to their appearance in the film.
Multiplayer games based on familiar stories face a very difficult problem, in that they cannot
duplicate those stories and must find other ways to exploit their settings. The reason is that it would
be highly unrealistic to run the same events over and over again, with players lined up around the
block to slay Darth Vader one more time. Games that are not based on popular stories have this
problem of dramatic unrealism as well, but to a lesser degree because they offer very broad plots with
many villains to slay in many different sequences. Solo player games do not have this problem, and
an excellent example is Star Wars Episode II: Revenge of the Sith. The player alternates running either
Obi-Wan Kenobi or Anakin Skywalker through the plot of the movie, until they meet at the end in
the battle that turns the latter into Darth Vader. No other players intrude to spoil the player’s sense
of immersion and dramatic coherence.
 2.5. STAR WARS GALAXIES 21
Just as the game was based on a series of movies, the movies were heavily based on a pre-existing
artistic genre and even to a remarkable extent on specific earlier works. In the most obvious way imag-
inable, the Star Wars movies are inspired by the Flash Gordon movie series of the 1930s (Kinnard, R.,
1988; Gordon, A., 1995). The very opening, when paragraphs of text scroll up and away describing
the previous episode of the story, is taken from the 1940 serial Flash Gordon Conquers the Universe, and
the famous Star Wars music was directly inspired by the classic Flash Gordon soundtracks. Audiences
may have been confused in the escape scene of the 1977 Star Wars movie, when Luke accidentally
destroys the controls for a light bridge and needs to hold Leia as they swing across a chasm on a
cable like Tarzan and Jane on a jungle vine. Several scenes of the 1938 serial, Flash Gordon’s Trip
to Mars involve a light bridge, including one in which the heroes escape. The Flash Gordon movies
were based on a comic strip, which in turn was based on the Buck Rogers comic strip which in turn
was based on serious science stories written by Philip Francis Nowlan, and Flash Gordon also drew
on the ten influential Mars novels by Edgar Rice Burroughs.
Although no one work of literature can claim all the credit for launching the genre from
which Star Wars drew or, indeed, for providing the mythic basis for all online fantasy games, the
novels of Edgar Rice Burroughs were tremendously influential (Porges, I., 1975; Lupoff, R., 1976).
Beginning with his first published story in 1911, Burroughs created a complex imaginary world, not
unlike Tatooine or World of Warcraft.The fourth novel in the series,Thuvia, Maid of Mars, introduced
the idea that women could be brave science-fiction warriors equally with men, and it included a
glossary of Martian terms. The Chessmen of Mars included instructions for playing a Martian game
with some affinities to Warcraft: Orcs and Humans. The fundamental premise has carried over into
many games, including Star Wars Galaxies: an advanced, decaying society has reverted to feudalism
with constant warfare, mixing anachronistic technologies such as swords and spaceships, and where
technology blends imperceptibly with magic. Luke Skywalker’s Tatooine has moisture vaporators
because the planet has lost its seas, whereas Burroughs’ Mars has atmosphere factories because the
air is leaking away into space. Luke uses the magical Force to aim his missile at the right point on
the Death Star, whereas the hero of Burroughs’ first novel uses telepathy to open the locked door of
the malfunctioning atmosphere system in order to save his entire world.
Most of the games described here owe a debt to the table-top game, Dungeons and Dragons,
and D&D acknowledges its debt to writers like Burroughs who created the century-old literary
genre of heroic science-fantasy that inspires even many of the most recent online games (Gygax, G.,
1979). Within the science fiction literary subculture, various technical terms are used to describe
different kinds of fiction (Bainbridge, W., 1986). Star Wars is space opera, whereas some of the other
games are closer to sword-and-sorcery.
Space operas are extravagant adventures set against a fanciful interplanetary background, in
which little attention is paid to scientific realism, and much of the technology functions as if by
magic but usually wrapped in the rhetoric of machinery. Sword-and-sorcery is a variety of fantasy,
typically concerning the adventures of brave heroes trekking through barbarian lands, wielding
primitive weapons against demons and wizards, but sometimes including science-fiction elements
22 CHAPTER 2. HISTORICAL-CULTURAL ORIGINS
as well. The point of this example is to show that a serious attempt to understand the mythos of a
modern gamelike virtual world will need to take account of the older culture to which it belongs,
reaching back a century and more, long predating the invention of Internet.
 23

CHAPTER 3

Technical Constraints
Many behind-the-scenes technical details about online games can be ignored by players, but four in
particular need to be understood: latency, sharding, graphics, and what, for want of a standard term,
I call rolecoding and which will be covered the next section. Latency is the delay caused by the need
to send information between each player over Internet and the central server that combines their
actions and controls the game, and it is a crucial if narrow topic, as Achterbosch et al. (2007) noted.
Sharding is the need to separate out some areas of the virtual world for selected players, either to
provide them with distinctive experiences or to manage the large number of players who may be
online at the moment. Graphics concerns how the virtual world and its contents are displayed, and
here we will focus on those aspects that are especially relevant to the user’s experience.

3.1 LATENCY
Griwodz and Halvorsen (2006) studied latency issues in Anarchy Online, which uses the standard
Internet protocols, Transmission Control Protocol (TCP) and Internet Protocol (IP). This is a
decades-old set of standards for transmitting packets of data, and one of its features is that it checks
to see if each packet reached its destination and will resend the packet if not. This is crucial for
sending documents over Internet, but it may not be entirely necessary for all information transmitted
in games. A subset protocol, User Datagram Protocol (UDP), does not verify that a packet reached
its destination and, therefore, tends to function more quickly if less reliably. A study of the Chinese
game ShenZhou Online, by Chen et al. (2006), indicated that TCP was not good for multiplayer
games because packet loss introduced too much latency.
Anarchy Online used the full TCP/IP with packet arrival verification. Consider two scenarios:
(1) an Omni-Tek player runs past an enemy Clan player in Anarchy Online, and (2) an Omni-Tek
player battles a Clan player. If the Omni-Tek player is merely running past the Clan player, the loss
of one packet of data may not be significant. Well-designed games can interpolate the positions of
characters, and they do not need to know the exact location of another character at every split second
of time. Thus, UDP would be sufficient for situations like this. However, if the two players battle,
as they may be expected to do at higher experience levels of Anarchy Online, a lost packet could be
disastrous. If one player blasts the other with a ray gun, but the other gets no record of having been
blasted, the two may wind up with different information about who killed whom.
Analyzing the actual data stream, Griwodz and Halvorsen found that Anarchy Online was
sending very little data back and fourth, usually fewer than four small packets (about 120 bytes) per
second per player. They comment that this means that UDP would not reduce latency much. They
found that the round-trip delay time was typically on the order of 250 milliseconds – a quarter of
a second – although not infrequently taking a full second, and in one case, data took 67 seconds to
24 CHAPTER 3. TECHNICAL CONSTRAINTS
make a round trip between a player and the game server. Players call this lag, and it is a frequent topic
of discussion in text chats of many games I have played. Many of the games have built-in latency
meters, and mousing over them displays the exact time required for recent transactions, which I have
seen as low as 90 milliseconds but more typically runs about 200, in my experience.
There have been many proposals to change the communication patterns in online games, for
example, moving to a peer-to-peer system that avoids the need to communicate through a distant
server, but they typically run into security problems. If the key variables – such as a player’s wealth
and weapon capabilities – are stored on a secure computer belonging to the game company, then it
will be difficult for players with hacking skills to modify the data to their advantage. More complex
network architectures have also been proposed, with some data located on local servers, and some at
central points (Aggarwal et al., 2006). However, the complexity of these ideas raises its own security
issues, and game companies would need to invest in creating a distributed system that had significant
maintenance costs, achieving only modest reduction in latency.
Research by Pantel and Wolf (2002) suggests that latencies even as low as 50 milliseconds
– one twentieth of a second – can degrade performance in real-time multiplayer games. There-
fore, Bosser, A. (2004, p. 263) has argued that online games require a tradeoff between gameplay
quality and the technical characteristics of Internet: “It is impossible to allow dynamic shared state
to change frequently and guarantee that all hosts simultaneously access identical versions of the
world.” A 200 millisecond latency would be very annoying had the game designers not covered it
over by the way they choreograph the action. When two players are battling, they do not usually have
control over every sword thrust, and many of the actions take considerable time for the character to
complete. A magical spell takes time to cast, and time to have its effect, thereby covering over the
latency so the player does not perceive it. Latency would be especially problematic in melee fighting
– toe to toe battles – so a variety of choreographed moves are used to hide it.
An extreme example is Boadicea, my Bear Shaman character in Age of Conan, a melee fighter
who tends to use a sledge hammer as her weapon. This game uses combos, powerful attacks that
require combinations of two key presses to execute. With the adjustable user interface, I have five
different combos set up on the ten number keys 1 through 0. I press the 1 key to start the sequence,
and then when an indicator flashes on the screen I quickly press 2 to have her start swinging the
sledge hammer. Then I press 3 and wait for another indicator, when I quickly press 4. By the time
I reach the 0 key, the fight may be over, but if necessary, I can return and do some of the two-key
combos again, although each is available only after a significant delay. As I am doing all this, I am
not really conscious of every hit that takes place in the battle, although I do often glance at the two
health meters to see whether Boadicea needs to sprint away from her opponent rather than finish
the fight. Boadicea constantly swings her sledgehammer, virtual blood splatters in all directions, and
sometimes big red stains even cover parts of the screen as if it were a bloody window. All of these
complex actions overwhelm my awareness, and Internet latency is not a factor.
In other contexts, even somewhat low latency can be highly visible. For example, a minor
fad in online gaming is the use of two or more computers and avatars simultaneously, called multi-
 3.1. LATENCY 25
boxing. I have observed a World of Warcraft player operate four characters simultaneously, on four
computers, but my own experience is limited to two at once. Having two characters online helped
me document the meetings of a scientific conference I organized inside World of Warcraft in May
2008 (Bohannon, J., 2008a,b) because they could take pictures at opposite ends of the group, ensure
that at least one would record the text chat reliably, and perform different roles as when I had pre-
programmed them to make formal announcements when I pressed a single key to run a macro. I
also used two accounts, computers, and characters to analyze in some detail how pairs of characters
could cooperate in action.
My best example was when Alberich the Dwarf hunter and Stephie the Gnome warlock went
questing together. They would identify an enemy to attack, or a wild animal to hunt, and Alberich
would move to a point just outside the distance that would have aggravated an attack. He would
go into attack mode, but given the distance, this would not yet initiate the attack. Stephie would
stand behind him and shoot a hostile magic spell at the target. The target would run to retaliate
against her but encounter Alberich first, who without any input from me would begin slamming
with his axe, bringing the enemy to a halt. I would operate Stephie, while Alberich would battle
autonomously, until the enemy was dead. The strangest part of all this was what I saw on the screens
of the two computers I had set up, side-by-side. The timing of events was very different on the two,
things happening very noticeably earlier for the character being run from that computer, and later
for the character being run on the other computer, because the two computers were connected only
over a very lengthy Internet connection – a foot apart in the physical world but literally a thousand
miles apart via Internet. This was most obvious if I hit both space bars simultaneously, making both
characters jump. The character that belonged to one computer would have finished jumping just
when the other one began.
Inspired by their study of Starsiege, a game designed for low-bandwidth Internet connec-
tions, Dyck et al. (2007) have outlined a number of principles for managing bandwidth and latency
in games that can be usefully applied to other online purposes, such as computer supported cooper-
ative work. Among these principles, data compression and good design are used to limit the amount
of information that must be transmitted, and adaptive flow and reliability priorities are used so the
system can degrade gracefully when latency rises.
Research on Internet latency can contribute to our understanding of the social dynamics in
games, as a study by Suznjevic et al. (2008) illustrates. Using a software protocol analyzer named
Wireshark, the team analyzed over a gigabyte of the Internet traffic on six personal computers
playing World of Warcraft. In descending order of size of data packets, four different kinds of activity
were compared: player-versus-player (PvP) combat, team raiding, completing quests, and economic
trading. The effect of latency naturally depends upon the characteristics of a game and the different
kinds of actions performed in it. Sheldon et al. (2003) studied the real-time strategy game Warcraft
III and found negligible impact on the overall outcomes, but they found some hindrance specifically
of exploring the game world.
26 CHAPTER 3. TECHNICAL CONSTRAINTS

3.2 SHARDING

Sharding is one of many names for dividing a virtual world into sections that are operated by different
parts of a cluster of Internet servers or, at least, different parts of a single database. Small shards
that are clearly demarked from the user’s perspective are typically called instances or dungeons. The
term instance implies there are two or more versions of the same section of one world running
simultaneously. For example, when a team of five players enters the Sunken Temple instance in
World of Warcraft, they will battle many monsters, but they will not encounter other players even
though several teams of players may be experiencing the Sunken Temple at the same time.
Some games make very heavy use of instances, such as Anarchy Online and Age of Conan,
which were launched in 2001 and 2008, respectively, by Funcom, a Norwegian company. A player
who undertakes a mission in Anarchy Online typically needs to go into a cave or a building, which is
a small instance, call it a microinstance. These are not tiny, and they may consist of one or two dozen
rooms connected by corridors and populated by enemies, but they are simpler than the typical World
of Warcraft instance. On occasion, the player will need to run away from several enemies, and they
can even run out of the instance, where the enemies cannot follow. After regaining strength, the
player can re-enter the instance and will find it about as it was before. The enemies that had been
killed remain dead, but the living ones may have moved around slightly, making it easier to conquer
them on a second attempt. Thus, Anarchy Online uses persistent instances, that are not automatically
reset when the characters leave them, as happens in other games.
Age of Conan contains instances of many different sizes and settings. Some are truly microin-
stances, shops or small houses of as little as one room – even a single prison cell – where an individual
player can have an experience that advances a storyline without affecting other players. Some are
medium-sized, such as Conan’s throne room. Other instances are very large, such as the major cities
and zones of the world, each containing dozens of independent players and also containing some
number of microinstances. In a few cases, the same medium-sized instance is presented as three
different ones with different names because they are visited by players belonging to different nations
who access it from different zones. At the moment of entering a medium-sized instance or an entire
zone, players may get the option to experience it as “normal,” in which the enemies are weak or
“epic,” in which they are strong.
One of the glories of World of Warcraft is that a high-level player can run across an entire
continent, perhaps, for a full hour, without experiencing any jolting transition from one instance to
another. As the character runs, a speck in the distance grows gracefully into a towering castle, and a
notch on the horizon smoothly becomes a pass between mountains. Yes, a character can walk across
the narrow bridge connecting Elwynn Forest with Westfall, but it is also possible to swim across the
river at any point. Going into or out of a city requires navigating through a twisting passageway, and
one suspects that the computer is furiously moving data around during those moments, but there is
no sense of disjunction. However, the heavy instancing in Age of Conan does not make it inferior, but
it merely emphasizes different qualities. Going from one zone to another always brings a temporary
 3.2. SHARDING 27
flash screen and a delay, and it can take place only at a designated portal, but this disjunction is
compensated for by the ability to select the difficulty of a zone when entering it.
In a paper chiefly about how instances permit load balancing across servers and portions of
the network, Lu et al. (2006, p. 2) note:
There are two extremes when determining how to sub-divide a virtual world for the
purposes of modeling player interaction (localized game play) and providing manageable
consistency:

• Geographic – world divided into regions at initialisation time to reflect the structure
of a virtual world.
• Behavioral – virtual world sub-divided to reflect the interaction patterns of players.

The highest level of sharding, practiced by most games, involves running multiple versions of
the entire world on different sets of servers, limiting a given character to only one. In my research
on World of Warcraft, I ran multiple characters on fully six servers, or realms as they are called. Some
of these were identical except for whatever characters were on them, and they are used to reduce
the load and thus the latency from any one of them. I have often employed the CensusPlus add-on
software to do a census of all characters currently in a given realm, finding that as many as 10,000
might visit the realm during a busy day, but no more than 4,000 would be on it at once. Indeed, on
several occasions, a server was so busy there was a waiting queue – and during the 2008 Lich King
expansion, I once saw that 1,000 characters were waiting as long as an hour to get in – and one day,
a census done right after I got in, confirmed the 4,000 maximum population.
Having multiple instances of a game, generally, means that some players cannot interact with
some others. Very commonly, popular multiplayer games place servers physically in different real-
world geographic locations, to serve players who live there with better connectivity (Beskow et al.,
2008), and even have a gametime clock that matches local time. Often, servers in different countries
employ different subscription systems as well, which means, for example, that World of Warcraft
Players in China, North America, and Europe usually cannot play with each other. On occasion,
a World of Warcraft realm will fill up and not permit any new players to enter it. At this point,
the Blizzard company that operates the game usually advertises free transfers from that realm to
specific others that are under-populated, while normally, it charges $25 to transfer a character. This
can present both opportunities and challenges to groups of players, who may find it difficult to
coordinate their decisions about moving in the short period of a few days when transfer is free.
A few games avoid this problem by dividing their world up in a way that does not separate
some players from others. A good example is EVE Online, which is set in outer space and where
travel from one solar system to another takes place through teleportation. All the players are on the
same cluster of servers, distributed by which solar systems they happen to be in at the moment,
but they are, potentially, all capable of interacting. On rare occasions, I ran into a very brief waiting
queue getting into EVE Online when the number of players currently online exceeded something
above 35,000.
28 CHAPTER 3. TECHNICAL CONSTRAINTS

3.3 GRAPHICS
This is not the place for a complete survey of computer graphics, which is a very complex field
possessing many thousands of scientific publications. Our goal is to note a few of the technical
aspects of graphics in online multiplayer games that have consequences for people’s experience of
the game and that interact in interesting ways with other topics considered here.The primary manner
in which users experience virtual worlds is visually, and a successful game must give the player a solid
sense of being in a particular place (Browning et al., 2006).
The key concept in understanding game graphics is that there are really two virtual envi-
ronments, not one (Bainbridge and Bainbridge, 2007b). First, there is the display model that the
user sees and which is the result of the graphics aspects of the software and data. Second, there is
the world model that the character experiences when moving and otherwise interacting with virtual
objects. For example, with respect to the display model, a wall is the surface the user sees, which will
appear quite differently from different angles. The world model of a wall establishes the fact that
the character cannot walk through it. Sometimes, these two models are generated by very different
parts of the program or database, and there may be considerable discrepancies between them. In
the display model, a wall may include a window plus niches and sculpture that realistically change
appearance as the character walks past, yet the world model may depict the same wall merely as a
plane that may not even be at exactly the identical apparent location as the graphically displayed
wall. Sometimes the discrepancy is so great that there is an unintended gap in the world model that
allows a character to walk through a wall and get into an area the game designers did not anticipate
would be reachable.
Depending upon how the virtual environment is programmed, both models may impose
significant demands on the user’s computer. For example, if there are several non-player enemies
inside a castle, they also should not be able to walk through the walls, even as they may pursue the
player through the twists and turns of the hallways. A common solution is to have the world model
be simpler than the display model, although there are, of course, different ways the designers can
trade off computer work between the two. Ideally, the discrepancies should be at a minimum, but
this is hard to achieve consistently.
A striking example is how Entropia Universe handles plant life, in comparison with buildings.
The plants on the planet Calypso are large, colorful, and exotic, some looking very much like terrestrial
moss magnified to hundreds of times its earthly size. However, all the original plants existed only
in the display model, and they did not exist in the world model. That is, one’s avatar could easily
walk through them, whereas the wall of a building felt solid. In August 2009, when the planet was
entirely recreated to accommodate a more advanced graphics engine, some new trees appeared solid,
existing in both models, whereas many of the old plants remained insubstantial and did not exist in
the world model.
A very difficult graphics issue that affects players is the problem of the user’s viewpoint when
the character is inside a building. Humans in the physical world do not have this problem because
our viewpoint is set by nature to be at the location of our eyes. However, users of virtual worlds,
 3.3. GRAPHICS 29
typically, set the viewpoint above and behind the character, often a considerable distance from the
eyes. This means that the point of view of the player could be obstructed by walls and other features
of the architecture. Different games handle this differently. Matrix Online had considerable trouble,
both because its rooms and corridors tended to be small, and because the graphics system would
compensate unpredictably, making the image bounce around and causing difficulty moving quickly
when the character was in a fight. Many other games have similar problems on staircases or other
narrow spaces where the character must turn along a complex path. More than a few times one of
my characters has fallen to its death because I could not tell where on a staircase or battlement it
was standing.
By definition, first-person shooter games (such as Chronicles of Riddick) make the user look
through the eyes of the character, for example, sighting along the barrel of a rifle toward the enemies
and showing only the hands of the character or sometimes a shadow. Of the examples discussed
here, Tabula Rasa was designed to be used either with a first-person shooter interface, or a role-
playing interface, complete with different key mapping schemes, but several others permit shifting
the viewpoint. Many players, especially inexperienced ones, have difficulty orienting themselves in
a first-person view because they lack many of the other sensory cues people ordinarily experience in
the real world, such as a physical sense of orientation and movement, and the peripheral vision that
often lets us know where our comrades or the exits are. Therefore, most multiplayer games encourage
a third-person view, which has the advantage of giving the person more warning when an enemy
attacks from behind. To compensate, rooms inside buildings are made unnaturally large, just as the
rooms in the houses depicted in television situation comedies have exaggerated dimensions.
Expert players may become accustomed to both first-person and third-person viewpoints, able
to switch back and forth, but games differ enough from each other that transferring this skill from
one to another is far from easy. Among the more spectacular skills in World of Warcraft, although
seldom useful in combat, is being able to run along the horizontal ropes of the numerous sailing
ships. The player goes into first-person view, places the rope directly in the center of the view, and
runs forward, reliably keeping the character’s feet on the rope without, however, being able to see
them.
Given the complexity of the virtual worlds and the needs of advanced players to handle
that complexity, new players often find learning how to use the graphic user interface a daunting
experience (Cornett, S., 2004). I was about 300 hours into World of Warcraft before I “found my zone”
when I could handle the game’s excellent interface intuitively, and every new game presents a new
learning curve that must be climbed. A survey carried out in ten European nations of people who
played four different kinds of online games showed that, indeed, there were a substantial number
of “hardcore” gamers who invested from ten to over thirty hours a week, with role playing games at
the high end of the distribution (Fritsch et al., 2006, 2007). Yet every hardcore gamer was once a
newbie, and the population of players grows only by attracting people who initially are strangers in
a strange land.
30 CHAPTER 3. TECHNICAL CONSTRAINTS
How the graphics displayed on the screen integrate with other modalities has significant im-
plications for user experience. For example, a group conversation in the real world is facilitated by
the fact that people can see who is talking, without needing first to learn how to recognize each per-
son’s voice. As Halloran et al. (2004) have shown, the absence of visual cues can impede cooperation
when a voice channel is used in online games, and adding a visual cue can be advantageous, as was
done in the racing game Midtown Madness. Moore et al. (2007) have analyzed communications in
several virtual worlds – including EverQuest, Star Wars Galaxies, and World of Warcraft. In addition
to obvious details like the difficulty of making a character point or look in the desired direction,
they noted more subtle issues such as the fact that online game text chat messages are not sent until
they are finished, whereas a verbal communication in the real world is heard and responded to long
before it is complete.
I have organized formal group meetings in Second Life, and a green indicator appears over the
head of the avatar of the speaker. This has the added advantage of indicating whose microphone
was inadvertently left open after speaking (responsible for annoying feedback of what somebody
else is saying, for those whose sound systems do not have good methods for preventing feedback).
Visual cues of who is speaking require that the voice system be integrated into the game, but even
after voice was added to World of Warcraft, many players preferred to talk over separate systems using
Skype, TeamSpeak, or Ventrillo.
Similarly, non-vocal sound effects supplement visual cues, but they are less well integrated
with them than in the real world. Jørgensen, K. (2008) studied how sound effects functioned during
group combat in World of Warcraft, especially noting the ambiguities of the meaning of sounds in
complex environments. When a character goes into invisible stealth mode, a distinctive sound is
heard, but it does not tell the player whether an enemy or friend has just stealthed, and thus whether
to worry or be reassured. Similarly, the sound when one character casts an enhancing buff on a
friend may have different meanings depending upon the context: “It may be a responsive signal that
confirms an avatar’s casting of the buff; it may identify a change in avatar state; and it may orient
that avatar with respect to other avatars.” Jørgensen also documents in some detail how the sound
effects successfully add information to what is seen on the screen, for example, confirming whether
a magical spell worked or not.
A very important dimension of the virtual environment is its architecture. The more advanced
the graphics technology is, the more complex the buildings can be. Actually, there are two kinds of
complexity. First, there are the graphic textures displayed on surfaces, such as the grainy appearance
of individual stones in the wall of a castle. In cases where a game demands much from the user’s
computer, but the user’s computer is only barely up to the task, the textures of somewhat distant
objects may often be degraded, and they show their full granularity only from close virtual distances.
Second, there is the complexity of structures experienced as the user interacts with them, for example
climbing the stairs inside a castle and entering its rooms. Although graphics constrains the gameplay,
the reverse is also true because good graphics must be designed to fit the nature of the game.
 3.3. GRAPHICS 31
McGregor, G. (2006) offered evidence on this point in a study that compared the role-playing
game World of Warcraft with the real-time strategy game Lord of the Rings: Battle for Middle Earth
2. Multiplayer role-playing games typically represent the user as a single avatar, through which the
user experiences the world, requiring that the avatar be able to go inside buildings, caverns, and even
swim through water. In a strategy game, the user often manipulates many characters at once and,
even when working with a single character, operates it at a distance, emotionally as well as visually.
Indeed, strategy games are sometimes called God games because the player operates from a lofty,
almost Olympian perspective. Thus, it is not surprising that the architectural structures in Lord of the
Rings: Battle for Middle Earth 2 cannot be entered, although McGregor complains that this greatly
reduces their aesthetic value. McGregor also notes that the game’s large complexes of buildings often
seem very strange, compared with World of Warcraft because they are devoid of people. Nothing quite
compares with the visual excitement of the trade district of Stormwind City in World of Warcraft,
where there are often fifty avatars rushing around in different directions, in addition to a dozen
non-player characters.
Perhaps, the ultimate in realism – or in paradox – in several of the games are the bathrooms.
I don’t happen to know of any examples in which the characters actually need to relieve themselves
in bathrooms, although they often need to eat food and drink liquids, but the bathrooms in Matrix
Online were actually useful. Figure 3.1 shows my character, Cosmic Engineer, in the men’s room
on an upper floor of an office building, which he reached by riding in an elevator. His mission was
to kill research subjects in a vile experiment, which accelerated the speed of the algorithms that
operated them to an alarming degree, but none of them were found going to the bathroom. Indeed,
the bathrooms tended to be private places where one could escape enemies. Often, Cosmic Engineer
would battle too many enemies at once to defeat them all, so he would kill one, then run away to
regain his lost health. A restroom was a good place to rest.
Notice that the bathroom in the picture really is a men’s room; the ladies’ rooms do not have
stand-up urinals. When a character needs to rest and recover health, sitting down speeds up the
process, and sitting on a toilet serves perfectly well. Indeed, intrepid ethnographic researcher that
he was, Cosmic Engineer tried this in ladies’ rooms as well, and found they did not discriminate
against him on the basis of his virtual gender. Some other games have bathrooms, such as Anarchy
Online and Tabula Rasa. Given that Tabula Rasa is a science fiction game, it was interesting to see
that future technology had dispensed with toilet paper. Some of the toilets could be flushed, and
when they did the water gushed upward, presumably to clean off the astronaut’s posterior.
Around the edges of Figure 3.1 are the elements of the graphic user interface, which is complex
in each of the popular games, but it is different in detail from one to another. At the top center are
two bars of icons used to control the character during battle, with the option of scrolling to other bars
that have the skill controls he uses when crafting virtual goods. In the upper right corner is the display
of his status, which as in many other games is represented by a colored bar graph, here with three
bars. The top (red bar) represents health which warns when death approaches. The second (yellow)
bar represents the “inner strength” used to power many actions. The third (blue) bar represents the
32 CHAPTER 3. TECHNICAL CONSTRAINTS

Figure 3.1: An Office Building Men’s Room in The Matrix Online.

growth of experience indicating one tenth of a level with blue dots below it to indicate how many
tenths had been achieved. The number 34 in this display box indicates that Cosmic Engineer has
reached level 34 of 50 at this point in his career.
In the lower right corner, Cosmic Engineer has set his map, which could be changed in shape
to meet his momentary needs, and there is also an atlas of the entire city he could consult. The map
tells him he is in the Vauxton neighborhood of the high-status Downtown part of the city, an area
undergoing urban renewal. The atlas and other Matrix lore tell him that the residents of this section
have an inferiority complex with respect to the very highest-status districts, like Creston Heights
and Stratford Campus, and they scorn the merely middle-class area they call South Vauxton.
Architecture expresses social status, as the atlas recognizes: “Vauxton is usually called ‘The
Vox’ by local residents, and anyone who calls it Vauxton is immediately tagged as an outsider.”
 3.3. GRAPHICS 33
“Unbeknownst to the South Vauxton residents struggling to get the acknowledgement they crave,
the move to separate South Vauxton from Vauxton is secretly financed and backed by the business
leaders of Vauxton who’d rather be rid of the undesirable south end.”
On a more practical level, the map tells Cosmic Engineer where he is in the city, and he
frequently consults the map, the atlas, and external online sources when he is seeking a resource or a
specific destination. The display tells him his current coordinates are: X: -954 Y: 225 Z: -339. The
X coordinate is how far east of the center of the Downtown he is, and it is negative, so he is on
the west side. The Z coordinate is how far north he is, and it is negative, so he is slightly south of
the center of the district. The Y coordinate is how high about ground level he is, and it is positive
because he is in a tall office building. Each game has a different coordinate system, so switching
from one to another can be confusing. For example, in Dark Age of Camelot the numbers can reach
50,000, because each unit equals an inch, and the numbers start at 0,0 in the northwest corner of
the zone and increase eastward and southward, rather than being measured from the center as here.
The image in the map shows Cosmic Engineer where he is on the particular floor of the
building, in a small room off a hallway not far from the elevator. Elevators are another good place
to hide from enemies, so it is wise to keep conscious of where they are. The map does not show
where the doorways are, but two different shades of red distinguish rooms he has already visited, and
perhaps cleared of enemies, from those he has not yet entered. Icons sometimes locate mission goals,
likely enemy locations, and locked doors for which keys must be found. Outdoors, such icons indicate
mission contacts, parks, monuments, nightclubs, vendors, and the hardline telephone booths used
for teleportation. The bottom center of the screen shows a compass that tells him which direction
he is facing, flanked by buttons that can open up many option choices such as the atlas.
To the left of center, on the bottom, is the control panel he uses for his simulacrum, a secondary
avatar not shown in this picture because he is another room resting after killing an enemy. In this case,
the secondary avatar is a submachine gun bodyguard who helps fight enemies, but whom Cosmic
Engineer is quite ready to sacrifice when his own life is in danger.
The lower left-hand corner of the display normally holds the multi-facetted text chat interface,
but at the moment, it shows only the immediate mission objective: “Kill all of the test subjects.”
At certain points in the mission, a communication window opens at the lower center of the screen,
giving the player orders and information, and when working with a team of other players, the text
chat area can become very cluttered. The faint rectangle at the upper left corner stands where the
status of an enemy would appear during a fight, but, currently, it merely contains the word “door,”
because the last thing Cosmic Engineer interacted with was the door to the men’s room.
This picture of a bathroom illustrates the complex relationship between the virtual world
and real world, including the great complexity of a graphic user interface that must be learned well
before it can be used well. In the three Matrix movies, users jacked into the system through a
direct brain interface at the base of their skulls, and they were not conscious that the world they
interacted with was inside a computer. Many researchers still hope to develop fully immersive virtual
reality in which the user acts in the normal manner by moving legs and arms and hands and fingers.
34 CHAPTER 3. TECHNICAL CONSTRAINTS
But at the present time, cumbersome user interfaces interact with increasingly realistic but still
obviously artificial graphic virtual worlds. Much of the psychological feeling of immersion comes
from emotional engagement with the lore, the quests, and the online social groups. At the same time,
the graphic display, including the user interface, is part of a system that controls human behavior
even as it liberates human fantasies.
 35

CHAPTER 4

Rolecoding and Social Control

Rolecoding is connected to security, privacy, access control, group membership and role playing. It
refers to rules built into the game, rather than enforced through persuasion or punishment, that
control the ways in which particular characters can interact with each other. The social roles of
online game characters, and thus of their players, are sharply defined in many ways. For example, my
Anarchy Online character, Nanobic, is an engineer working for the Omni-Tek Corporation. Figure 4.1
shows him sharing Christmas with his robot friend, Tobor.

Figure 4.1: Nanobic and Tobor Celebrating in Anarchy Online.

Most other kinds of AO characters cannot create and control robots, and most games described
here strictly impose a division of labor upon characters and thus upon players. One could argue,
36 CHAPTER 4. ROLECODING AND SOCIAL CONTROL
without too much exaggeration, that the rules of the game constrain Nanobic as much as Tobor and
thus that the player is transformed into a kind of robot as well. The player’s chief options are which
game to subscribe to, and what kind of character to select from a pre-defined set.

4.1 SYSTEMS OF RULES

Given that much of the action in online games involves “killing” other characters, some of whom
are operated by other players, and stealing virtual goods from them, it is hard to say what the basis
of morality could be. Some rules are written into the user agreements that every player must accept,
others are programmed into the code as firm technical barriers against certain kinds of behavior,
and others are informally adopted by groups of players within these virtual worlds. What are we to
make, both technically and ethically of the following case described by Craft, A. (2007, p. 205)?

On April 18, 2005 Istvaan Shogaatsu, the leader of a group of mercenaries called the
Guiding Hand Social Club in EVE Online, announced that his organization had just
completed one of the biggest acts of theft and betrayal in the history of virtual worlds.
Members of the Guiding Hand spent a year infiltrating a rival organization before assas-
sinating their leader and stealing in-game assets valued at 16,500 US dollars, effectively
shattering the trust within the organization’s social network setting its members back
months of playing time.

Several games make a distinction between servers where player-versus-player (PvP) combat
is encouraged, and those where both players must agree to have a duel before one can attack the
other. Thus, the populations of players differ across them, with more aggressive players choosing
the PvP realms. In EVE Online, which emphasizes PvP play for advanced players, much of the
galaxy is wide open, but a few hundred solar systems have a security force which deters one player
from attacking another. It is very common for PvP games to suppress PvP in the areas for starting
players or otherwise vary the constraints from one area to another. A few World of Warcraft realms are
designated for role playing (RP), suggesting but not requiring that players stay in character. I even
saw one example in which a guild of players had two text chat channels, one for RP communications
and one for talk about the players’ “real world” lives.
Building the rules into the game, so that it is impossible to violate them, is a cost-effective
solution from the standpoint of game designers. For example, in many games when one player agrees
to buy something from another, a trade window opens up, and the exchange does not complete until
both players click buttons indicating they agree. Thus, there is no possibility for one player to steal
from another. However, if one player agrees to perform a service for another, there may be no practical
way to build a safety mechanism into the code of the game, to ensure that the money does not get
paid until the service has been performed, and that the money will indeed be paid after satisfactory
service.
The typical solution to this problem of trust is to build into the game a system for managing
groups of players, like the guilds in World of Warcraft. Each guild is founded by a guild master,
 4.1. SYSTEMS OF RULES 37
who can set the rules of access to the guild management system by members of different ranks –
for example, specifying which categories of members can recruit new members and expel existing
members. If one member abuses the trust of others, the victims can talk with an officer of the guild,
who may admonish the troublesome member, or may even boot him or her from the guild. Most
recent games that have such player organizations incorporate rather complex management systems
that allow the leaders of the guild to make a range of decisions about such things as how many ranks
their group will have and what privileges each rank will enjoy.
A guild master in World of Warcraft, as I learned with the three guilds I myself ran, has a guild
control window in the user interface where such rules can be set. Several relate to communication
permissions. Can members of the given rank speak and/or listen in the guild chat and in a chat
reserved for officers of the guild? Do they have the power to set the guild’s “message of the day”
that appears in the text chat whenever a player logs into the character? In the guild part of the social
module in the user interface, any member can see a list of the guild members, including whether
or not they are currently online and what zone of the virtual world they are in. Also accessible is a
window displaying any information a fellow guild member wants to post about the character, called
a public note. Officers of the guild can also place an officer note there, which may not be visible to the
player who owns the character, and might be critical of the player. Thus, three permissions the guild
master may give high ranking members is editing a player’s public note and seeing or editing the
officer note.
Of course, important powers are inviting new people to join the guild, removing inactive or
troublesome members from the guild, and promoting or demoting up and down the ladder of ranks.
Many guilds allow all but the very lowest, provisional rank of membership is to recruit new members,
but reserve promotions, demotions and removals are for officers to handle. Often, one sees automatic
announcements of these changes in the guild chat. When a new person joins, that player gains access
to the guild chat, which announces by name the character has just joined. The members who are
online exclaim “welcome,” and some immediately offer advice to the newcomer. Often, a barrage of
messages announces a series of promotions, and people share congratulations. At the middle ranks,
promotions tend to depend simply upon being an active member and not causing trouble. The guild
part of the interface provides information on when a character was last online, and some guilds
regularly demote those who have not played within the past month, and expel members who have
been absent for a long time, although practices in this area vary widely. Expulsions and voluntary
defections by members are also automatically announced in the chat, and they frequently provoke
discussion.
Some of the largest or best-organized guilds conduct a fair amount of their business on
websites and through private forums that require a player to register and seek permissions, and these
are also used to screen new recruits. Thus, part of the rolecoding takes place outside the game itself,
in these web-based social computing sites operated by guilds or comparable groups associated with
one game or another.
38 CHAPTER 4. ROLECODING AND SOCIAL CONTROL
The World of Warcraft guild-master module also provides control over the guild bank or vault,
which itself is accessed only at banks in major cities in the virtual world. There, characters may
contribute virtual money or goods of various kinds, such as armor they do not need themselves, or
useful virtual objects they have crafted using specialized professional skills. The guild master can
decide whether a given rank has any access to the guild bank at all, and only those with permission
can donate items or even see how much is already there. Some guilds allow only officers to withdraw
items or money, and this gives them a nice social function in distributing those valuables among
needy members. But the guild master can follow any policy in giving a rank the ability to withdraw,
including setting limits on how much can be taken in a single day.

4.2 DEVIANT BEHAVIOR

If some rules are programmed into the game mechanics, while others are a matter of customary
behavior of players, what are the boundaries of the concept cheating, and how can different forms
of it be controlled? Perhaps the most blatant form involves using special software or hardware to
change the data of a game, to favor the user. This is common with solo-player games, and many
people use GameShark hardware or software, feeling it is not cheating because in solo play no other
player is put at a disadvantage by the use of GameShark. Of course, once people share the results of
their solo play – “I just finished Game X in record time!” – a competitive element enters the ethical
equation.
Online multi-player games typically prevent this kind of cheating by keeping a secure cen-
tralized database containing all competition-relevant statistics for each player. To reduce latency and
cost, game developers might want to move to a peer-to-peer system, and, indeed, peer-to-peer is
often used for handling updates of game content, but not of actual gameplay. With the goal of ulti-
mately designing cheat-proof peer-to-peer systems, Webb and Soh (2007); cf. Li et al. (2004) drew
upon earlier researchers to develop a typology of cheats. Peer-to-peer systems are especially vulner-
able because many of these cheats involve transmitting wrong information that may have significant
impact upon competition even if the central controller of the peer-to-peer network seeks periodically
to synchronize information across all players. Invalid commands can be sent or inconsistency across
players is created by sending different updates to different people. Suppressed update, timestamp,
and replay cheats alter the time at which information is sent. In spoofing, one player masquerades as
another. Webb and Soh define five other kinds of cheats that are more subtle and can be carried out
with central server games that lack the vulnerabilities of peer-to-peer, so I shall expand upon their
analysis of those types here.
Exploit is the common gaming term for exploiting a bug or other weakness in a game, and a
few of the games actually prohibit exploits in their user agreements. Many advanced players, however,
consider exploits to be the highest form of play, deserving more honor than mere obedience to the
rules. The most powerful expression of this viewpoint was offered by Robinett, W. (2003) who did
essentially this when he programmed the very first easter egg into the computer game Adventure
way back in 1978. Easter egg is the term for a valuable item that is hidden somewhere in the virtual
 4.2. DEVIANT BEHAVIOR 39
environment and can be used to do something unusual. The game company Robinett was working
for, Atari, did not want its game creator’s names to be publicized, but he felt he deserved recognition
for this pioneering effort. Therefore, he placed a gray dot at one point in a maze, which was the key
to a secret room accessible at another point, in which “Warren Robinett” would flash on the screen
in rainbow colors. As game designer, Robinett was playing a metagame with the company and with
the ordinary players. Exploits are a way in which players can win a metagame against the companies
and designers as well as the ordinary players who lack the knowledge or the motivation to take the
game to a higher level.
An example of a type of exploit requiring two or more players was offered by Wright et al.
(2002), who studied the early online first-person shooter game, Counter-Strike. Typically, two teams
of commandos are fighting in a complex but well-defined territory. Characters who are killed do not
immediately come back to life, so their players are allowed to observe events but not participate in
them, even shifting the location from which their ghosts were observing.To prevent dead players from
functioning as spies, they are blocked from using the usual in-game chat communications system.
However, some players discovered they could still communicate through a subsidiary channel, so
they could scout out the enemy deployment and report back to the still-living members of their
team. This exploit is called ghosting, and is considered here as an exploit rather than placing it in the
collusion category (below), because it depends upon a technical vulnerability of the system, but it
also could be described as an espionage hack.
Espionage hacks are actions based on special technical knowledge that give the player informa-
tion he is not entitled to have. As Cikic et al. (2008) note, “Quite a few cheats actually revolve around
knowing things that the player is not supposed to know (yet). Famous examples are the well-known
maphacks and wallhacks, where the player sees what happens in areas he cannot currently see, or has
not ever seen. These cheats are some of the hardest to detect, prevent, and prosecute, as even a total
replay of the situation, with all possible data available, will often not prove that the player cheated,
he may just have guessed well.” Although some of these cases involve running software that lets the
player see from a different location than the game itself would provide, many of the ones that are
possible today are exploits more than hacks.
The general response from game developers is to level the playing field, sometimes achieving an
advantage for themselves in so doing, and the response from players has been to share the knowledge
so widely that it ceases to confer an advantage. For example, in World of Warcraft, a player cannot
see sections of a map of the territory unless that player’s character has already visited those sections.
The map starts largely blank, and it fills in only step by step as the character legitimately acquires
information about each section. Typically, the full map is already on the player’s computer, hidden
among the graphics files, and could be located and decoded, given sufficient computing skill, what is
called a maphack. However, once players had visited all parts of a map, they could take a screenshot
of it and post it on their websites, so every new player would have access. Blizzard Entertainment,
the company that produces World of Warcraft, did them one better, printing all the maps in a series
40 CHAPTER 4. ROLECODING AND SOCIAL CONTROL
of atlases, along with additional charming drawings of each and every small town in its vast virtual
world.
Wallhacks are exploits that allow a player to see through a supposedly opaque wall – or to
reach through a supposedly solid wall – and often depend on a local flaw in the game’s database.
For example, when I first started playing World of Warcraft, wallhacks could be done on two sunken
ships. To legitimately reach the objectives in these two ships, a character had to swim down into
the water and battle obnoxious amphibians called murlocs, all the while holding a virtual breath.
Several dives were necessary, and if the character were “killed,” there would be time for the dead
murlocs to respawn so the battle would start all over from the beginning. However, it was possible
to swim down outside each ship, to a certain point in the stern, and grab the quest object through
the solid wall. After a few months, this vulnerability was closed, and the wallhack no longer worked.
As with many other kinds of exploit, the ultimate cure is adding preventive measures to the game
code, rather than punishing offenders.
Given that cooperation among players is an important part of many online games, they of-
fer a range of groupware technologies to manage raiding parties, long-lived guilds, and permanent
category divisions such as combat factions. What Webb and Soh (2007) call collusion is cooperation
between players who are nominally in competing groups. I reported a remarkable sequence of collu-
sion and counter-collusion in my book about World of Warcraft (Bainbridge, W., 2010c, p. 136–137)
It is possible for a player to have both Horde and Alliance characters in the same realm, as I myself
did, and the following incident was observed by Etacarinae from the Alliance side, and Incognita
from the Horde side. The only collusion between them concerned placing them at the right spots
to collect data, but I could just as easily have used collusion between them to influence events.
Etacarinae was going about some ordinary business in Stormwind, the Human city, when
a rumor came over the local Alliance text chat that the Horde was about to attack. Presumably, a
player was using a Horde character as a spy and passing the information to the Alliance through
a second Alliance character. Etacarinae and several others ran to the front gate of the city, ready
to block entry. They grew restless as the minutes passed without incident; then, a second message
said that the Horde raiding party was crossing the bridge from Westfall. Thinking to cut off the
Horde advance at that point, the defenders rushed to the bridge but saw no invaders. Rushing back
to Stormwind, they discovered they had been tricked, and the Horde raiders had penetrated deep
into the city before being killed. Probably, it had all been a trick. A player loyal to the Horde had
used an Alliance character to lure the Alliance players first to the main gate and then over to the
bridge, leaving the entrance undefended.
When the attack was over, Etacarinae’s guild began organizing a retaliation. Because guild
members have some basis of trust, it is relatively safe for them to plan attacks using the text chat that
is reserved for members.The player who tricked the Stormwind defenders did not need to belong to a
guild and, indeed, could have created an Alliance character only minutes before for the sole purpose of
placing disinformation in the local Alliance chat. The guild decided to attack Orgrimmar, one of the
Horde cities, which required long distance travel. Travel takes time, and it reveals intentions, because
 4.3. GAME MASTERS AND MENTORS 41
one of the non-guild Alliance characters could have been the Horde spy. I switched over to my Horde
character, Incognita, and I headed for Orgrimmar. Word came over the Horde chat that the Alliance
was planning an attack on a different Horde city, Silvermoon. This, I knew, was disinformation. The
obvious attack point on Orgrimmar is the main gate on the south side, but the Alliance planned to
attack the west gate, instead, so that was where I placed Incognita. This retaliatory attack was less
successful than the original, but both included disinformation communicated by means of collusion
between characters nominally belonging to opposed factions. On several other occasions when one
of my characters participated in raiding by a well-organized guild, the leadership did not announce
the target until they were very close, and they chose routes that disguised their destination.
Cheating is not the only crime in virtual worlds, and among the most prominent other
offenses is griefing, taking pleasure from causing harm to other players. Foo and Koivisto (2004)
have suggested there are four distinguishable types of griefing. Harassment annoys other players
through shouting slurs in the text or voice chat, spamming a channel repeatedly, placing one’s avatar
in the way of the other player’s avatar, and disrupting events. Power imposition is more violent,
often consisting of repeatedly killing a weaker player. Scamming involves such misdeeds as breaking
promises and deceiving other players about one’s identity. Greed play is like stealing, for example,
ninja looting by taking the loot from a dead non-player character that rightfully belongs to another
player, or kill stealing that joins a fight at the last moment hoping to get credit for the kill.
Rolecoding systems in some games effectively prevent some but not all of these behaviors. For
example, many text chat systems have an ignore command that forever prevents a spammer from
sending you a message, and many games assign loot and kill credit to the first player who attacked
the enemy, thereby preventing ninja looting and kill stealing. Rolecoding affects groups as well as
individuals. For example, team quests and instances are designed for a certain number of players,
require the exercise of a particular combination of skills for success, and provide standardized methods
of communication between players while they are on the team (Bardzell et al., 2008; Chen et al.,
2008).
Barnett et al. (2010) carried out a questionnaire study to determine what made players angry
at other players in World of Warcraft. Griefing and harassment were among the four factors they
found, but two others that proved very important have not been studied extensively in previous
work. Most anger-causing scenarios involve the failure of one player to live up to the expectations
of another, notably in raids or instances where teamwork is essential, and other scenarios concern
simple time-wasting. Interestingly, feeling angry did not depend on believing that the other player
intentionally misbehaved. This may suggest a blind area in prior research, because while there has
been much discussion of intentional norm-violation, there has been little about incompetence.

4.3 GAME MASTERS AND MENTORS

Online roleplaying games are an outgrowth of old-style pencil and paper roleplaying games, including
tabletop games using cards or dice exemplified by Dungeons and Dragons released in 1974. Even
earlier, the 1949 board game Cluedo (called Clue in North America) had players take different roles
42 CHAPTER 4. ROLECODING AND SOCIAL CONTROL
in a murder mystery. The rules of Cluedo were unambiguous, and each time the game was played one
of the characters selected by the drawing of a card would be secretly assigned guilt for the murder.
Thus, as with other popular board games, any reasonable set of people could understand and follow
the written rules. Dungeons and Dragons was complex, and it permitted considerable elaborations, so
it required one person to take the role of game master. As the current Wikipedia article for the game
says, “A Dungeon Master serves as the game’s referee and storyteller, while also maintaining the
setting in which the adventures occur”1 . When roleplaying games were created online, many of the
former functions of the game master were encoded into the software, but a few functions remained
that human beings had to perform.
With the role of game master in mind, Tychsen, A. (2006); Tychsen et al. (2005) says that
role-playing games have five common features. First, these games are “storytelling with rules,”
which implies that someone or something must perform the storytelling function. Second, the
players operate characters inside the fictional reality, through which they interact. Third, each par-
ticular game reality rests upon a unique premise: “This is a shared understanding among the game
participants of the game setting, the starting point of the game, the rules and similar framework
properties” (Tychsen, A., 2006, p. 76). Fourth, these games are usually guided either by a human
game master, or, in the case of many electronic games, by the game engine. Fifth, role-playing games
like those considered here require at least two people – often many – one of whom may be the game
master.
One distinction across games of many kinds is whether the game master can also be a player.
The banker in Monopoly, for example, is in no position to cheat, so being banker does not confer
an advantage that prevents the banker from also playing. Massively multiplayer online games would
probably never be profitable if all the traditional game master functions were performed by employees
of the game company, but not all of these functions can be automated. Thus, some game master
functions may be given to the players, while a few are reserved for employees.
One of the functions of the traditional game master is educational, teaching new players and
guiding more experienced players to achieve a satisfying play experience. MMORPGs, generally, do
this automatically, by establishing special areas of the virtual world for newcomers and giving them a
set of beginner missions that teach them about both the game mechanics and the fictional premise.
Both Tale in the Desert and Age of Conan place newcomers on an island, which they cannot leave
until they have completed their training. In the case of Tale in the Desert, the training emphasizes
technology, for example, how to make stone blades from slate and use them to cut boards from
wood taken from the trees. In Age of Conan, the player learns the mythos about how Conan’s arch
enemy, the Stygian sorcerer Thoth-Amon, is preparing a magical army to attack Aquilonia, and he
learns how civilization is threatened by wild tribes along all of its borders as this major war brews.
Many games use delivery quests to teach players about the virtual world’s geography, even at very
advanced experience levels, sending the player by land, sea or air to distant locations, supposedly

1 http://en.wikipedia.org/wiki/Dungeons_and_Dragons
 4.3. GAME MASTERS AND MENTORS 43
to give supplies or information to a remote ally, but, actually, to make the player familiar with the
destination.
Entropia Universe handles training very differently. It offers no explicit quests or missions
although the rich environment of the planet Calypso suggests many reasons a player might want to
explore. Rather it urges each new player to select a human mentor, a role taken by more experienced
players, who will help the newbie identify personal goals and the means to achieve them. When
I entered Entropia Universe, I decided that my research would work better if I did not seek a
mentor inside the world, and I used a variety of web-based sources instead. However, one of those
resources was a very fine book-length guide written by one of the mentors, using the name Alice in
Wonderland2 .
A second function of game masters, which overlaps the first, is to help players solve problems.
In rare cases, these may involve data corruption or bugs where only an employee of the game
company has the power to help. Entropia Universe provides a fascinating example of a serious
problem experienced by some players and solved by others, which I myself experienced. I had set
myself the goal of reaching all of the teleporters which provide easy transit from place to place,
but as with many games, each one must be reached on foot before it is added to the player’s list
of useable teleporters. With just four left to go, I got totally stuck at a resurrection station in an
advanced zone, which was surrounded by vicious monsters. I tried many times to sneak out, but I
was repeatedly killed and resurrected back at the station. I sought help from the Calypso Rescue
Team, an organization of helpful players, and two of them quickly came to lead me out, one killing
the monsters and the other healing him from their damage3 .
Age of Conan has an organization, called the Followers of Asura, who are player volunteers
who help other players and also assist the development team find and fix bugs. To add drama to
their work, they are described as a mystical secret society: “The history of the Followers is shrouded
in mystery. Little is known about their actions or motivations, but tales of their knowledge and
assistance are legend. The Followers have come to the aid of many a Hyborian in times of need, even
King Conan himself ”4 . One of the more established medieval fantasies, Dark Age of Camelot, has a
similar program in which volunteers work with Mythic company employees to share game master
duties:

The goal of the Knights of the Round Table program is to harness the knowledge and
experience of the customer base to assist the developers in ensuring that Camelot remains
stable, balanced, and fun to play. The Knights are players who volunteer their free time
to assist the Dark Age of Camelot development team by acting as a conduit between
Mythic and the larger player community, providing feedback and testing of upcoming
changes, and suggesting new ideas on ways to improve Camelot. Each Knight has an

2 http://rp.apachenet.de/downloads/Entropia_Guide.pdf
3 http://v2.euforces.com/
4 http://cs.ageofconan.com/Public/
44 CHAPTER 4. ROLECODING AND SOCIAL CONTROL
area of responsibility. Some Knights are advocates for their realm while others speak on
behalf of the citizens of all realms and the game, at large5 .

4.4 LEGAL REGIME

With EverQuest as his example, Lastowka, G. (2009) suggests that the legal regulation of an online
game has three distinct frameworks. First, as fictive text a game is a work of art, belonging to the
company that created it, and regulated by intellectual property law. This categorization is somewhat
problematic, because it leaves open the question of to what extent the players are also creators, and
where the line should be drawn between the intellectual property of the company and that of the
players. Second, as a computer game EverQuest has rules, and many of these rules are built into
the programming codes so that players cannot violate them. This contrasts with ordinary games, in
which the rules are enforced by the players themselves or in some cases by referees, and raises the
question of how online games would define and enforce rules if they were not built into the code.
Third, as an online community, EverQuest could have norms above and beyond those written into
the code, but this raises two immediate questions: is the EverQuest community distinct from the
wider society and thus able to have its own rules? Is the EverQuest community unified, or is it really
a set of separate communities that overlap and interact as social groups do within the game?
On Lastowka’s first point, Humphreys, S. (2009) argues that the current laws and contracts
give all the power to the companies hosting games and virtual worlds, without recognizing the
contributions made by the players. Users cannot negotiate the user agreements with the companies
and are forced to accept complex rules they may not understand. Typically, these agreements give the
company the right to expel a user without granting any appeal, despite the fact that user may have
invested much time advancing their characters and creating social relations. Indeed, the current legal
regime does not recognize the affective and social contributions of the players, and it conceptualizes
the game in an old-fashioned manner equivalent to fully-scripted works of art like novels. Humphreys
suggests that hosting an online community is much more like a service than a product, and thus the
wrong traditional rules are being applied.
Grimes, S. (2006) has not only laid out the intellectual property issues for games that are jointly
created by companies and players, but she has argued that American law is moving rapidly toward
corporate rather than community ownership of culture. She cites some specific court cases, including
one involving a small Mexican gold farming sweatshop, where employees played Dark Age of Camelot
to earn virtual currency to sell to ordinary players, something that would not seem to violate society’s
laws, unless society is required to enforce the user agreements games require. Dibbell, J. (2007) has
explored the world of Chinese gold farmers, for whom playing World of Warcraft is a livelihood as
well as occasionally fun. Other examples include EverQuest’s campaign to stop people from selling
virtual property on eBay. If people really “own” the items they earn in the games, then of course
they should be able to sell them, but the companies argue they own the items on the basis of their
intellectual property rights and user agreements. A key issue is political. If people can advance in the
5 http://www.camelotherald.com/knights/
 4.4. LEGAL REGIME 45
games by paying gold farmers or hiring others to advance their characters for them, then economic
inequalities from the outside world infect the game, and players are no longer rewarded in terms of
their skill and effort but how big their real-world bank accounts are.
 47

CHAPTER 5

Personality and Motivation

Players bring different styles, goals, habits, values, and temperaments to their gaming; their person-
alities influence their behavior and thus the experiences of other players. In addition, the games are
filled with non-player characters who must behave in an interesting and somewhat realistic manner;
they need personalities as well in the increasingly sophisticated social environments of online games.

5.1 PSYCHOLOGICAL THEORIES AND TYPOLOGIES

Traditionally, personality is conceptualized in psychological terms as a persistent predisposition to
feel and behave in ways characteristic for the individual (Lindzey et al., 1968). Throughout human
history, people have described each other in terms like hot and cold, dominant and submissive, kind
and cruel. Early in the twentieth century, psychologists began to systematize these folk concepts.
An influential research method using questionnaires was pioneered by Raymond Cattell, although
he himself believed that observational and experimental data were also needed. In an iterative
process extending across many studies, researchers would develop a long list of questionnaire items,
for example, phrases describing a person in ordinary language that expressed a range of concepts
about personality. People would rate these items in terms of how well or poorly each one described
themselves, or they described some particular other person they knew well.Then statistical techniques
like factor analysis would be used to identify which items cluster together along what number of
general dimensions of meaning. Through such methods, Cattell came up with fully 16 personality
dimensions, although subsequent researchers have felt that this is too great a number and have not
been able to confirm much of his research using more modern variants of his methods.
The most widely cited personality classification system is the so-called “Big-5” dimensions,
which exist in somewhat different variants. The so-called OCEAN version (McCrae and Costa,
1996, p. 67) is:

1. Openness to experience: fantasy, aesthetics, feelings.

2. Conscientiousness: competence, order, dutifulness.

3. Extraversion: warmth, gregariousness, assertiveness.

4. Agreeableness: trust, straightforwardness, altruism.

5. Neuroticism: anxiety, angry hostility, depression.

This system is solidly rooted in self-report psychological assessments studies but also supported
by observational studies of behavior.To be sure, there are questions about the exact meanings of these
48 CHAPTER 5. PERSONALITY AND MOTIVATION
five supposedly distinct dimensions, notably about “openness to experience,” which is sometimes
equated with “intelligence,” and cross-cultural studies show local variations as well as a degree of
commonality (Yang and Bond, 1990).
Arguably the five dimensions should have implications for the behavior of a game player.
Extraversion implies sociability, and characters may differ in the degree that they participate in
guilds and other groups, versus solo questing in the virtual world. Both conscientiousness and agree-
ableness may be valuable for cooperation with other players to achieve shared goals, and good
co-workers would logically score high on these characteristics. These two dimensions also have
also been connected to control of impulses that might otherwise express themselves through vio-
lence (Jensen-Campbell et al., 2007; Jensen-Campbell and Malcolm, 2007), and collective violence
is a major form of cooperation in games. Neuroticism might imply a disability in cooperating reliably
over time with others, whereas openness to experience could encourage exploration and discovery
in this virtual world as well as in the “real world.”
Despite its prominence, the Big-5 model has limitations. Largely based on self-report ques-
tionnaire data, backed up by some modest behavioral studies, it has two debatable features. First, it
assumes that humans should be distinguished in terms of a small number of uncorrelated dimen-
sions, rather than, for example, by a very large number of psychological scales that may have complex
relationships to each other and are often relevant only to some people. The statistical methods used
to create and validate the Big-5, such as factor analysis, are based on the dimensional assumption,
and thus they do not seriously test it. Second, its empirical measures chiefly concern mild behaviors,
such as those relevant to sedate psychology classrooms and college campuses. For example, rather
than wondering whether conscientiousness and agreeableness inhibit violence, one might postulate a
primary personality dimension measuring propensity to violence that cannot be measured accurately
by the bland questionnaire items of the Big-5.

5.2 GAME-BASED THEORIES

Violence, at least in attenuated form, is central to most online games, and thus it is conceivable that
a personality theory deriving from games rather than classrooms would have some advantages over
the Big-5, and on average it could be no more unrealistic than those based on research inside the
artificial academic world. There is at least one candidate for such a theory, the four motivations for
online game playing suggested by Bartle, R. (1996, 2004):
1. Achievement within the game context (achievers).
2. Exploration of the game (explorers).
3. Socializing with others (socializers).
4. Imposition upon others (killers).
Bartle suggests that these four categories can be mapped in two dimensions: acting versus
interacting and orientation toward players versus toward the world. Of those oriented toward other
 5.2. GAME-BASED THEORIES 49
players, killers act and socializers interact. Of those oriented toward the virtual world, achievers act
and explorers interact. But one could also connect these ideas with standard psychological concepts.
Achievers might be high on need for achievement, socializers high on need for affiliation, and killers
high on need for power, the three dimensions of psychologist McClelland, D. (1961) personality
theory, which does not map perfectly upon the Big-5. Explorers, in contrast, could reflect openness
to experience, the most ambiguous of the Big-5 dimensions.
Yee, N. (2006) has explored the motivations of players from an empirical perspective, while
mindful of Bartle’s theory-based concepts, using a 40-item questionnaire scale. A starting point was
an open-ended question administered to player of EverQuest: “Why does EverQuest appeal to you?”
Yee culled the written responses for different reasons why the game appealed to different people, and
he also drew upon Bartle’s publications for additional items. He then administered the motivation
scale to 6,675 players of such games and performed statistical factor analysis, identifying five distinct
dimensions of variation in the data. Here is how Yee defines the five most important motivational
factors:

1. Relationship: “To interact with other users, and their willingness to form meaningful relation-
ships.”

2. Manipulation: “To objectify other users and manipulate them for... personal gain and satisfac-
tion.”

3. Immersion: “Being in a fantasy world... being ’someone else’.”

4. Escapism: “To temporarily... escape from real-life stress and problems.”

5. Achievement: “To become powerful in the context of the virtual environment.”

Exploratory factor analysis often leaves a number of items either standing alone or in two-item
groupings that are hard to interpret. Yee saw some evidence of three additional factors that could in
principle be defined more precisely in future research that included many more similar items. One
concerned how much people like to lead others, the second emphasized what people believe they
learn about themselves through playing the game, and the final one distinguished people who prefer
solo play versus team play.
Statistical clustering methods like factor analysis work best in identifying major dimensions
of variation that relate to most people, but it is certainly possible that some motivations are highly
idiosyncratic. If a motivation was salient for a tiny fraction of players but not meaningful to many
others, it would not show up in factor analysis, nor might the usual methods of item creation produce
measures that could detect it. However, researchers could start from a different premise and discover
such motivations, and some of them might be theoretically significant.
One motivation that immediately comes to mind, but it may not yet have been studied
systematically, concerns fans of the particular mythos on which a game was based. For example, Star
Wars or Conan fans might not be particularly oriented toward online games, but they start playing
50 CHAPTER 5. PERSONALITY AND MOTIVATION
Star Wars Galaxies or Age of Conan in order to experience a continuation of the enjoyment they have
felt watching the movies and reading the books. This very special motivation might correlate with
Yee’s immersion and escapism dimensions, but it also has a special quality we might call cultural
commitment – valuing the distinctive qualities of a mythos, perhaps, simply from familiarity. An
analogy in the real world would be nationalism. People differ in how nationalistic they are, but
nationalists differ from each other in terms of which nation they value. Of course some worlds, such
as EverQuest, were not based closely on a pre-existing mythos, and thus cultural commitment might
not matter for them, or it would exist only at a high level of abstraction, for example, being a fantasy
fan rather than a science fiction fan.
Yee, N. (2009) has shown that different games attract and reinforce somewhat different per-
sonality orientations, comparing EverQuest primarily with World of Warcraft but also with others. In
particular, EverQuest seems especially capable of building strong relations of friendship and mutual
aid. To begin with, people self-select by entering this somewhat mild fantasy game, which requires
a good deal of patience, rather than a first person shooter or other more violent game. As people
play EverQuest, they find that they absolutely must team up with others. Compared with World of
Warcraft, EverQuest characters have more specialized abilities. For example, clerics are good at heal-
ing but lack good attack spells, whereas the equivalent class in World of Warcraft, priests, can acquire
powerful attack spells as well as healing. Thus, characters with different strengths gain immeasurably
by teaming up. In addition, the temporary death of a character is more costly in EverQuest, and that
cost can be reduced through the help of other players. Put the other way around, World of Warcraft
is more readily soloable than EverQuest. Yee believes that one consequence of EverQuest’s natural
sociability is that many online friendships become real world friendships.
Holsapple and Wu (2007) have suggested a different framework for understanding player’s
motivations, with seven categories: (1) fantasy daydreams, (2) role projection allowing the person to
experience a new role, (3) escapism from problems and pressures in the real world, (4) enjoyment
and the experience of fun, (5) emotional involvement that makes the person feel a part of the
game, (6) arousal whether mere excitement or profound inspiration, and (7) behavior reflecting the
probability of playing the game without attributing a specific motive.
Online games often have their own theories of personality, embodied in attributes of the
character that provide more or less technical advantage under various circumstances, but which also
can be interpreted in more psychological terms. A good example is Matrix Online, which had a
system of five attribute variables which the player could build up differentially in the process of
gaining experience levels: perception, focus, reason, belief, and vitality. At the very beginning when
creating a character, the player must choose one of the ten personality types listed in Table 5.1, each
of which has a different distribution of 40 points across the five attributes (Bedman, 2007).
Note that each of the ten types is high in one attribute, low in another, and average in
the remaining three. Arguably, the five attributes are a five-factor theory of personality dimensions,
competing with the Big-5 popular among academic psychologists. Some of the dimensions even seem
comparable, focus with conscientiousness, and perception with openness to experience – although
 5.3. THEORETICAL DEBATES 51

Table 5.1: Point distributions across attributes and personalities in Matrix Online.

Personality Type Perception Focus Reason Belief Vitality

(awareness of (control, (problem (confidence, (physical
environment) concentra- solving) conviction) wellbeing)
tion)
Detached Spectator 11 5 8 8 8
Devoted Ascetic 8 8 8 11 5
Fanatic Self Improver 8 8 8 5 11
Inquisitive Genius 8 8 11 5 8
Lunatic Fringe 8 8 5 11 8
New Age Hippy 8 11 5 8 8
Secluded Introvert 5 11 8 8 8
Suspicious Cynic 11 8 8 5 8
Troubled Intellectual 8 5 11 5 8
True Believer 5 8 8 11 8

the intellect interpretation of openness in the Big-5 seems closer to reason in the Matrix-5. One of
the ten Matrix types, Secluded Introvert, even uses the introversion-extraversion language of one
of the Big-5. Here is how the game described that type: “You never had many friends. Well, except
the ones in your head. They said the world didn’t need you. They had it wrong. You didn’t need the
world.”
The true believer personality type seems to harmonize with the messianic quality of the Matrix
mythos and the orientations of the friends of the main character in the three Matrix movies who
come to believe he is The One who will save the world. It emphasizes conviction in one’s beliefs,
while minimizing awareness of anything in the environment that might challenge these beliefs.
Its description in the game is: “Possible and impossible are the states of mind. Everything is not
what it seems. Sometimes the senses lie. And you knew why all along.” Traditionally, psychologists
described true belief as an attribute of the pathological authoritarian personality (Adorno et al.,
1950). Frankly, the scientific literature in that area is hard to evaluate because it grew out of analysis
of fascist mentalities in the emotional wake of World War II, was developed by avowedly left-
wing scholars—although some who criticized right-wing mentalities like Lipset and Raab (1970)
and Bell, D. (1963) evolved from socialists into neo-conservatives during their own lives—and true
belief might be considered a virtue by conservatives rather than an authoritarian vice.

5.3 THEORETICAL DEBATES

Psychologists and social scientists have developed a very large number of concepts and measurement
scales to distinguish individuals in terms of their styles of thought, feeling, and behavior. Many of
these concepts overlap while some are very specialized, yet potentially a large fraction could usefully
be applied to players of multiplayer games and inhabitants of non-game virtual worlds. Personality
concepts vary greatly in terms of how well supported they are by scientific evidence, in terms of
52 CHAPTER 5. PERSONALITY AND MOTIVATION
how meaningful they are to non-scientists, and thus how useful they would be for non-technical
discussions about games. Unfortunately, some widely-held theories are probably wrong, and some
of the most popular psychological concepts are not based in good science.
An example is the Myers-Briggs Type Indicator, a widely used self-report questionnaire instru-
ment that supposedly divides people into a small number of distinct types. Based on ninety-year-old
ideas of Carl Jung, Sigmund Freud’s most mystical prominent follower, the MBTI seems to fit what
many ordinary people think about the personalities of others, but it has numerous deficiencies as a
scientific instrument. First of all, it thinks in terms of types, whereas the Big-5 thinks in terms of
dimensions. It may not be valid to capture the rich complexity of an individual in terms of either
scheme, but typological thinking is worse than dimensional thinking because it often assigns people
to separate categories on the basis of very small real differences. The MBTI was created decades ago
by people who lacked advanced scientific training. It tends to get what empirical support it has from
studies by practitioners who have a vested interest because they use the test in professional counsel-
ing. Studies by well-trained academics raise severe doubts about the test’s overall validity, although
one of Jung’s concepts, extraversion, found its way into the Big-5 as well (McCrae and Costa, 1989).
While recognizing the importance of over-arching concepts like the Big-5, it is essential also
to acknowledge that a school of personality research connected to the psychoanalytic tradition has
examined the distinctive characteristics of individuals (Smith et al., 1956; Murray, H., 1981). Such
person-oriented research is sometimes called ideographic, in distinction to the nomothetic research
that seeks general principles that apply across individuals, and it is fundamental to any serious effort
to capture the personality or roles belonging to any individual human being. Thus, players bring their
full personalities to an online game, expressing themselves through their avatars and characters. At
the same time, game designers create distinctive personalities for the most important non-player
characters. Dark Age of Camelot has gone so far as to write extensive dialog for its many quest-givers,
sometimes requiring the player to communicate with them by typing words into text chat as if they
were communicating with a human, in which the non-player characters talk at length about their
feelings and past experiences. Age of Conan allows the player to ask pre-set personal questions of its
quest givers, or to skip this banter, and depending upon the personality of the player, the player may
learn much about the personality of the non-player character.
One area of public controversy about online games is the possibility that some people – espe-
cially impressionable young people – might become addicted to them. This concern naturally con-
nects to the idea that some personality types might be more vulnerable to addiction. Wang and Chu
(2007) have explored this possibility in the light of a recent social-psychological theory that dis-
tinguished harmonious passion from obsessive passion. Vallerand et al. (2003, p. 756) say, “Passion can
fuel motivation, enhance well-being, and provide meaning in everyday life. However... passion can
also arouse negative emotions, lead to inflexible persistence, and interfere with achieving a balanced,
successful life.” The former is an harmonious form of passion, while the latter is negative, obsessive,
and can amount to addiction. Psychiatrists have long recognized that some people have obsessive-
compulsive personalities, and in recent years, distinctions have been made between sub-varieties of
 5.3. THEORETICAL DEBATES 53
obsessiveness, for example, in terms of whether or not the individual suffers from an unusual level
of anxiety that may fuel compulsive behaviors that serve no positive purpose other than to hold
the anxiety in check. However, it is important to distinguish obsessive-compulsive disorder from
mere conscientiousness, one of the Big-5 dimensions of normal personality, and, perhaps, the best
benchmark is the degree to which the behavior serves the person’s general life goals in a manner that
is emotionally satisfying without degrading the individual’s ability to hold a job and enjoy intimate,
real-world social relationships.
Wang and Chu administered a ten-item harmonious-obsessive measurement scale to 404
online game players in Taiwan, using an online questionnaire, as well as a scale measuring Internet
addiction. A quarter of these respondents were female; they averaged about eighteen years old, and
they typically spent about eighteen hours per week playing online games. Indeed, persons scoring
high on addiction tended to score at the obsessive end of the harmonious-obsessive dimension.
With studies like this, there is always a question of whether the results indicate a causal connection:
obsession causes addiction. An alternate interpretation is that obsession and addiction are overlapping
concepts, and some items on each scale to some extent measure the concept of the other scale. But, at
the very least, such studies deepen our understanding of the phenomenon and show that it is possible
to carry out a range of studies on the personalities of game players, based on diverse concepts.
As Sweetser and Wyeth (2005) correctly argue, it would be a mistake to ignore the ways in
which successful games satisfy a wide range of players, quite independently of their personality types.
They suggest that the best characteristic a game can have is the capacity to give the player a positive
experience of flow. This is frankly a rather subjective concept, proposed by a humanistic psychologist,
Mihály Csíkszentmihályi, and it has not found favor among rigorous experimentalists in psychology
or among cognitive scientists. Yet it is attractive to many members of the general public, and for that
very reason may point to key elements of successful, positive human experiences. In the context of
computer games, Sweetser and Wyeth say that flow has eight elements:
(1) A task that can be completed;
(2) The ability to concentrate on the task;
(3) That concentration is possible because the task has clear goals;
(4) That concentration is possible because the task provides immediate feedback;
(5) The ability to exercise a sense of control over actions;
(6) A deep but effortless involvement that removes awareness of the frustrations of everyday life;
(7) Concern for self disappears, but sense of self emerges stronger afterwards; and
(8) The sense of the duration of time is altered.
Except for the last two points, it is obvious how advancing toward completion of a quest in a
well-designed fantasy game can provide these feelings. I find the seventh point, about the self, rather
54 CHAPTER 5. PERSONALITY AND MOTIVATION
mystical, but clearly it relates to a feeling of immersion in the game character and some sense of
refreshment from playing the game. The final point, an altered sense of the flow of time, is part of
the fundamental definition of multiple realities in phenomenology (Schütz, A., 1971), and it ought
to be measurable in rigorous studies, so it seems like a reasonable hypothesis to test.

5.4 NON-PLAYER CHARACTER PERSONALITY

The example cited above from Matrix Online combines two notions of personality, that of the player
with that of the character. Many researchers in and around the field of artificial intelligence have
begun exploring how to incorporate principles of real human personality variation in the non-player
characters (NPCs) they create. Kshirsagar and Magnenat-Thalmann (2002) used Baysian methods
to model NPC mood changes. It is not uncommon for NPCs in existing games to express simulated
emotions, such as rage or fear during combat, either roaring and attacking or whimpering and running
away. What this study added was the ability to give different NPCs different habitual emotional
reactions to events and different common moods, based on three of the five dimensions of the Big-5
theory: “A Neurotic person will change moods often, and tend to go into a negative mood easily. On
the other hand, an Extravert person will tend to shift to a positive mood quickly in a conversation.
An Agreeable person will tend to go to positive mood more often, but frequent mood changes may
not be shown” (Kshirsagar and Magnenat-Thalmann, 2002, p. 110)
Su et al. (2007) also based their research with virtual humans on the Big-5 theory, building a
fuzzy rule-based system to combine the influence of different personality characteristics to control
the interactions and movements of the virtual character. Starting with the assumption that a char-
acter can be high or low on each of the five dimensions, the system produced 32 different rules, such
as: If (Openness is Low) and (Conscientiousness is Low) and (Extraversion is Low) and (Agree-
ableness is Low) and (Neuroticism is Low) then (CharacterType is Hinder) and (BehaviorType is
Cold). Arya and Di Paola (2007) developed a system to control facial expressions of virtual humans
based on a simpler two-dimensional model: dominance versus submission, and cold versus warm
relations with people.
Personality psychology is by no means the only discipline that contributes useful frameworks
for understanding players and creating non-player characters. For just one example, there is a well
established tradition for understanding emotional meanings in terms of a three-dimensional semantic
differential. As originally proposed by Osgood et al. (1957), the semantic differential was constructed
in a manner rather similar to the Big-5 but focused on the judgments people make of things and
events as well as of people. The first dimension is evaluation, how good or bad people judge the
thing to be. The second dimension is potency, and the third is activity. A sleeping tiger is low on
activity and high on potency; whether it should be evaluated positively depends on whether it is in
a cage. Sociologist Heise, D. (2004) has developed a system for giving artificial agents the semantic
differential scores of a large number of nouns and verbs, plus a theory about how to combine their
meanings into sentences, that potentially would allow the agents to act like human beings in a variety
of circumstances.
 55

CHAPTER 6

Avatars and Characters

The player is typically represented in the game environment by one or more characters that have
distinctive characteristics such as gender, race, class, profession, mission, and experience level. The
relationship between players and their representations inside a gamelike virtual world is a complex
one, and it undoubtedly varies by the personality of the player, the nature of the game, and the past
history of experiences. Sociologists have long believed that much can be learned about human beings
from studying the roles and games they play (Pangburn, W., 1922; Stanton et al., 1956).

6.1 BUILDING A BOND WITH AN AVATAR

Let me tell you about my two main World of Warcraft characters, Maxrohn the level 75 Human priest
and Catullus the level 80 Blood Elf priest. Just now, I logged into each in turn and entered “/played”
into the text chat. The system then told me the total time I had run each of them over the past two
and a half years: Maxrohn = 33 days, 9 hours, 52 minutes, 46 seconds; Catullus = 32 days, 15 hours,
35 minutes, 38 seconds. Given that those are 24-hour days, I have invested nearly 800 hours in each
of them. This is more time than I estimate I have invested in any of my co-workers at my real-world
job, which I have held for seventeen years, if only because employment relationships in real-world
jobs these days tend to be fragmentary and fleeting. The point is that many game players invest so
much time in their characters, often “sharing” intensely emotional experiences with them, that they
can become extremely meaningful.
We also invest meaning in characters when we create and develop them. Maxrohn was my
first priest, and I named him after my uncle, Max Rohn, who was an Episcopal priest but had died
some years before. Max was something of an adventurer as well as a cleric, who held churches on
remote islands in the sea and once taught me a judo move that could break a man’s arm. Maxrohn
developed a rich personality over my 800 hours working with him, that was a mixture of my own
personality and what I recalled about my uncle. Not many players may name their characters after
colorful, deceased relatives, but all probably invest in their characters some of their prior feelings
about real people they have known in their lives.
It is also true that highly-developed characters develop a will of their own. As radical as this
idea may seem, I suggest it is literally true, given how the human mind functions. As the extreme
cases of multiple-personality neurosis and veteran movie actors illustrate, a human being can adopt
multiple roles that operate like semi-autonomous personalities. This is how we are able to interact
with other people, by forming in our own minds models of the other person’s thought processes and
desires. In the case of Catullus, I originally thought of him as a reflection of the ancient Roman poet,
Catullus, perhaps, with a little of Julius Caesar thrown in, and even some elements of my two Latin
teachers. Like the ancient Romans, Blood Elves are arrogant, technically competent, hedonistic,
56 CHAPTER 6. AVATARS AND CHARACTERS
and cynical. Those are my character’s characteristics. My Catullus (2008) has actually published a
book chapter under his own name, in which he debates his own existence and his right to become
autonomous from my control. How real is that?
To consider a game character to be somewhat real, we need to grant reality to the world he
lives in. Based on consideration of narrative theory and a range of previous scholarship, Ryan, M.
(2001) has suggested that a player’s relationship to an electronic game can be described in terms of
two largely independent dimensions: Internal-External and Ontological-Exploratory. The Internal-
External dimension concerns how much the individual becomes immersed inside the game world,
whether by identifying closely with an avatar or because the computer technology produces a highly
realistic world the user experiences from a first person perspective. The Ontological-Exploratory
dimension concerns how much the user can affect what happens in the virtual environment, very
much at the Ontological extreme of the dimension, versus hardly at all at the Exploratory extreme.
Thus, the most realistic situation combines Internal with Ontological, but the two dimensions are
conceptually separate, and Ryan shows there are examples of the four logical categories created by
crosscutting these two dichotomies.
Users differ greatly in terms of their orientation toward immersion and how they conceptualize
their digital representations.The naive view of many observers and some players holds that the digital
image that hops when the user presses the space bar directly represents the user. This view is implied
by the term avatar, which originally meant the terrestrial reflection of a Hindu deity. The avatar
is simpler than the deity or user, perhaps, representing just some aspects, but it does possess the
identity of the deity or user to a significant degree. Non-game virtual worlds like Second Life tend
to use the term avatar, but online games do not, preferring the term character. One reason is that
successful games encourage users to have multiple characters, with different capabilities and perhaps
different personalities to match, thereby giving the user a more varied experience and keeping the
user in the game longer. This suggests a third possible dimension crosscutting the two defined by
Ryan, perhaps called Unitary-Multiple. All these are matters of degree, and the Unitary situation is
comparable to reading a novel, whereas the Multiple situation is like reading a collection of short
stories about different characters.
A few gamelike virtual worlds do not represent the player by a humanoid avatar, notably
Jumpgate and EVE Online, which use spaceships instead. Figure 6.1 shows the very complex display
in EVE Online, just as I am being defeated and my ship is about to explode. In theory, the ship is
piloted by a character named Theo who is a priest of the Amarr Empire, and a thumbnail picture of
him is in the upper left corner of the screenshot. But the primary experience of EVE is piloting the
ship myself. The navigation aids in this game are excellent, such as the list of nearby targets at the
right of the picture, but take many hours to learn. We do not need an avatar when we drive a car in
the real world, so why do we need one to drive a spaceship in a virtual world?
First-person shooters differ in terms of the degree to which they assign a specific fictional
identity to the first-person character, or they let the users feel they are operating their gun directly.
Much of the time I spend in Second Life is creating virtual objects and putting programs into them,
 6.1. BUILDING A BOND WITH AN AVATAR 57

Figure 6.1: Being Defeated in EVE Online.

during which time my avatar is an annoying hindrance. So, for all the legitimate research interest in
player-avatar relations, we must keep in mind that there are at least two other alternatives: characters
not conceptualized as representing the player, and working directly through an impersonal user,
interface with neither a character nor an avatar.
The games themselves need not be constrained by these categories, and they sometimes
combine them as in Pirates of the Burning Sea. Figure 6.2 shows my first ship in combat with a pirate
who will shortly be sunk by my cannon. My ship is in the center, and if I zoom the viewpoint inward,
I can see my men working on its deck. Usually, it is more effective to zoom back and see more of
the surrounding sea, watching the ship condition display in the upper left corner, and having no
consciousness of any avatar but the vessel. However, I often have the choice to board the other ship,
at which point I see the battle from the standpoint of my avatar, Edmund Bainbridge, brandishing
his cutlass. Pirates of the Burning Sea is a fairly realistic depiction of commerce and navel competition
in 1720, and I selected for my avatar an actual ancestor of mine who was 18 years old in that year and
had the correct personality to be a freetrader. Depending on the mission, my experience switches
between a focus on the avatar or a focus on the vessel. Given the word’s multiple meanings, vessel
58 CHAPTER 6. AVATARS AND CHARACTERS
would be a good generic term to use whenever a player is represented by something other than an
in-game person.

Figure 6.2: A Sea Battle in Pirates of the Burning Sea.

6.2 THE QUALITY OF AVATAR RELATIONSHIPS

A number of researchers are exploring the extent to which avatars and game characters express
a person’s ideal self, are similar to the individual’s own personality, or possess identities of their
own (Schultze and Leahy, 2009). People have a variety of relationships with their game characters,
and Blinka, L. (2008) did a questionnaire study of EverQuest and World of Warcraft players to deter-
mine whether the level of maturity of the player shaped this variation. A factor analysis distinguished
three dimensions of the relationships. Identification is the degree to which the player feels similar
to the character, and compensation is the degree to which the player would wish to be more like the
character.The strength of identification was greater among young and presumably immature players,
but the data did not permit fully testing age differences with compensation. Immersion is the extent
to which the player feels emotionally involved with the character, for example, feeling ashamed or
proud of what the character does and thinking about the character even while not playing. This
factor does not correlate with age, but it may be stronger among men and among unmarried people.
 6.2. THE QUALITY OF AVATAR RELATIONSHIPS 59
Blinka noted that only 4 percent of women play a male character, but 23 percent of men play a female
character.
Playing these games extensively conceivably could produce maturity in the sense discovered
by Blinka because advanced players often have multiple characters, and thus they develop a range
of kinds of relationships to them. My own example, having run about three dozen characters for at
least a work week each, may be extreme, but the whole point of games offering a variety of characters
is to encourage subscribers to stay longer by exploring many of them.
In a questionnaire study of World of Warcraft players, Bessiere et al. (2007) had people rate
themselves, their ideal self, and their main character along a number of psychological scales. They
found that the character was more like the ideal self along three of the Big-5 personality dimensions:
conscientiousness, extraversion, and neuroticism. This tended to be especially the case for people
who felt depressed and had low self-esteem. The question then arises whether playing a successful
character could possibly reduce depression and raise self-esteem of the player, even outside the game
situation. Current research cannot answer this question, but we can imagine longitudinal studies to
explore it rigorously. For present purposes, the point is that player-character relations may change
over time for any given player or category of players, as well as differing from one player to another.
To compare avatar-creation systems and explore physical appearance dimensions of the player-
character relationship, Ducheneaut et al. (2009) administered questionnaires to people who use
three very different virtual worlds (VWs), World of Warcraft, Maplestory, which is a two-dimensional
multiplayer role-playing game, and the non-game virtual world, Second Life. Males were more likely
to create an avatar of the opposite gender than females were, and older people tended to make
their avatars appear younger. Players in all three environments considered the style and color of
their character’s hair to be a high-impact decision, and overweight players tend to prefer slimmer
characters. This study also asked respondents to rate both themselves and their main characters
on the Big-5 personality dimensions, finding they rated their characters more conscientious and
extraverted but less neurotic than they considered themselves to be. However, the novel findings of
this study concern physical appearance:
Our data also show that VWs are used to experiment with digital bodies that are often
very different from a user’s. In a large majority of cases, VWs users create a digital
identity that looks close to Western ideals: leaner, younger, more fashionable versions
of themselves. This trend is particularly prevalent for older users and those with weight
issues, but it also benefits these users more – they are more satisfied and more attached to
their enhanced online selves than average.While such “perfect bodies” therefore appear to
have clear benefits to some, it is also worth mentioning that it may cause VWs to become
much less diverse than physical environments in the long run, which may in turn lead
users to a fairly stereotyped vision of what a human body should be (Ducheneaut et al.,
2009, p. 1160).
Huh and Williams (2010) have looked closely at gender-switching in EverQuest 2, employing
a rich dataset combining a questionnaire administered to a large number of players with informa-
60 CHAPTER 6. AVATARS AND CHARACTERS
tion directly about their characters from the game servers. The most striking finding was that few
players switch genders when creating a character, and, apparently, free exploration of a new identity
is not a major factor motivating people to play these games. A set of suggestive findings chiefly
concerning female players indicated that this general rule did not apply to everybody. One category
of female players did appear to be ambivalent about their conventionally defined gender, adopted
male characters, and behaved in what might be described as a hyper-masculine fashion.
Fron et al. (2007) note that scholars have tended to dismiss the tendency of humans to explore
different ways of appearing by trying on various kinds of clothing and dressing up for public display:
“From a feminist perspective, we can look at dress-up, which has historically been associated with
female play, as fulcrum for the creation of more gender-inclusive games, as well as a bridge to more
gender-balanced approaches to game studies.”
Barr et al. (2006) have noted that avatars frequently change their appearance, behavioral abil-
ities, and other characteristics during solo-player computer games, and this is certainly true for
the longer-duration online role-playing games. Both World of Warcraft and Entropia Universe allow
characters to get haircuts. Quests in many games put the character in disguise. For example, several
quests in both Dark Age of Camelot and World of Warcraft involve donning a disguise to get into an
enemy camp to overhear their secret plans.
More research will be needed to determine the range of meanings people attach to gender
swapping and other situations in which players create game characters that are very different from
themselves (Hussain and Griffiths, 2008). However, I must suggest that the answer to such questions
may not be stable, either across games or across the passing years. For example, when people first
started interacting online, many observers feared this reflected or could cause social isolation, yet
those concerns largely evaporated as online communications matured. It may be that a significant
fraction of people attracted to online games today have unusual motivations, and that as this new
medium grows, its users will become more like the general public.
It is also true that each player gains experience that may radically change relationships with
characters. In my own case, I must admit some trepidation when I created my first female character,
a hesitation that had vanished by the time I had created my tenth. From my own perspective, they
are characters with some degree of individual integrity of their own, just like the characters of many
types created by a novelist, and I don’t think of them as avatars any of which represent my self-image
or my ideal. Different people will approach these issues in a wide variety of ways that are likely to
shift as the industry matures, giving researchers many years of interesting work to understand this
historic evolution.

6.3 SECONDARY AVATARS

As we have noted, experienced players of these games tend to operate multiple characters. If a player
concentrates on one character to the exclusions of others, then it is the player’s main, and the others
are alts. But we have also referred to secondary avatars, and there has been very little research to
my knowledge about these avatars of avatars. The games conceptualize them in various different
 6.3. SECONDARY AVATARS 61
ways, for example, a coder character in Matrix Online, who has acquired the right programming
skills can create a proxy, to help fight or even to heal the main avatar. The user interface panel for
controlling a proxy has displays for health and inner strength, just like the main character, plus two
sets of buttons with which to give commands. Three buttons at the left set the temperament of the
assistant: aggressive, passive, or fighting enemies only when they attack. A set of four buttons on the
right tell the assistant whether to attack the selected target, follow the main avatar, stay in its present
location, or cease to exist. On occasion I had told a proxy to stay where he was, then I walked as far
as a virtual city block away and called him to come. Surprisingly, even when the route was complex,
he often found his way.
Perhaps the archetype of secondary avatars is a hunter’s pet in World of Warcraft. First, a
hunter must tame a wild animal, give it a name, and begin to train it. At even a fairly low level of
experience, the pet’s control bar includes common commands like attack, follow, and stay – or moods
like aggressive, passive, or defensive. Also the pet may be made to growl at the enemy to get it to
focus on the pet rather than the main character, prowl in stealth mode so the enemy won’t notice it,
and claw or rake at the enemy. Other user interface options feed or heal the pet, and the player even
can look though the eyes of the pet when it goes scouting.
While games differ, something like a convention has emerged about the primary controls
used to operate a secondary avatar. Here are primary modes and commands Lord of the Rings Online
offers for bear pets, plus a couple of the more complex options:

Aggressive Mode - Your pet will enter combat when valid targets are detected.

Passive Mode - Your pet will not enter combat unless commanded to.

Guard Mode - Your pet will enter combat to defend you or itself.

Attack - Your pet will attack your currently selected target.

Follow - Your pet will move towards you and then act according to its mode.

Stay - Your pet will stay in its current position unless its mode requires combat.

Toggle Assist - When enabled, your pet will attack targets that you attack.

Roaring Challenge - This skill forces the target to attack the bear.

Sign of the Wild: Rage - This sign asks your animal friend to go into a rage, attacking faster
and making itself look as threatening as possible.

Extreme examples of specialized secondary avatars can be found in Age of Conan, where
necromancer characters can command entire teams. Figure 6.3 shows Eridanos and his team of six,
just as they retrieve a relic from Atlantis called the Phoenix of the North in a heavily defended ice
cave in the Eiglophina Mountains, after killing a witch named Mithrelle. In the center, proudly
dressed in his white robe, is Eridanos the level 52 Stygian necromancer. Immediately behind him
62 CHAPTER 6. AVATARS AND CHARACTERS
are two deathless acolytes, an arcanist and a mage, who supplement his damaging spells. The other
four assistants are undead minions. At the far left are a life-stealer that takes life from the enemy and
gives it to the team, and a mutilator that inflicts bleeding wounds. On the right side of the picture
are a corruptor that adds additional holy damage to each attack and a harvester that gives the team
mana and stamina.

Figure 6.3: Eridanos with Secondary Avatars in Age of Conan.

Games differ in terms of how much they personalize secondary avatars. In World of Warcraft,
for example, the player names a hunter’s pet, and when it dies, it is revived with the same identity,
whereas a Warlock’s minion is summoned from another realm and cannot be given a distinctive
name. The proxies in Matrix Online appear at random male or female each time one is summoned,
dressed like local gang members with unpredictable weapons specialization, and cannot be named.
In order to name the robot Tobor in Anarchy Online, I needed to write a simple macro program and
 6.4. FACING THE END 63
run it whenever I generated a robot, so that game allowed me to personalize the secondary avatar
but did not require it. In the much more recent game, Lord of the Rings Online, Rumilisoun easily
gave her sturdy bear the permanent name Beorn, her aloof raven the name Poe, and her stealthy
lynx the name Lamhainn. Given their complexity and their potential to be meaningful to players,
secondary avatars would be a fruitful topic for future research.

6.4 FACING THE END

Perhaps the most profound question for player-character relations is death. Lisbeth Klastrup has
written repeatedly about this, and she notes that death in the games always allows the possibility
of resurrection, whereas death for players is permadeath, from which there is no obvious worldly
salvation. Klastrup, L. (2006) suggests that in this context death has four possible meanings:

• Avatar death as in-game mechanics (concrete death).

• Death as a symbolic event visualized in a variety of forms.

• Avatar death as narrativised event in the life of an avatar.

• The death of a player.

In another essay, Klastrup, L. (2008) describes the in-game mechanics of death in World of
Warcraft. First, the player sees the character fall on the ground, then a dialog box opens asking if the
player wants the character to resurrect at the nearest graveyard. If no other player casts a resurrection
spell, the character teleports to the graveyard in six minutes, even without a click in the box. The
character is not immediately restored to life at the graveyard but exists in a shadowy state, in which
the world appears colorless, until either the player decides to accept the offer of resurrection from
an angel, which leaves the character temporarily weakened, or the character runs back to its corpse
and resurrects itself.
Games differ in how death is handled, and some do not even permit a player’s character to die.
In Lord of the Rings Online, the character merely retreats in a weakened condition. In EVE Online,
the character escapes from his or her spaceship in a pod, which seems gentle enough except that the
loss of a spaceship is extremely costly. In Dark Age of Camelot, so the mythos says, magical stones
fell from heaven after the death of King Arthur, and deceased avatars revive at whichever of these
stones they most recently bound themselves to, which can be very far away from where they died,
rather than nearby as in World of Warcraft. The character must pay a healer to restore lost strength to
them, and they can return to the scenes of their deaths to regain some lost experience. This is done,
dramatically enough, by kneeling in prayer at their own tombstones.
Klastrup’ second meaning of death, as a symbolic event, suggests that the death and resurrection
of a player’s character may have emotional, philosophical, or even religious meaning for the player.
Two of my own favorite avatars, Cosmic Engineer and William Bridgebain, died permanently when
Matrix Online and Tabula Rasa were closed down. A total of thirty others died, nearly permanently,
64 CHAPTER 6. AVATARS AND CHARACTERS
when I cancelled my subscriptions after completing the intended research, although many game
companies keep the characters on file in cases the player wants to reactivate the account. Yes, it is sad
to lose an old friend, although players undoubtedly differ in the degree to which they conceptualize
the termination of their characters in this way.
Klastrup’ third meaning recognizes that the life of an avatar is a story, and death can conclude
a chapter of that narrative, as when a brave comrade sacrifices life for his or her guildmates in an
heroic battle.The fourth meaning, permadeath of a player, raises the challenging question of whether
characters could live on after their players’ deaths, perhaps, operated by some form of artificial
intelligence that had used machine learning to archive the players style. Readers who dismiss this
idea as foolish and superficial should meditate upon the fact that they can interact with a deceased
person’s avatar on the Aldor Rise in Shattrath City, in World of Warcraft. At age 28, a player named
Dak Krause died of leukemia, and his level 70 Night Elf huntress, Caylee Dak, was placed in the
game as a memorial. Her one function in the game is to receive a sentimental poem about death
sent to her from a little girl character, and to bless the person who delivers it.
 65

CHAPTER 7

Virtual Professions and

Economies
Long-duration games often invest a good deal of the action into the production and sales of virtual
goods by players, using elaborate systems of resource gathering, item crafting, and exchange through
complex auction houses. One of the ways to win status is to excel in one or more professions. For
example, in Entropia Universe, the public text chat abounds with messages like these:

• Max aka Cayenne found a deposit (Lytairian Dust) with a value of 72 PED!

• Aleksandra Mayday Kowalski found a deposit (Copper Stone) with a value of 99 PED!

• Ericsson Cozzmo Kramer killed a creature (Atrox Young) with a value of 135 PED!

• Team “Jedi & Trooper” killed a creature (Hogglo Young) with a value of 61 PED!

• Tom Psychodadiks Delonge killed a creature (Aurli Watcher) with a value of 286 PED and
has been recorded in the Hall of Fame!

• Anunka Mithra Skarpnes manufactured an item (Zombie Arm Guards (M,L)) worth 829
PED and has been recorded in the Hall of Fame!

“PED” in these messages refers to Project Entropia Dollars, ten of which nominally equal a
US dollar, so that 829 PED equals $82.90. Thus, people are winning money as well as fame, and
depending upon your perspective, that sounds either like a game or like capitalism. In some games,
practicing a profession can be an alternate way of advancing in the main status contest, comparable to
completing quests and killing enemies, depending on the specialization of the character. An engineer
in Star Wars Galaxies does not get any status advantage from killing an enemy, but he does from
collecting materials from the environment and assembling them to create a robot, whereas a Jedi
knight chiefly gains status by slaughtering enemies with a light saber. More commonly, however,
online games have separate systems of honor and advancement for ordinary adventuring versus
specialized professions, permitting any character to do both in whatever mixture the player desires,
although access to economic resources depends heavily upon advancement in combat experience.

7.1 WORK IN STAR WARS GALAXIES

To compare the extremes in careers of MMORPG characters, I ran two some distance up the
status ladder in Star Wars Galaxies, an engineer named Algorithma Teq and a Jedi named Simula
66 CHAPTER 7. VIRTUAL PROFESSIONS AND ECONOMIES
Tion. While Simula was gallivanting around doing heroic quests for characters like Jabba the Hut,
Algorithma was plodding around the planet Tatooine laboriously collecting raw materials and using
a series of tools to craft them into valuable virtual devices, including the tools themselves.
Many games allow characters to gather various resources from the natural environment, but
they differ in the restrictions they place on which characters can gather which resources, what
virtual tools are required to gain a resource, and what information is provided to the player about
the resources that exist in the vicinity. In Entropia Universe, any player can acquire and develop
proficiency with prospecting tools, but there are no hints about what resources might be nearby
before the tool sets off a little explosion or sends a probe down a hole, and doing so uses up one of
the explosives or probes the player has earlier purchased. In Tale in the Desert, a character looking
for slate to make blades knows to walk along a shoreline until a slate icon appears at the top of the
screen and to click on the icon to tell the character to bend down and pick up the stone. A miner or
herbalist in World of Warcraft can set the little map in the interface to display a gold dot indicating
resources but only nearby and only for characters trained in the appropriate profession.
When Algorithma went prospecting, she had to have the right tool and decide when to use it,
although doing so was costly only in terms of time. She had to tell the tool exactly which resource to
search for, and Figure 7.1 shows the display after she has found some Didoreo, which is a variety of
Agrinium, which is a variety of Aluminum, which is in the non-ferrous metal category. On the left,
the display shows a map of the vicinity, with percentages indicating the chance of getting Didoreo
at each location. When she started, at some distance from her final location, the percentages were
very low, even 0%, but in one direction, they were slightly higher, so she walked that way. Again and
again, she would prospect, see which direction had the higher percentages, and go in that direction,
until she found a good place to stop and extract the needed Didoreo. Eventually, even a rich deposit
becomes depleted, and she would need to resume prospecting.
Many games require a character to have both skills and instructions, as well as raw materials,
before being able to create a particular virtual object or substance. Depending upon the game, skills
are acquired through training, which may cost money, or through experience doing similar but easier
work in the past. Instructions take the form of a recipe or schematic, which may be found in the field,
bought from a vendor, traded between players, or automatically acquired at some level of general
experience.
Early in her development, Algorithma made a chemical survey device, which she could then
use in her own prospecting. Figure 7.2 shows what her display looked like at step two out of six.
The display explains, “This tool scans the local environment for chemical resources than can be
extracted with the appropriate types of harvesters and even has the ability to acquire minute samples
of the resources it detects.” This implies that an advanced engineer will need a whole toolkit of
specialized devices, each of which may be manufactured in this manner. To make this device, she
needs a generic crafting tool, which she can also use to make more advanced or specialized crafting
tools able to do more advanced projects. She also needs the schematic, which has “instructions” (she
cannot read them, but they permit the required manufacturing steps) on how to make and assemble
 7.1. WORK IN STAR WARS GALAXIES 67

Figure 7.1: Prospecting for Resources in Star Wars Galaxies.

four subcomponents, in this case: the assembly enclosure, the controller mechanism, the scanner
assembly, and the storage unit. On the left side of the display are the relevant resources she already
possesses, and she has most recently selected amorphous gemstone, which she earlier intentionally
prospected for and collected. At this point, she is ready to assemble the subcomponents, which she
will do by clicking the [Assemble] button at the lower right.
Although this is an extreme example, very commonly games represent creation of economic
value in terms of gathering raw materials, gaining technical abilities, and going through a complex
series of steps to achieve the desired goal. This is a nice metaphor for how any real economy works,
aside from the market mechanisms that handle investment and exchange. But it is also a model of
how computer programming works, based on algorithms which are precise sets of instructions for
achieving a given goal in a finite number of steps.
68 CHAPTER 7. VIRTUAL PROFESSIONS AND ECONOMIES

Figure 7.2: Making a Simple Device in Star Wars Galaxies.

A final point about the division of labor in Star Wars Galaxies is valuable both in understanding
how gameworlds evolve and in recognizing the care that researchers must exercise when relying upon
publications. In November 2005, a major update called the New Game Enhancements radically
transformed the way the different professions and character lines worked, including letting players
start a character as a Jedi and greatly simplifying the development lines for the existing professions
such as engineers. Prior to this transformation, a guidebook had been published, Star Wars Galaxies:
The Total Experience (McCubbin et al., 2005), which describes the situation prior to this change in
great detail, so a researcher or student must not assume it describes the game as it existed after this
revolution. Reportedly, many players who were heavily invested in the earlier system left the game
in disgust, but players who joined later may have found the new system more comfortable.

7.2 PRODUCTION IN WORLD OF WARCRAFT

The most successful internal economy of any multiplayer game is the one in World of Warcraft at
the very least because so many players engage in it, rendering the marketplace rich with the virtual
 7.2. PRODUCTION IN WORLD OF WARCRAFT 69
resources they gather and exchange. To illustrate how statistical analysis can be performed in studies
of games, I will use two quantitative datasets I collected at some effort. One dataset covers professions
and several other variables about 1,096 members of the Alea Iacta Est guild on the Earthen Ring
server, all those who were level 10 or above in the period February 17-22, 2008, a guild that will be
described more fully later. To draw a sample for the other dataset, I used the CensusPlus add-on
program to identify of all characters online at any point during Saturday, January 12, 2008, on two
contrasting other servers, Emerald Dream and Scarlet Crusade. The former is a PvP server, while
the latter is not, so to get a balanced sample, I removed 97 cases at random from Emerald Dream.
The actual data in both cases came from manually looking up each of these characters in World of
Warcraft’s online database, The Armory1 .
A character in World of Warcraft can perform any of the three secondary professions – fishing,
cooking, and first aid – but only two of the ten primary professions that were available for most of
the 2007-2008 period: alchemy, blacksmithing, enchanting, engineering, herbalism, jewelcrafting,
leatherworking, mining, skinning, and tailoring. Developing skill in any of these takes a considerable
investment of time and effort, so while characters may switch professions, they tend to be locked
into the ones they learned when they were around level 10 in general experience.
At a first approximation, there are two categories of professions, gathering and crafting. Gath-
ering professions collect resources from the landscape. Miners collect ore from which metals and
gems can be extracted; herbalists collect plants that serve as ingredients for potions, and skinners
take leather from the corpses of animals. These resources can be traded directly to other players,
placed on the auction house, or sold to non-player characters called vendors. Thus, they are economic
resources. In harvesting these resources, characters compete with each other, but skinning also has
an implicit cooperative aspect. Often, a skinner comes across the corpses of dead animals killed by
non-skinners and can take their hides without the inconvenience of killing them first. A gatherer
gains skill mainly by gathering, with rare trips to a non-player trainer, and increasing skill is required
to gather increasingly valuable resources.
Crafting professions involve manufacturing useful goods from raw materials, so they have a
dual relationship to the division of labor and economic exchange. First, the crafter must obtain the
necessary raw materials, either by gathering them personally or by purchasing them. For example, an
engineer can make a nice, functioning telescope called the “ornate spyglass,” by assembling 2 bronze
tubes, 2 whirring bronze gizmos, 1 copper modulator, and one 1 moss agate gem. Each tube, in turn,
requires 2 bronze bars and 1 weak flux; each gizmo requires 2 bronze bars and 1 wool cloth; each
modulator requires 1 copper bar, 2 linen cloths, and 2 handfuls of copper bolts. A handful of bolts,
in turn, requires 1 copper bar. The moss agate gem is somewhat rare, so engineers tend to buy it on
the auction house even if they have the mining profession and use it to get the copper plus the tin
that makes bronze when alloyed with copper.
Second, the crafter sells at least some of the goods produced, in order to have the money needed
to buy materials. In addition, but not involving other players, crafters must purchase frequent training

1 http://www.wowarmory.com/
70 CHAPTER 7. VIRTUAL PROFESSIONS AND ECONOMIES
from non-player trainers to gain the skills required to make particular goods, in an ascending ladder of
abilities. One profession, enchanting, has aspects of gathering as well as crafting because enchanters
can obtain some of the materials they need by disenchanting already-enchanted objects they loot
from non-player enemies.
There are two ways of evaluating the utility of a profession: in terms of its value for the
individual player, or for its value to other players in the individual’s guild. Among the most obvious
kinds of utility for the individual are whether the profession can craft the armor that the given
player can wear. Mages, priests, and warlocks can use only cloth armor, which can be crafted by
tailors. Below level 40, druids, hunters, rogues and shamans can use only leather armor, made by
leatherworkers. Paladins and warriors use metal armor, made by blacksmiths. Thus, one important
variable is whether or not characters can make their own armor.
All characters loot the cloth required by tailors, but leatherworkers rely upon skinners to
provide their raw materials, and blacksmiths similarly rely upon miners for the smelted metals
they require. Therefore, it is worth asking whether leatherworkers and blacksmiths have the second
profession required for their own craft.Those who wear cloth armor, naturally are least well protected,
so they may benefit from the craft of enchanting more than other classes. Furthermore, enchanters
can craft magic wands, which are used by the same three classes that wear cloth, so enchanting and
tailoring are linked indirectly, as well.
Table 7.1 shows the percentages for each class practicing each profession where the numbers
are at least 20 percent, across the two data samples. Printed in bold are the professions that produce
the given class’s armor – such as blacksmithing for warriors – plus the profession that supports the
particular kind of armor-making. We can see that mining is popular across most classes, and this
is the case because the metal and gems obtained through mining are needed by fully three of the
crafting professions: blacksmithing, engineering, and jewelcrafting.

7.3 DIVISION OF LABOR IN PROFESSIONS

The professions practiced by characters in many online games represent a sophisticated system of
division of labor (Durkheim, E., 1933), and they possess complex economic interconnections. We
shall first look at connections between major professions in World of Warcraft, then briefly compare
with three other examples.
Table 7.1 shows that each World of Warcraft class rates highly the gathering profession that
provides the raw material for its own kind of armor. Characters creating their own armor and
gathering the raw material for it themselves are more self-reliant and, therefore, less dependent
upon their group, than other characters, all else being equal. On the other hand, some characters
practice two of the three pure raw-material professions (skinning, mining, and herbalism), often
with the aim of selling the materials to non-player vendors or on an auction house. Other pairs of
professions also have practical links, so it is worth analyzing the connections between professions.
Table 7.2 shows how popular each profession is, plus its statistical connection with other
professions. The correlations are the Pearson r coefficient, which ranges from -1.00 through zero
 7.3. DIVISION OF LABOR IN PROFESSIONS 71

Table 7.1: Professions Preferred by World of Warcraft Classes.

Class Alea Iacta Est (1,096) Two Realms (1,420)
Warrior Mining 71% Mining 73%
Blacksmithing 37% Blacksmithing 53%
Herbalism 28%
Paladin Mining 80% Mining 71%
Blacksmithing 34% Blacksmithing 40%
Rogue Skinning 50% Skinning 43%
Mining 36% Leatherworking 37%
Leatherworking 32% Mining 31%
Engineering 22% Engineering 23%
Herbalism 21%
Hunter Skinning 59% Skinning 51%
Mining 43% Mining 41%
Leatherworking 33% Leatherworking 36%
Engineering 21%
Druid Skinning 63% Herbalism 43%
Herbalism 42% Skinning 40%
Leatherworking 31% Alchemy 32%
Alchemy 20% Leatherworking 24%
Shaman Skinning 41% Mining 43%
Herbalism 38% Skinning 30%
Mining 32% Jewelcrafting 24%
Alchemy 26% Leatherworking 24%
Leatherworking 20% Herbalism 23%
Warlock Tailoring 53% Tailoring 48%
Enchanting 38% Enchanting 39%
Mining 29% Mining 31%
Skinning 24%
Mage Tailoring 46% Tailoring 44%
Enchanting 46% Enchanting 41%
Mining 29% Mining 30%
Herbalism 22%
Priest Tailoring 54% Tailoring 53%
Enchanting 41% Enchanting 36%
Herbalism 27% Mining 27%
Mining 25% Herbalism 20%
Skinning 21%
72 CHAPTER 7. VIRTUAL PROFESSIONS AND ECONOMIES

Table 7.2: Correlations between Professions in World of Warcraft.

Alea Iacta Est (1,096) Two Realms (1,420)
Profession Percent Correlations Percent Correlations
Alchemy 14.3% Herbalism 0.66 14.7% Herbalism 0.78
Tailoring -0.28 Skinning -0.25
Skinning -0.31 Mining -0.33
Mining -0.33
Blacksmithing 7.6% Mining 0.31 12.1% Mining 0.37
Skinning -0.21 Skinning -0.21
Tailoring -0.24
Enchanting 19.6% Tailoring 0.54 19.8% Tailoring 0.48
Herbalism -0.21 Herbalism -0.20
Leatherworking -0.21 Leatherworking -0.20
Mining -0.26 Skinning -0.20
Skinning -0.28 Mining -0.26
Engineering 10.3% Mining 0.29 13.2% Mining 0.30
Skinning -0.25 Skinning -0.21
Herbalism 23.4% Alchemy 0.66 19.3% Alchemy 0.78
Enchanting -0.21 Enchanting -0.20
Tailoring -0.22 Leatherworking -0.21
Leatherworking -0.23 Mining -0.32
Mining -0.33
Jewelcrafting 6.4% Mining 0.24 9.9% Mining 0.24
Skinning -0.20
Leatherworking 15.0% Skinning 0.55 15.7% Skinning 0.66
Tailoring -0.20 Enchanting -0.20
Enchanting -0.21 Herbalism -0.21
Herbalism -0.23 Mining -0.35
Mining -0.34
Mining 40.6% Blacksmithing 0.31 40.4% Blacksmithing 0.37
Engineering 0.29 Engineering 0.30
Jewelcrafting 0.24 Jewelcrafting 0.24
Enchanting -0.26 Enchanting -0.26
Tailoring -0.28 Skinning -0.26
Alchemy -0.33 Tailoring -0.30
Herbalism -0.33 Herbalism -0.32
Leatherworking -0.34 Alchemy -0.33
Leatherworking -0.35
Skinning 36.3% Leatherworking 0.55 26.9% Leatherworking 0.66
Blacksmithing -0.21 Enchanting -0.20
Engineering -0.25 Blacksmithing -0.21
Tailoring -0.28 Engineering -0.21
Enchanting -0.28 Alchemy -0.25
Alchemy -0.31 Mining -0.26
Tailoring 18.7% Enchanting 0.54 16.6% Enchanting 0.48
Leatherworking -0.20 Alchemy -0.30
Herbalism -0.22
Skinning -0.24
Alchemy -0.28

to +1.00, and measures the negative or positive associations between pairs of variables. There is a
built-in negative correlation because selecting one profession uses up one of the two choices and thus
makes it less likely to select any other profession. All coefficients shown are statistically significant,
and the table focuses just on those with a respectable magnitude, at least 0.20 from zero in either
direction. The very strong connections between alchemy and herbalism (0.66, 0.78) reflect the fact
that herbs are among the raw materials needed by alchemists to make their magic potions.
Less than half of the characters could make the armor they themselves wore. This implies
they often relied upon other players to create their armor for them, and they produced other goods
 7.3. DIVISION OF LABOR IN PROFESSIONS 73
to provide in turn. It implies that the entire system is set up in such as way as to encourage coop-
eration between players in industry as well as battle. However, one of the motivations for having
multiple characters is to practice multiple trade skills, and the multiple characters belonging to one
player can send goods and virtual money to each other through the in-game mail system. Note
that both possibilities encourage players to keep paying their subscriptions. Economic exchange
with other players builds social bonds that become valuable in themselves and thus a source of
commitment (Homans, G., 1950). Developing multiple characters increases the time a player must
play.
Compared with World of Warcraft, Dark Age of Camelot does not have a complete, modern
economic system of resource extraction, production and exchange. There are six trade skills: alchemy,
armorcraft, fletching (arrow-making), spellcraft, tailoring, and weaponcrafting. Production is not a
full economic system. Characters can learn only certain crafting skills, depending on the class to
which they belong, and they cannot make anything they themselves cannot use. Raw materials must
be bought from non-player characters, rather than gathered from the environment. There is no
elaborate auction system, such as in World of Warcraft and other more recent games, and trading
takes place between individual players. In the case of spellcrafting and alchemy, this makes perfect
sense because the crafter often enhances an item the customer already possesses. But the primitive
nature of exchange also fits a Dark Ages society where the economy is largely based on barter.
The Matrix Online allowed any character to practice any number of skills, although gaining
them all was a tremendous labor and players tended to specialize. When near one of the hardline
telephone booths that functioned as an interface to the fictional mainframe computers, a player could
switch various program modules in and out of inventory, gaining access to the corresponding skills,
limited only by the number that could be carried at once by a character at the given experience level,
and by the fact the player needed to have earned the skill and placed it in inventory. Thus, players
who wanted to experience different specializations, whether of crafting or combat, did not need to
have multiple characters, a fact that compensated to some extent for how tedious leveling up tended
to be at the higher levels.
Because it lacks quests and similar lore-based adventures, Entropia Universe relies more heavily
than the other examples on gathering resources and crafting virtual items. Prospecting for mineral
resources requires explosive charges and probes that are destroyed in the process of using them, and
they need to be bought from vending machines where weapon ammunition, animal decoys, and
other perishable items need to be bought as well. This was a key part of the larger economic system,
because the money to buy these commodities comes from exchanging external currencies like US
dollars for the internal currency, Project Entropia Dollars.
Crafting must be done at machines found in cities or outposts, but this is not an inconvenience,
because at the same location can be found the character’s storage locker and an auctioneer from which
materials can be bought. Once an item had been made, it can be sold to a vending machine for a
set price, which was usually done when the player was crafting for the purpose of building up the
character’s skill level and thus making things that no other player might want, but it also can be
74 CHAPTER 7. VIRTUAL PROFESSIONS AND ECONOMIES
sold to another player through the auctioneer. Despite the fact that the auction system in Entropia
Universe tends to have far fewer items for sale than the World of Warcraft system, it works well,
because many of the items for sale indeed find customers, and patient players can check the system
frequently to see what had just been put up for sale and either buy it outright for a high price or
place a bid just above the previous high bid. Given the great effort required to gain a high skill level,
players tend to specialize, and there is no economic advantage in having multiple characters.
 75

CHAPTER 8

Social Relations Inside Games

On the one hand, online games bring people together in complex visual environments created by
a marriage of art and technology, and, on the other hand, they construct intricate societies both of
players and of non-player characters that are often fractured into competing virtual clans. Social
science research methods are appropriate for studying online communities, and the results may
inform social science more generally because these are, after all, “real” communities.

8.1 EMERGENT SOCIAL ORGANIZATION

Based on extensive ethnography inside World of Warcraft, Nardi and Harris (2006) have classified
different levels of social organization that mark collaborations between players.They distinguish two
kinds of unstructured collaboration, strangers in the fight and random acts of fun. An example of the
fight classification is buffing a stranger. Some classes of character, such as a priest, have the ability
to cast a positive spell on another character, called a buff, at the cost of a small amount of mana. If
the spell caster is not currently in a battle, this mana replenishes quickly and thus costs the buffer
nothing. Some buffs briefly diminish the recipient’s health, and some buffs require the recipient to
be at a high experience level, so a buffer must exercise some care. However, players very often buff
other players they encounter at random, as a gesture of benevolence and an expression of their own
competence.
Another “strangers in the fight” example is providing a kill assist when encountering a player
locked in combat with a non-player enemy. Here, too, some care is required. In my own experience, I
often encounter a player battling a well-matched non-player opponent and feel it would be insulting
to the other player to immediately kill the opponent, so I stand and watch. In many MMORPGs,
it is easy to monitor the health of both combatants, and if I see that the other player is likely to
succeed, I let him or her accomplish the victory alone, which I feel is more honorable. My character
stands back so as not to constitute a distraction, and after watching an heroic battle, I will bow at the
victor, thus showing respect. However, if the health of the other player becomes imperiled, then a kill
assist is warranted, because in all the games there are penalties for being killed. By my observation,
many players typically do not follow this somewhat subtle etiquette, and they provide a kill assist
even when it is not wanted.
Nardi and Harris also describe structured collaborations with friends and strangers, in four
subcategories. First, there are relatively temporary but well-structured small parties and larger raids,
as well as a player’s friends list. Many quests and instances in essentially all of the MMORPGs require
more than one player to complete, and five is the maximum number who can belong to a brief party
in World of Warcraft. The computer interface lets a player see the health conditions and approximate
locations of each of the others, and a special text chat channel opens up for the party members. Raid
76 CHAPTER 8. SOCIAL RELATIONS INSIDE GAMES
groups can have as many as 40 members, structured as a collection of parties each with a sub-leader,
and I have occasionally participated in raids that were so big they required two or even three raid
groups, reaching a hundred total membership. Parties and raid groups can use voice chat as well,
although in raids it is often set so that all participants may listen but only sub-group leaders and the
chief leader may speak. The leader may also send warning messages that display prominently across
the center of each person’s computer screen. As Nardi and Harris note, players tend to form friend
relationships linking them in pairs who can see when the other person is online and perhaps invite
the friend to join a party. Some of these mutually agreed upon friend relationships last a long time,
but some are of very short duration, as are parties and raids.
Two other major categories identified by Nardi and Harris are guilds, about which there is
a very extensive scientific literature, and battlegrounds which have not been studied closely. As we
shall shortly see in detail, guilds are enduring, structured groups with anywhere from a handful of
members, to one or two hundred, to, in rare cases, as many as 5,000 members – noting that members
are characters, not players, and a serious player typically has several characters in their favorite
guild. Battlegrounds are special instances where two large teams may compete, one representing the
Alliance faction and the other the Horde faction, and they can recruit characters from more than one
realm (shard) of the game. Since Nardi and Harris did their research, arena teams have become very
popular, enduring groups the size of small parties that regularly compete in tiny battleground-like
instances, rather in the way that amateur basketball teams do, although basketball players rarely “kill”
their opponents.
The final major collaboration category identified by Nardi and Harris concerns brief two-
person dyads in two subcategories, duels and trades. As the name implies, a duel is combat between
two players, but they always end with the loser falling on one knee and admitting defeat, rather
than being “killed.” Duels can often be seen at Goldshire, a short walk from the gate of Stormwind
City, although they may happen anyplace. One player challenges another, who is free to accept or
reject the challenge, and a tall flag appears to tell others that a duel is in progress. The winner may
challenge all comers, which is one reason that a few locations like Goldshire have become traditional
dueling sites.
A trade is a peaceful economic exchange, in which the user interface ensures that neither
party will steal from the other. When one character proposes a trade to another, both see a trade
window, which has one side for each character, where the virtual goods or money to be given are
placed. Only after both players accept the trade does it take place. A special text chat channel for
the local geographic zone is labeled trade, and players are encouraged to use it to advertise whatever
business they want to do. However, very often players who have no business use the trade channel
for ceaseless chattering, and the Barrens local trade chat has become a venue for joking rather than
the intended economic behavior.
In a related study, Nardi et al. (2007) showed that the text chat was an effective channel through
which players could learn in three different ways: fact finding, developing tactics or strategy, and
being socialized to the norms and values that constitute game ethos. Furthermore, they argued, social
 8.2. EXAMPLES OF GUILDS 77
learning in these games was emotional as well as cognitive. Papargyris and Poulymenakou (2004)
argue that players of online multiplayer games learn lessons that can be transferred to real-world
organizations, in such areas a teamwork, negotiation, and management of projects.

8.2 EXAMPLES OF GUILDS

Table 8.1 compares three of my typical research characters and their guilds in different worlds, based
on information a member of each guild could see in the user interface on September 24, 2009.
The first thing to note is that the three guilds, and indeed the three games, are rather similar. One
difference is that Dark Age of Camelot has an experience level cap of 50, whereas the maximum is
80 in the other two. All three have multiple servers, and characters can interact only if they are on
the same server. At this point in time, World of Warcraft had 241 North American servers, 112 of
which were PvP servers where conflict between players increased the importance of belonging to a
supportive guild. Age of Conan had 8 North American servers, 4 of them PvP. Dark Age of Camelot had
effectively five “clusters” of servers, but one of them, Ywain, had over ninety percent of the players.
The majority of players had recently moved to Ywain from other clusters that had populations too
low to sustain a viable society, and about 4,000 players every evening were rapidly forming new
friendships and guilds. Gaheris was the second-most populous server and was distinctive in that the
three cultures (Albion, Midgard, and Hibernia) were at peace and could belong to the same guild.

Table 8.1: Typical Characters and Guilds in Three Games.

World Age of Dark Age of
of Warcraft Conan Camelot
Maximum Level 80 80 50
Server Earthen Ring Ironspire Gaheris
Character Sciencemag Boadicea Reitsche
Faction Horde none Midgard
Race Blood Elf Cimmerian Viking
Class hunter bear shaman valkyrie
Level 20 30 35
Guild Science Warlords Legend
Members 333 154 265
Mean Level 27.3 47.6 44.9
Top Level 7.8% 16.2% 74.0%

I created the Science guild myself for the May 2008 scientific conference I organized in World
of Warcraft with assistance from the magazine Science whose website is www.sciencemag.org, so I
named the guild master Sciencemag. After the conference, I handed control over to members who
wanted to continue, and the guild has survived for over a year. The reason the average member is
at a somewhat low level, and only 7.8 percent have reached the top level 80, is that many members
78 CHAPTER 8. SOCIAL RELATIONS INSIDE GAMES
are inactive, having joined for sake of the conference. The Warlords exist in a recent and not very
popular game, so it is not surprising that few have reached the experience ceiling. At the opposite
extreme, the Legend guild is composed overwhelmingly of top level characters because their game
was popular when it came out but is now eight years old. Indeed, dedicated Dark Age of Camelot
players tend to have multiple top-level characters, and many of them own substantial virtual villages
in the game, something also possible in Age of Conan but not in World of Warcraft.
To get a perspective on how characters with different qualities combine to form a guild,
Table 8.2 tabulates race versus class for fully 5,042 members of Alea Iacta Est in World of Warcraft,
who have reached at least level 10 of experience at the beginning of October 2009. This guild had
grown substantially since the data on professions cited above were collected a year and a half earlier,
and it is one of very largest groups in the games described here. It got its start as an adjunct of the
most popular World of Warcraft podcast, The Instance, which had transmitted fully 164 professional-
quality weekly programs by the time the new data were collected1 . While not representative in the
sense of statistical sampling, I believe this guild is actually the best for research on the culture of
World of Warcraft because the members are so knowledgeable, and the podcast functions as the news
center for the dedicated player community.

Table 8.2: Alea Iacta Est Members in World of Warcraft.

Class Blood Elf Orc Tauren Troll Undead
Death Knight 12% 21% 13% 9% 12%
Druid 0% 0% 47% 0% 0%
Hunter 14% 30% 11% 25% 0%
Mage 15% 0% 0% 11% 20%
Paladin 28% 0% 0% 0% 0%
Priest 12% 0% 0% 15% 19%
Rogue 8% 5% 0% 11% 15%
Shaman 0% 18% 15% 23% 0%
Warlock 12% 8% 0% 0% 25%
Warrior 0% 17% 14% 5% 9%
Total 100% 100% 100% 100% 100%
Cases 1,980 714 1,082 527 739

There are currently ten races of characters in the game (two more have recently been announced
for the third game expansion), and the five listed in Table 8.2 are the ones that belong to the Horde
faction. The death knight class, which was added in the second major expansion, is the only one
available to characters of all races. Notice that 28 percent of Blood Elves are paladins, and 47 percent
of Tauren are Druids, while no other races have characters in this class. Every zero in the table

1 http://www.myextralife.com/wow/
 8.3. QUANTITATIVE RESEARCH ON GUILDS 79
represents a forbidden cell, in which the indicated combination of race and class is not available
currently in the game.
Twenty-eight percent of these characters are female characters, and they are distributed across
all races and classes. Females are somewhat overrepresented among two non-violent magical classes,
priests (13 percent of female characters versus 7 percent of the males) and warlocks (13 percent versus
8). Just 3 percent of female characters are warriors, compared with 9 percent of male characters. By
far the most popular race for female characters is the Blood Elves, to which 45 percent of the females
belong, compared with 11 percent female among Orcs at the opposite extreme. The Blood Elves are
very beautiful, even angelic-looking people, and when a female Blood Elf dances, she duplicates the
moves of that popular entertainer, Britney Spears.
Races differ very little in their capabilities, other than in the classes available to them, but
their appearance and backstory lore differ greatly. Except for the Orcs and Trolls who start the game
in the same zone, each race has a different experience of the first twenty experience levels because it
begins in a different area of virtual geography. Priests are the most effective healers, whereas warriors
are the most effective “tanks” who directly engage and hold the enemy during combat. Mages and
rogues are especially good at inflicting damage on an enemy from a distance, while a warrior holds
the foe and a priest heals the warrior. I tend to think that paladins and death knights are good
characters for solo play because they combine some of the qualities of warrior and priest. However,
hunters and warlocks are also good for solo play because each of them has a secondary avatar that
can serve as a tank. It is important to realize that the members of Alea Iacta Est are characters, not
players, and each dedicated player tends to have multiple characters that can play different roles in
group activities. When a raid or questing party is being formed, one often sees players log off from
one character and log back on with a different character whose abilities are better suited to the group
challenge.

8.3 QUANTITATIVE RESEARCH ON GUILDS

Many multiplayer games permit the user to create programs, often in the Lua scripting language, that
run through the game, so long as the program does not confer a competitive advantage, and a number
of researchers have begun using such programs to collect useful data (Ducheneaut et al., 2006). For
example, Pittman and Dickey (2007) adapted the existing CensusPlus open source program to collect
data on 12,000 World of Warcraft characters in 32,000 individual playing sessions over a period of five
weeks. CensusPlus queries the interface’s “who” routine that lets players look for other players with
specified characteristics, then tabulates how many characters in different categories are currently
online, and saves the full dataset as a text file. This study chiefly looked at patterns of activity
over the hours of the day and across virtual geographic zones, finding that many characters were
online only briefly and disproportionately in certain zones. Tarng et al. (2008) did a similar study
but with the aim of predicting when players of World of Warcraft might cancel their subscriptions,
while Feng et al. (2007) used comparable methods to study the varying workload in EVE Online.
Ideally, one would want to do quantitative research with full access to the data on the host computer,
80 CHAPTER 8. SOCIAL RELATIONS INSIDE GAMES
and Kwok and Yeung (2005) showed how this could be done with a text-based gameworld called
RockyMud.
Using automatic censuses of characters online at many points in time, Ducheneaut et al. (2007)
found that the average size of World of Warcraft guilds tended to get larger over a period of months,
and that character churn (movement between guilds) tended to go down. This suggests that the
social structures of games may consolidate over time, and that guilds may serve different functions
for players at different levels of experience and with different personal goals.
On the basis of a player interview study, Williams et al. (2006) constructed a typology of
World of Warcraft guilds. Social guilds are numerous but often small collections of friends or family
members who enjoy questing together as an extension of their social relationships. In a PvP guild,
the goal is competing with other PvP guilds in a manner analogous to competition between athletic
teams. A raiding guild is like a PvP guild but large and dedicated, participating in the very large-scale
combats staged on battlegrounds. More problematic, and difficult to study because they seemed to
be rare, were role-playing guilds that supported members’ desires to become their characters. In a
quantitative study of guilds of World of Warcraft players in Taiwan, Chen et al. (2008) identified five
types of guilds in terms of the expertise and dedication of their members: newbie guilds, small but
somewhat stable guilds, very stable large guilds, elite guilds composed of high-level players, and
mid-level guilds whose progress seemed to have stalled and which declined.
 81

CHAPTER 9

Implications for External

Society
Social scientists have long believed that real-world society can be analyzed in terms of a game, and that
play is an essential part of becoming human (Von Neumann and Morgenstern, 1944; Huizinga, J.,
1949; Winnicott, D., 1971). Long, N. (1958) even argued that local communities could be described
as ecologies in which several related games play out simultaneously, such as banking, newspaper
publishing, churches, labor unions, construction contractors, manufacturing, and politics. However,
one must not rely too heavily upon a metaphor, and researchers will need to be alert to the many
ways in which online games are different from other spheres of human experience, the better to
understand both (Ducheneaut, N., 2010).

9.1 THE ONLINE GAME PENUMBRA

The first important fact to note about the relationship between multiplayer online games and the
surrounding society is that a considerable amount of online interaction takes place around the games
but outside them. I call this the penumbra of the gaming universe, the blurry edge of its shadow. The
more popular a game is, the more heavily populated its penumbra will be, and thus the more diverse
and intense the para-game activities will be.
Of course, part of this penumbra is created by the game companies themselves, in the form
of spin-off projects. Especially notable are the novels and shorter fiction stories that have been
written to flesh out the mythos of a game. The most striking example is “Of Blood and Honor”
by Metzen, C. (2002), the man primarily responsible for the lore and storyline in World of Warcraft,
which tells the tragedy of a strange friendship between the Human aristocrat Tirion Fordring and
the Orc warrior Eitrigg, in a manner that contrasts family loyalty against moral principles in a
philosophically sophisticated debate. A comparably challenging example is the online novel Prophet
Without Honour by Tørnquist, R. (2001), one of the creators of Anarchy Online. Novelettes can be
found on the website of EVE Online, and the European distributor of Dark Age of Camelot has
posted much lore in five languages. Even Star Wars Galaxies, based on a mythos for which many
novels already existed, was extended in this form (Whitney-Robinson, V., 2004).
Most popular games have their own wikis and other user-created guides. The two most
stunning examples concern the most popular MMORPG, World of Warcraft: Wowwiki the wiki
encyclopedia and WoWHead the tactics-oriented game guide. Wowwiki was originally launched in
November 2004, and by September 1 2009, it offered fully 77,187 content pages and 40,184 images,
82 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY

serving and created by 1,530,994 registered users1 . On that same date, WoWHead offered the official
text and user-created instructions for completing fully 8,093 quests. It also provided often detailed
descriptions of 1,232 status-conferring achievements, 34,166 virtual items, and 46,565 magical spells.
All the games also have forums, where players can ask questions of other players and sometimes get
good answers, and amateur websites covering one or another aspect of the game. One of the ways
to gain status outside a game but in connection to it is to post proud accomplishments on the web
in a manner that is informative to other players, for example, the stunning user-created catalog of
spaceships in EVE Online 2 .
In some cases, fans have modified existing games to add features and content. One education-
ally relevant example concerns the strategy game Rome: Total War, which can be played in a limited
multiplayer mode but is usually considered to be s solo-player game in which the user commands
Roman legions against the barbarians. Feeling that the barbarians were not depicted with as great
historical accuracy as the Romans, an online network of players created a modified version of the
game, called Europa Barbarorum – “Europe of the Barbarians” – adding content that improved the
accuracy and educational interest of the game. Although overshadowed by the great popular suc-
cess of commercial online multiplayer games, considerable creativity is shown by many networks
of enthusiasts who create and modify games following open source principles (Scacchi, W., 2004).
As Kow and Nardi (2010) have shown, there exists an active worldwide community of people who
program add-on or mod programs to enhance the user’s experience of World of Warcraft, deserving of
study in its own right.
To emphasize the fact that online role-playing games blend the human person with the
computer system, McArthur, V. (2008) has coined the term ludic cyborg. While the debate rages
about the extent to which people become psychologically immersed in virtual worlds, a very different
perspective looks at the way these environments are themselves technically and socially integrated
into the surrounding world. This is most obvious in the case of pervasive LARPs, live-action role-
playing games that are conducted in the real world but utilizing Internet for some of the actions and
communications. Pervasive computing is a term, especially popular in Europe, for what American
computer scientists calls ubiquitous computing or ubicomp. The idea is that with universal wireless or
cellphone connectivity and the convergence of all popular forms of electronic communication into
one, people will be playing online games that are integrated into their lives in a variety of ways.
Crabtree and Rodden (2008, p. 481) describe hybrid ecologies, “which marry mixed reality
environments and ubiquitous computing environments together to bridge the physical-digital divide.
Hybrid ecologies are new class of digital ecology that merge multiple environments, physical and
digital, together. Collaboration in these emerging environments is characterized by ‘fragmented
interaction’ in that it is mediated by interaction mechanisms that are differentially distributed.”
Their introductory example is a pervasive game called Uncle Roy All Around You, a research-oriented
game played in the real world but getting clues over Internet, on a quest to find a particular real-

1 http://www.wowwiki.com/WoWWiki:About
2 http://www.eveonlineships.com/
 9.2. WHAT PEOPLE LEARN IN ONLINE GAMES 83
world location. Some players are on the street, using mobile technology, whereas others are entirely
online and represented by avatars, and both use a kind of geographic information system that tracks
movements across the city. Crabtree and Rodden (2008, p. 490) suggest four processes that achieve
this hybridization:

• Media spaces LINK physical spaces through digital mediums.

• Mixed reality environments FUSE physical and digital environments.

• Ubiquitous computing environments EMBED the digital into physical environments.

• Hybrid ecologies MERGE multiple environments, physical and digital.

It would be wrong to imagine that these insights are limited to pervasive LARPS or to new
kinds of games that might be invented in the future. Lindtner et al. (2008) argue that World of
Warcraft is already part of a hybrid ecology, based on research looking at how the game is played in
China and integrated into people’s lives there, with the possibility that different national cultures
might be hospitable to different kinds of hybrid ecologies. Notably, many players in China access
the game in a kind of game-oriented cybercafé called a wang ba, where players often sit next to each
other, each operating a computer, as they quest together across virtual continents:

In the wang ba, players used a mix of digital artifacts and software tools, such as external
chat clients, in-game chat clients, mobile phone numbers, and virtual characters, as well
as resources in the wang ba to engage in a rich play experience. The “game” was not
simply the software files downloaded on a player’s machine or accessed at the wang
ba; it was, rather, a collage of artifacts and data collectively assembled by its engaged
participants. For example, for many players exchange of real life data, such as mobile
phone numbers, information about professional careers, and physical location allowed
them to express trust in players met online (Lindtner et al., 2008, p. 376).

Examples from western countries are different but numerous. I remember shortly before
the Lich King expansion was released in stores, questing in World of Warcraft while at home and
discovering from the text chat that some of my guildmates were playing on laptops using wireless
while waiting in line for the store to open so they could purchase it, naturally socializing with the
other players in line as well as on line.

9.2 WHAT PEOPLE LEARN IN ONLINE GAMES

Over the years, considerable effort has been invested in creating computer games for educational
purposes, and today much of the emphasis is on multiplayer games with the hope that students
will motivate each other. There are two fundamentally different conceptions of how a game can
be educational. First, it can be a delivery vehicle for conventional curriculum, for example, drilling
school children in the multiplication table, but there is good reason to doubt that much of the current
84 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
school curriculum will be worth delivering by whatever means to students whose lives will extend
to the very end of the twenty-first century and perhaps beyond. Second, games can teach habits
of thought, including strategies for exploration and creativity, whether or not some conventional
curriculum is also delivered at the same time.
A good example is Quest Atlantis, a well-established multiplayer game designed to help stu-
dents deepen their comprehension of issues facing our own society by helping the fictional people of
Atlantis decide how to deal with a host of social and environmental problems (Barab et al., 2007).
After achieving a good deal of success documented in many publications, its creators went a big step
further, developing a programming toolkit so teachers and students could create their own quest
narratives. This enhancement is based on two principles: “First, employing narrative to stimulate
creative work is a valuable and broadly applicable method. Second, creativity is best framed within
an active social network. This project characterizes the children’s activity in terms uniquely suited to
the new millennium: collaborative, distributed, negotiated, and critical”3 .
Based on observational research in Lineage, World of Warcraft, and Second Life, Steinkuehler, C.
(2008), (Steinkuehler and Chmiel, 2006) has identified a number of valuable kinds of learning that
players may gain even in games that were not designed with educational goals in mind. One example
is how the logical styles of thought required for science are exercised when a World of Warcraft player
decides which abilities from three alternative talent trees to give a particular character. In addition to
scientific habits of mind, Steinkuehler suggests five other modes of cognition that player may learn.
First, they gain experience in socially distributed cognition, where a network of individuals make
discoveries and decisions they could not have made alone, using complex systems of communication.
Second, they learn to collaborate to achieve practical goals. Third, they gain skills using specialized
forms of language, not unlike those employed in science-based professions but also developing
narrative skills. Fourth, handling the games’ complex user interfaces and online information resources,
their computer literacy improves. Finally, they learn general mechanisms that may lead to success in
many real-world contexts, such as what Steinkuehler calls reciprocal apprenticeship in which people
tutor each other by sharing information and actively debating its meaning.
Playing online multiplayer games may enhance the user’s understanding of cultural diversity,
both because the games themselves are cultural systems and because other players come from dif-
ferent real-world cultures. A remarkable feature of MMORPGs and other virtual worlds is how
international they are. The offices of World of Warcraft are in Irvine, California, but it is owned by
a transnational corporation headquartered in France, and only about a quarter of the players live in
the United States. Age of Conan is based on stories written in Texas but was created by a Norwegian
company. Entropia Universe hails from Sweden, and EVE Online is an Icelandic production. Korea
is among the most prominent nations producing MMORPGs, and others were created in Japan and
China. To the extent that games aim for an international audience, they intentionally blend cultural
elements from diverse sources. In so doing, and by using Internet, they bring people together across
cultural and national boundaries. I remember one time when I helped attack the Horde city, Or-

3 http://www.nsf.gov/awardsearch/showAward.do?AwardNumber=0855852
 9.2. WHAT PEOPLE LEARN IN ONLINE GAMES 85
grimmar, mentally in World of Warcraft but physically in the United States, knowing that the warrior
beside me was in Canada, and the married couple leading the charge about twenty feet ahead of us
were in England.
The cultural fusion becomes especially interesting when players speak different native lan-
guages. Standing in front of a weapon store in Tabula Rasa, you would see the sign cycle between
English, French, German, and Korean, and you could download patches that would make the non-
player characters and user interface employ any of these four languages. Some users of the non-game
virtual world, Second Life, use automatic language translation systems. I have seen a Brazilian type
something into the text chat in Portuguese, after which a bot would immediately render it into rough
English, after which the Brazilian with some knowledge of English would make corrections. The
Japanese game Final Fantasy XI was designed to be played by people who lacked a common lan-
guage, so it incorporated an extensive list of phrases that could be selected by one member of a team
and would display in the various correct languages to other members of the team. Interestingly, the
players themselves then developed a set of additional words adapted from various languages, what is
technically called a pidgin, to supplement the stock phrase list (Nolen, C., 2007). Players who want
to improve their facility with one of the world’s major languages can now play an online game in that
language, so long as they recognize that the vocabulary may not be entirely standard (Rankin et al.,
2008).
Like the great classic works of literature, online games draw upon existing culture and influence
the player’s perception of the real world, for example, serving as a metaphor for resource competition
between nations, and often functioning as a political allegory. Some argue that these games support
the norms and values of the surrounding culture. For example, Rettberg, S. (2008, p. 20) writes:

World of Warcraft is both a game and a simulation that reinforces the values of Western
market-driven economies. The game offers its players a capitalist fairytale in which
anyone who works hard and strives enough can rise through society’s ranks and acquire
great wealth. Moreover, beyond simply representing capitalism as good, World of Warcraft
serves as a tool to educate its players in a range of behaviors and skills specific to the
situation of conducting business in an economy controlled by corporations.

However, it could just as well be said that World of Warcraft is explicitly critical of capitalism
and of political practices related to it. Many of the quests reveal the elite to be selfish liars, for
example, when the rulers of Stormwind send teams of players into the Deadmines to kill Edwin
VanCleef, leader of the so-called Defias Bandits. Only when the player brings his severed head
back to Stormwind, is it revealed that he was actually the honorable leader of a political movement
to seek rights for exploited workers. Another unjust quest requires the assassination of Colonel
Kurzen, a renegade military officer who has broken away from Stormwind’s Alliance and is living
in the Stranglethorn Vale jungle with his private army. He is based on the character Colonel Kurtz
in the movie Apocalypse Now that satirized the Vietnam War, who was based in turn on Kurtz in
Joseph Conrad’s novella “Heart of Darkness” that criticized European colonialism in Africa.
86 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
Players who are alert to such niceties will experience World of Warcraft as a critique of capitalist
imperialism, not a defense of it (Langer, J., 2008). Expanding the range of the game’s critique, the
second expansion late in 2008 added several missions that require torturing prisoners, thus making
the player think about proper interrogation techniques in the real-world War on Terror and slaughter
peaceful whales while listening to them sing to each other. Since its very beginning in 2004, the game
has included many quests incorporating the value of the environmental movement. For example,
Figure 9.1 shows two scenes in Windshear Crag, where capitalist gnomes of the Venture Company
have chopped down a beautiful forest and made a wasteland, using harmful technologies to fuel their
greed. On the left, we see Ozma battling a deforester, with a huge earth-wrecking machine in the
background. On the right, Catullus battles a gnome operating an XT-9 logger.

Figure 9.1: Chopping Down Deforesters in World of Warcraft.

It is an open question whether most players understand the alegories embedded in many of
the games, not to mention the literary references or philosophical principles. Many of the most
popular games, including World of Warcraft, depict warring societies, none of which have a monopoly
on morality. It is entirely possible that these games teach people that loyalty is only valuable when
it is expedient, that stealing other people’s resources is entirely appropriate, and that the world is a
battleground where selfish coalitions compete for survival. A few games lack competing factions of
players and teach a different morality, notably Tabula Rasa and Lord of the Rings Online, but it is a
discouraging fact that these idealistic games are less popular.
 9.3. RESEARCH OPPORTUNITIES 87

9.3 RESEARCH OPPORTUNITIES

Many social scientists have begun conducting research in online games and in non-game virtual
worlds, on the premise that social behavior is much easier to observe in these computerized environ-
ments, yet is comparable to behavior in the outside world (Bainbridge, W., 2007). Whether concep-
tualized as laboratories or as new regions of the wider human world, they can be used to develop and
test theories of human interaction, psychology, and culture. For example, economist Castronova, E.
(2005a,b) argues both that an increasing fraction of human economy will take place in virtual envi-
ronments, so they need to be studied as important phenomena in their own right, and that it can be
fruitful to compare results from research in different virtual worlds that follow somewhat different
economic principles.
Writing in a medical journal, Lofgren and Fefferman (2007), explained how an unintended
disease epidemic in World of Warcraft showed how serious epidemiologists could use a virtual world
to model the actual spread of infectious diseases. Many games represent disease or poisoning as
temporary and often progressive degradation of the character’s ability, that under adverse conditions
can lead to death. Usually, these virtual diseases are not contagious, but in September 2005, World
of Warcraft included an infectious disease called “corrupted blood” in a new 20-man raid instance
called Zul’Gurub. Given that the teams that enter this instance are fairly large, it must have seemed
interesting to allow them to contract a disease from each other, as healer characters attempted to
cure their companions even as the infection spread among them. The designers believed they had
constructed the instance in such as way that would prevent the infection from spreading outside, but
perhaps not enough beta testing had been done to evaluate this hope. Furthermore, while the illness
was seldom life-threatening to the level 58-60 characters who could enter the instance, it was often
fatal to lower-level characters once it had escaped the confines of the instance.
Leaving Zul’Gurub by the normal route rendered the disease inactive, but apparently the
designers had failed to notice that players could teleport out in different ways, and the new pro-
gramming code did not deactivate the disease for all of them. Also, the hunting animals used by
hunter characters could acquire and transmit the disease. To prevent their animals from dying,
hunters would often deactivate these secondary avatars, and the code did not cure deactivated an-
imals. Apparently, non-player shopkeepers in the towns and cities also contracted and spread the
disease, but they usually were unaffected by it. So, on one level, the epidemic was the result of an
innovative game idea, the consequences of which were not fully worked out in the programming.
The result was an epidemic that may have infected thousands of players’ characters per hour.
Lofgren and Fefferman (2007, p. 627) argue that proper collection of data during a virtual
epidemic such as this could explore social factors that operate with real-world plagues, but at far less
human cost:
Whereas the epidemic of Corrupted Blood within World of Warcraft was the result of
unintended interactions between different elements of the game, it nevertheless shows
the potential of such scenarios for the study of infectious disease. One of the major con-
straints in studies of disease dynamics in animals is that epidemiologists are restricted
88 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
largely to observational and retrospective studies. In nearly every case, it is physically
impossible, financially prohibitive, or morally reprehensible to create a controlled em-
pirical study where the parameters of the disease are already known before the course of
epidemic spread is followed.

A team of researchers at UCLA (Kafai et al., 2007) has actually conducted an experiment
by introducing an infectious disease inside the educational children’s virtual world, Whyville. The
symptoms of Whypox were mild: red pimples appeared on the avatars, and they started sneezing
“achoo” into the text chat. This was done primarily for educational purposes, as the children could
track the spread of the disease, post theories about it, and predict its course.The team used a spectrum
of research methods to study what happened:

(1) Log files that recorded all Whyville-based actions of consenting Whyville partici-
pants, including information about locations visited, time spent there, and chat content
(∼70 million data points); (2) online surveys (pre- and post-epidemic) that asked partic-
ipants (with a combination of multiple choice and open-ended items) about their science
and technology interests, understanding of infectious disease, and experiences and pref-
erences in Whyville activities; (3) field notes and video recordings of classroom students
and after-school club participants while on Whyville; (4) face-to-face interviews with
selected participants about their Whyville interactions, and (5) embedded ethnography
that chronicled Whyville community life before, during, and after the virtual epidemic.

Like a real disease, Whypox impeded activities. The sneezing in the text chat corrupted some
commands and interrupted conversations.The students learned about the course of an epidemic from
an information resource that graphed the cases, rising over three days to infect 4,000 characters, then
dropping to a low infection rate after most of the population pool had already contracted it. Students
tended to generalize excessively from their own personal experience in describing and explaining
the course of the disease, rather than thinking scientifically like professional epidemiologists. Their
language in describing Whypox in the text chat tended to emphasize symptoms; perhaps, because
there were so many different ways they conceptualized other aspects of the disease, such as its
transmission modalities, they employed a wide variety of words. Some of the children actually faked
the infection, by intentionally sneezing in the text chat. The educational experience seems to have
been quite rich as the students shifted from a concern for their own health or illness to an awareness
of what was happening to other people to the development of over-arching insights about disease
transmission in general.
Bradley and Froomkin (2004) have argued that games and other virtual worlds could be
used as laboratories for testing laws and regulations before they are applied in the outside world.
Three and four decades ago, real-world social-science experiments were often used to pre-test new
government policies before they were enacted on a wide scale, but politicians were disappointed
with the results and such studies are far less often done today. The classic example was the 1976
Transitional Aid Research Project (TARP) that at great cost experimented with different programs
 9.3. RESEARCH OPPORTUNITIES 89
to assist prisoners released from jail adapt to society (Rossi et al., 1980, 1982; Rossi and Wright,
1984). One pessimistic reading of the findings was that providing short-term financial support and
help in finding a job failed to reduce the chance a prisoner would return to a life of crime (Zeisel, H.,
1982a,b). The researchers on the project themselves were split, some feeling the study showed that
such programs do no good, and others using elaborate statistical techniques to try to find hidden
benefits. I can imagine replicating this study inside a multiplayer online game, imprisoning players
for being excessively violent, then providing some of them with training and resources to take up
a peaceful profession, and seeing if they remain violent. To be sure, we might not want to base
real-world government politics entirely on the results of gaming experiments, but they could help
strengthen the social science relevant to those policies.
Violence inside games may have implications for violence outside them. In an experimental
study that compared new players of Asheron’s Call 2 with a control group, Williams, D. (2006) found
some evidence that players became more sensitive to real-world dangers, but only if the dangers
were similar to those faced in the game. Massively multiplayer online games put intense pressures on
players to join into groups and competing factions, and most of them discourage neutrality in the face
of social conflict (Medler, B., 2008). Logically enough, the game designers think that conflict builds
group loyalty, and group loyalty commits the player to the particular game, rather than hopping from
one to another as soon as each character reaches the maximum experience level. The social scientific
study most relevant to this issue is actually a very old one, the famous Robber’s Cave Experiment
carried out by a team headed by Sherif et al. (1988). Twenty-two boys attended a special summer
camp in 1954, where they were divided into two competing groups that called themselves Rattlers
and Eagles and developed strong in-group solidarity and out-group hostility by competing intensely
against each other.
Multiplayer online games can become laboratories where humanity experiments with new
norms, values, and institutions that might later be transferred to the outside world. Whether these
innovations will be new forms of conflict or cooperation remains to be seen. Castronova, E. (2007)
has suggested that the economic systems of popular games may change people’s sense of justice in the
real world, for example, causing them to reject inheritance of wealth on the principle that everybody
should start the game of life with the same resources.
The quests in these games could enhance the wider human quest for utopia, as one of them
explicitly claims: “The objective of A Tale in the Desert is to build the ideal civilization by perfecting
the Seven Disciplines of Man, with each discipline containing seven tests”4 . Many of the forty-nine
tests require people to cooperate, but so do many of the quests in other games, whose guilds prototype
communities of the future. At the risk of ending this lecture about fantasies on an especially fanciful
note, these online multiplayer games may be etopias – electronic utopias that have the potential to be
the proving grounds for real-world social innovation and are already meeting places for real-world
subcultures and social movements (Bainbridge, W., 2009).

4 http://atitd.org/wiki/tale4/Tests
 91

Bibliography
Abbott, H. Porter. 2003. “Unnarratable Knowledge: The Difficulty of Understanding Evolution by
Natural Selection.” Pp. 143–162 in Narrative Theory and the Cognitive Sciences, edited by David
Herman. Stanford, California: CSLI Publications. 1.3

Abbott, H. Porter. 2008. The Cambridge Introduction to Narrative. New York: Cambridge University
Press. 1.3

Achterbosch, L., Robyn Pierce, and Gregory Simmons. 2007. “Massively Multiplayer Online Role-
Playing Games: The Past, Present, and Future,” Computers in Entertainment 5(4). 1.4, 3

Adorno, T., Else Frenkel-Brunswik, Daniel Levinson, and Nevitt Sanford. 1950. The Authoritarian
Personality. New York: Harper. 5.2

Aggarwal, S., Justin Christofoli, Sarit Mukherjee, and Sampath Rangarajan. 2006. “Authority
Assignment in Distributed Multi-Player Proxy-Based Games.” In Proceedings of 5th ACM
SIGCOMM Workshop on Network and System Support for Games. New York: ACM.
DOI: 10.1145/1230040.1230068 3.1

Arya, Ali, and Steve Di Paola. 2007. “Multispace Behavioral Model for Face-Based Af-
fective Social Agents,” EURASIP Journal on Image and Video Processing, article 48757.
DOI: 10.1155/2007/48757 5.4

Bainbridge, William Sims. 1986. Dimensions of Science Fiction. Cambridge, Massachusetts: Harvard
University Press. 2.5

Bainbridge, William Sims. 2007. “The Scientific Research Potential of Virtual Worlds,” Science, 317
(27 July): 472–476. DOI: 10.1126/science.1146930 9.3

Bainbridge, William Sims. 2009. “Etopia,” Networker 13(1): 36–35.

DOI: 10.1145/1516035.1516045 9.3

Bainbridge, William Sims. 2010a.“Science,Technology and Reality in The Matrix Online and Tabula
Rasa.” In Online Worlds, edited by William Sims Bainbridge. Guildford, Surrey, England: Springer.
1.2

Bainbridge, William Sims. 2010b.The Warcraft Civilization. Cambridge, Massachusetts: MIT Press.
1
92 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
Bainbridge, William Sims. 2010c. “When Virtual Worlds Expand.” In Online Worlds, edited by
William Sims Bainbridge. Guildford, Surrey, England: Springer. 1, 4.2

Bainbridge, William Sims, and Wilma Alice Bainbridge. 2007a. “Electronic Game Research
Methodologies: Studying Religious Implications,” Review of Religious Research 49: 35–53. 2.4

Bainbridge, Wilma Alice, and William Sims Bainbridge. 2007b. “Creative Uses of
Software Errors: Glitches and Cheats,” Social Science Computer Review 25: 61–77.
DOI: 10.1177/0894439306289510 3.3

Barab, Sasha, Tyler Dodge, Hakan Tuzun, Kirk Job-Sluder, Craig Jackson, Anna Arici, Laura Job-
Sluder, Robert Carteaux, Jr., Jo Gilbertson, and Conan Heiselt. 2007. “The Quest Atlantis Project:
A Socially-Responsive Play Space for Learning.” Pp. 159-186 in The Educational Design and Use of
Simulation Computer Games, edited by B. E. Shelton and D. Wiley. Rotterdam, The Netherlands:
Sense Publishers. 9.2

Bardzell, Shaowen, Jeffrey Bardzell, Tyler Pace, and Kayce Reed. 2008. “Blissfully Productive:
Grouping and Cooperation in World of Warcraft Instance Runs.” Pp. 357-360 in Proceed-
ings of the ACM 2008 Conference on Computer Supported Cooperative Work. New York: ACM.
DOI: 10.1145/1460563.1460621 4.2

Barnett, Jane , Mark Coulson, and Nigel Foreman. 2010. “Examining Player Anger in World of
Warcraft.” In Online Worlds, edited by William Sims Bainbridge. Guildford, Surrey, England:
Springer. 4.2

Barr, Pippin, Robert Biddle, and Judy Brown. 2006. “Changing the Virtual Self: Avatar Transforma-
tions in Popular Games.” Pp. 83-90 in Proceedings of the 3rd Australasian Conference on Interactive
Entertainment. Perth, Western Australia, Australia: Murdoch University. 6.2

Bartle, Richard A. 1996. “Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs,” Journal of
MUD Research 1; online at
http://www.brandeis.edu/pubs/jove/HTML/v1/bartle.html;
retrieved, November 15, 2008. 5.2

Bartle, Richard A. 2004. Designing Virtual Worlds. Indianapolis, Indiana: New Riders. 5.2

Baudrillard, Jean. 1994. Simulacra and Simulation. Ann Arbor, Michigan: University of Michigan
Press. 1.2

Bedman (aka Captain Stack). 2007. “The Matrix Online for PC: Windows;”
http://www.gamefaqs.com/computer/doswin/file/931849/35989 5.2

Bell, Daniel (ed.). 1963. The Radical Right. Garden City, New York: Doubleday. 5.2
 9.3. RESEARCH OPPORTUNITIES 93
Beskow, Paul B., Knut-Helge Vik, Pål Halvorsen, and Carsten Griwodz. 2008. “Latency Reduction
by Dynamic Core Selection and Partial Migration of Game State.” Pp. 79-84 in Proceedings of
the 7th ACM SIGCOMM Workshop on Network and System Support for Games. New York: ACM.
DOI: 10.1145/1517494.1517511 3.2
Bessiere, Katherine, Fleming Seay, and Sara Kiesler. 2007. “The Ideal Elf: Identity Exploration in
World of Warcraft.” CyberPsychology and Behavior, 10: 53-535. 6.2
Blinka, Lukas. 2008. “The Relationship of Players to their Avatars in MMORPGs: Differences be-
tween Adolescents, Emerging Adults and Adults,” Cyberpsychology: Journal of Psychosocial Research
on Cyberspace, 1, online at
http://www.cyberpsychology.eu/view.php?cisloclanku=2008060901;
retrieved November 15, 2008. 6.2
Bohannon, John. 2008a. “Scientists Invade Azeroth,” Science, 320:1592. 3.1
Bohannon, John. 2008b. “Slaying Monsters for Science,” online version of Science,
http://www.sciencemag.org/cgi/content/full/320/5883/1592c,
retrieved November 19, 2008. 3.1
Bosser, Anne-Gwenn. 2004. “Massively Multi-Player Games: Matching Game Design with Techni-
cal Design.” Pp. 263-268 in Proceedings of the 2004 ACM SIGCHI International Conference on Ad-
vances in Computer Entertainment Technology. New York: ACM. DOI: 10.1145/1067343.1067378
3.1
Bradley, Caroline, and A. Michael Froomkin. 2004. “Virtual Worlds, Real Rules,” New York Law
School Law Review, 49:103-146. 9.3
Browning, David, Steven Stanley, Michael Fryer, and Nicola J. Bidwell. 2006. “Emplacing Ex-
perience.” Pp. 96-103 in Proceedings of the 2006 International Conference on Game Research and
Development. Perth, Western Australia: Murdoch University. 3.3
Castronova, Edward. 2003. “Network Technology, Markets, and the Growth of Synthetic Worlds.”
Pp. 121-134 in Proceedings of the 2nd Workshop on Network and System Support for Games. New
York: ACM. DOI: 10.1145/963900.963912 2
Castronova, Edward. 2005a. “On the Research Value of Large Games: Natural Experiments in
Norrath and Camelot,” CESifo Working Paper Series No. 1621, Indiana University Bloomington.
DOI: 10.1177/1555412006286686 9.3
Castronova, Edward. 2005b. Synthetic Worlds: The Business and Culture of Online Games. Chicago:
University of Chicago Press. 9.3
Castronova, Edward. 2007. Exodus to the Virtual World: How Online Fun is Changing Reality. New
York: Palgrave Macmillan. 9.3
94 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
Catullus. 2008. “Letter to a Supernatural Being.” Pp. 247-255 in Human Futures: Art in and Age of
Uncertainty, edited by Andy Miah. Liverpool, England: Liverpool University Press. 6.1

Chen, Chien-Hsun, Chuen-Tsai Sun, and Jilung Hsieh. 2008. “Player Guild Dynamics and Evo-
lution in Massively Multiplayer Online Games,” CyberPsychology and Behavior, 11:293-301.
DOI: 10.1089/cpb.2007.0066 8.3

Chen, Kuan-Ta, Chun-Ying Huang, Polly Huang, and Chin-Laung Lei. 2006. “An Empirical
Evaluation of TCP Performance in Online Games.” Proceedings of the 2006 ACM SIGCHI
International Conference on Advances in Computer Entertainment Technology. New York: ACM.
DOI: 10.1145/1178823.1178830 3.1

Chen, Vivian Hsueh-hua, Henry Been-Lirn Duh, and Hong Renyi. 2008. “The Changing Dynamic
of Social Interaction in World of Warcraft:The Impacts of Game Feature Change.” Pp. 356-359 in
Proceedings of the 2008 International Conference on Advances in Computer Entertainment Technology.
New York: ACM. DOI: 10.1145/1501750.1501834 4.2

Cikic, Sabine, Sven Grottke, Fritz Lehmann-Grube, and Jan Sablatnig. 2008.“Cheat Prevention and
Analysis in Online Virtual Worlds.” In Proceedings of the 1st International Conference on Forensic
Applications and Techniques in Telecommunications, Information, and Multimedia. New York: ACM.
4.2

Cornett, Steve. 2004. “The Usability of Massively Multiplayer Online Roleplaying Games: Design-
ing for New Users.” Pp. 703-710 in Proceedings of the SIGCHI Conference on Human Factors in
Computing Systems. New York: ACM. DOI: 10.1145/985692.985781 3.3

Crabtree, Andy, and Tom Rodden. 2008.“Hybrid Ecologies: Understanding Cooperative Interaction
in Emerging Physical-Digital Environments.” Personal and Ubiquitous Computing, 12:481–493.
DOI: 10.1007/s00779-007-0142-7 9.1

Craft, Ashley John. 2007.“Sin in Cyber-Eden: Understanding the Metaphysics and Morals of Virtual
Worlds,” Ethics and Information Technology, 9:205-217. DOI: 10.1007/s10676-007-9144-4 4.1

De Camp, L. Sprague, Catherine Crook de Camp, and Jane Whittington Griffin. 1983. Dark Valley
Destiny: The Life of Robert E. Howard. New York: Bluejay. 2.3

Dibbell, Julian. 2007. “The Life of the Chinese Gold Farmer,” New York Times online, June 17, 2007;
http://www.nytimes.com/2007/06/17/magazine/17lootfarmers-t.html 4.4

Ducheneaut, Nicolas. 2010. “Massively Multiplayer Online Games as Living Laboratories: Oppor-
tunities and Pitfalls.” In Online Worlds, edited by William Sims Bainbridge. Guildford, Surrey,
England: Springer. 9
 9.3. RESEARCH OPPORTUNITIES 95
Ducheneaut, Nicolas, and Robert J. Moore. 2004. “The Social Side of Gaming: A Study of
Interaction Patterns in a Massively Multiplayer Online Game.” Pp. 360-369 in Proceedings
of the 2004 ACM Conference on Computer Supported Cooperative Work. New York: ACM.
DOI: 10.1145/1031607.1031667 2.5
Ducheneaut, Nicolas, Robert J. Moore, and Eric Nickell. 2007. “Virtual ‘Third Places:’ A Case Study
of Sociability in Massively Multiplayer Games,” Computer Supported Cooperative Work 16:129-166.
DOI: 10.1007/s10606-007-9041-8 2.5
Ducheneaut, Nicolas, Nick Yee, Eric Nickell, and Robert J. Moore. 2006. “Building an MMO
With Mass Appeal: A Look at Gameplay in World of Warcraft,” Games and Culture 1:281-317.
DOI: 10.1177/1555412006292613 8.3
Ducheneaut, Nicolas, Nick Yee, Eric Nickell, and Robert J. Moore. 2007. “The Life and Death of
Online Gaming Communities: A Look at Guilds in World of Warcraft.” Pp. 839-848 in Proceedings
of the SIGCHI Conference on Human Factors in Computing Systems. New York: ACM. 8.3
Ducheneaut, Nicolas, Ming-Hui Wen, Nicholas Yee, and Greg Wadley. 2009. “Body and Mind:
A Study of Avatar Personalization in Three Virtual Worlds.” Pp. 1151-1160 in Proceedings of
the 27th International Conference on Human Factors in Computing Systems. New York: ACM.
DOI: 10.1145/1518701.1518877 6.2
Durkheim, Emile. 1933. The Division of Labor in Society. New York: Free Press. 7.3
Dyck, Jeff, Carl Gutwin, T.C. Nicholas Graham, and David Pinelle. 2007. “Beyond the LAN:
Techniques from Network Games for Improving Groupware Performance.” Pp. 291-300 in Pro-
ceedings of GROUP’07. New York: ACM. DOI: 10.1145/1316624.1316669 3.1
Feng, Wu-chang, David Brandt, and Debanjan Saha. 2007. “A Long-term Study of a Popular
MMORPG.” Pp. 19-24 in Proceedings of the 6th ACM SIGCOMM Workshop on Network and
System Support for Games. New York: ACM. 8.3
Foo, Chek Yang, and Elina M.I. Koivisto. 2004. “Defining Grief Play in MMORPGs: Player
and Developer Perceptions.” Pp. 245-250 in Proceedings of the 2004 ACM SIGCHI Inter-
national Conference on Advances in Computer Entertainment Technology. New York: ACM.
DOI: 10.1145/1067343.1067375 4.2
Frasca, Gonzalo. 2003.“Simulation versus Narrative: Introduction to Ludoloogy.” Pp. 221-235 in The
Video Game Theory Reader, edited by Mark J.P. Wolf and Bernard Perron. New York: Routledge.
Freud, Sigmund. 1930. Civilization and its Discontents. New York: Cape and Smith.
Fritsch, Tobias, Benjamin Voigt, and Jochen Schiller. 2006. “Distribution of Online Hardcore Player
Behavior: (How Hardcore Are You?).” Proceedings of 5th ACM SIGCOMM Workshop on Network
and System Support for Games. New York: ACM. DOI: 10.1145/1230040.1230082 3.3
96 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
Fritsch, Tobias, Jochen Schiller, and Benjamin Voigt. 2007. “Personal Behavior and Virtual Frag-
mentation.” Pp. 60-63 in Proceedings of the International Conference on Advances in Computer
Entertainment Technology. New York: ACM. DOI: 10.1145/1255047.1255059 3.3

Fron, Janine, Tracy Fullerton, Jacquelyn Ford Morie, and Celia Pearce (aka Ludica). 2007. “Playing
Dress-Up: Costume, roleplay and imagination.” Philosophy of Computer Games Online Proceedings;
http://lcc.gatech.edu/˜{}cpearce3/PearcePubs/LudicaDress-Up.pdf 6.2

Gordon, Andrew. 1995. “Star Wars: A Myth for Our Time.” Pp. 73-82 in Screening the Sacred:
Religion, Myth and Ideology in Popular American Film, edited by Joel W. Martin and Conrad E.
Ostwalt. Boulder, Colorado: Westview Press. 2.5

Grimes, Sara M. 2006. “Online Multiplayer Games: A Virtual Space for Intellectual Property De-
bates?” New Media and Society 8:969-990. DOI: 10.1177/1461444806069651 4.4

Griwodz , Carsten, and Pål Halvorsen. 2006. “The Fun of Using TCP for an MMORP.” In Pro-
ceedings of the 2006 International Workshop on Network and Operating Systems Support for Digital
Audio and Video. New York: ACM. DOI: 10.1145/1378191.1378193 3.1

Gygax, Gary. 1979. Advanced Dungeons and Dragons, Dungeon Masters Guide. New York:
TSR/Random House. 2.5

Halloran, John, Geraldine Fitzpatrick, Yvonne Rogers, and Paul Marshall. 2004. “Does It Matter
if You Don’t Know Who’s Talking?: Multiplayer Gaming with Voiceover IP.” Pp. 1215-1218 in
Extended Abstracts of CHI ’04 Conference on Human Factors in Computing Systems. New York: ACM.
DOI: 10.1145/985921.986027 3.3

Heise, David R. 2004. “Enculturating Agents With Expressive Role Behavior.” Pp. 127-142 in Agent
Culture: Human-Agent Interaction in a Multicultural World. Mahwah, New Jersey: Erlbaum. 5.4

Holsapple, Clyde W., and Jiming Wu. 2007. “User Acceptance of Virtual Worlds: The
Hedonic Framework,” DATA BASE for Advances in Information Systems 38:86-89.
DOI: 10.1145/1314234.1314250 5.2

Homans, George C. 1950. The Human Group. New York: Harcourt, Brace. 7.3

Huh, Searle, and Dmitri Williams. 2010. “Dude Looks Like a Lady: Gender Swapping in an
Online Game.” In Online Worlds, edited by William Sims Bainbridge. Guildford, Surrey, England:
Springer. 6.2

Huizinga, Johan. 1949. Homo Ludens: A Study of the Play-Element in Culture. London: Routledge
and Kegan Paul. 9

Humphreys, Sal. 2009. “Norrath: New Forms, Old Institutions.” Game Studies 9(1);
http://gamestudies.org/0901/articles/humphreys 4.4
 9.3. RESEARCH OPPORTUNITIES 97
Hussain, Zaheer, and Mark D. Griffiths. 2008. “Gender Swapping and Socializing in Cyberspace:
An Exploratory Study,” CyberPsychology and Behavior 11:47-53. DOI: 10.1089/cpb.2007.0020
6.2

Jensen-Campbell, Lauri A., Jennifer M. Knack, Amy M. Waldrip, and Shaun D. Campbell. 2007.
“Do Big Five Personality Traits Associated with Self-control Influence the Regulation of Anger
and Aggression?” Journal of Research in Personality, 41: 403–424. DOI: 10.1016/j.jrp.2006.05.001
5.1

Jensen-Campbell, Lauri A., and Kenya T. Malcolm. 2007. “The Importance of Conscientiousness in
Adolescent Interpersonal Relationships,” Personality and Social Psychology Bulletin, 33: 368-383.
DOI: 10.1177/0146167206296104 5.1

Jonsson, Staffan, Markus Montola, Annika Waern, and Martin Ericsson. 2006. “ Prosopopeia:
Experiences from a Pervasive LARP.” In Proceedings of the 2006 ACM SIGCHI Inter-
national Conference on Advances in Computer Entertainment Technology. New York: ACM
DOI: 10.1145/1178823.1178850 1.1

Jørgensen, Kristine. 2008. “Audio and Gameplay: An Analysis of PvP Battlegrounds in World of
Warcraft,” Game Studies 8(2);
http://gamestudies.org/0802/articles/jorgensen 3.3

Kafai, Yasmin B., David Feldon, Deborah Fields, Michael Giang, and Maria Quintero. 2007. “Life in
the Time of Whypox: A Virtual Epidemic as a Community Event.” Pp. 171-190 in Communities
and Technologies, edited by C. Steinfield, B. Pentland, M. Ackerman, and N. Contractor. New
York: Springer. 9.3

Kinnard, Roy. 1988. “The Flash Gordon Serials,” Films in Review 39(4):194-203. 2.5

Klastrup, Lisbeth. 2006. “Death Matters: Understanding Gameworld Experiences.” In Proceedings of

the 2006 ACM SIGCHI International Conference on Advances in Computer EntertainmentTechnology.
New York: ACM. DOI: 10.1145/1178823.1178859 6.4

Klastrup, Lisbeth. 2008. “What Makes World of Warcraft a World? A Note on Death and Dying.”
Pp. 143-166 in Digital Culture, Play and Identity: A World of Warcraft Reader, edited by Hilde G.
Corneliussen and Jill Walker Rettberg. Cambridge, Massachusetts: MIT Press. 6.4

Klastrup, Lisbeth. 2009. “The Worldness of EverQuest: Exploring a 21st Century Fiction,” Game
Studies 9(1);
http://gamestudies.org/0901/articles/klastrup 1.3

Kow, Yong Ming, and Bonnie Nardi. 2010. “Culture and Creativity: World of Warcraft Modding
in China and the U.S.” In Online Worlds, edited by William Sims Bainbridge. Guildford, Surrey,
England: Springer. 9.1
98 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
Krzywinska, Tanya. 2006. “Blood Sythes, Festivals, Quests, and Backstories,” Games and Culture
1:383-396. 2
Kshirsagar, Sumedha, and Nadia Magnenat-Thalmann. 2002. “A Multilayer Personality Model.”
Pp. 107-115 in Proceedings of the Symposium on Smart Graphics. New York: ACM.
DOI: 10.1145/569005.569021 5.4
Kwok, Michael, and Gary Yeung. 2005. “Characterization of User Behavior in a Multi-Player Online
Game.” Pp. 69-74 in Proceedings of the 2005 ACM SIGCHI International Conference on Advances
in Computer Entertainment Technology. New York: ACM. DOI: 10.1145/1178477.1178486 8.3
Langer, Jessica. 2008. “The Familiar and the Foreign: Playing (Post)Colonialism in World of War-
craft.” Pp. 87-108 in Digital Culture, Play and Identity: A World of Warcraft Reader, edited by Hilde
G. Corneliussen and Jill Walker Rettberg. Cambridge, Massachusetts: MIT Press. 9.2
Lastowka, Greg. 2009. “Planes of Power: EverQuest as Text, Game and Community.” Game Studies
9(1);
http://gamestudies.org/0901/articles/lastowka 4.4
Li, Kang, Shanshan Ding, Doug McCreary, and Steve Webb. 2004. “Analysis of State Exposure
Control to Prevent Cheating in Online Games.” Pp. 140-145 in Proceedings of the 14th International
Workshop on Network and Operating Systems Support for Digital Audio and Video. New York: ACM.
DOI: 10.1145/1005847.1005878 4.2
Lindtner, Silvia, Bonnie Nardi, Yang Wang, Scott Mainwaring, He Jing, Wenjing Liang. 2008.
“A Hybrid Cultural Ecology: World of Warcraft in China.” Pp. 371-382 in Proceedings of
the ACM 2008 Conference on Computer Supported Cooperative Work. New York: ACM.
DOI: 10.1145/1460563.1460624 9.1
Lindzey, Gardner, and Elliot Aronson (eds.). 1968. The Handbook of Social Psychology. Reading,
Massachusetts: Addison-Wesley. 5.1
Lipset, Seymour Martin, and Earl Raab. 1970. The Politics of Unreason: Right-wing Extremism in
America. New York: Harper and Row. 5.2
Lofgren, Eric T., and Nina H. Fefferman. 2007. “The Untapped Potential of Virtual Game
Worlds to Shed Light on Real World Epidemics,” The Lancet Infectious Diseases, 7: 625-629.
DOI: 10.1016/S1473-3099(07)70212-8 9.3
Long, Norton E. 1958. “The Local Community as an Ecology of Games.” American Journal of
Sociology 64: 251-261. DOI: 10.1086/222468 9
Lu, Fengyun, Simon Parkin, and Graham Morgan. 2006.“Load Balancing for Massively Multiplayer
Online Games.” Proceedings of 5th ACM SIGCOMM Workshop on Network and System Support for
Games. New York: ACM. DOI: 10.1145/1230040.1230064 3.2
 9.3. RESEARCH OPPORTUNITIES 99
Lummis, Michael, and Edwin Kern. 2006. World of Warcraft Master Guide. Indianapolis, Indiana:
BradyGAMES/DK. 1.3

Lupoff, Richard A. 1976. Edgar Rice Burroughs and the Martian Vision. Westminster, Maryland:
Mirage Press. 2.5

Maher, Mary Lou, Kathryn Merrick, and Owen Macindoe. 2005.“Can Designs Themselves Be Cre-
ative?” Pp. 11-135 in Proceedings of the Sixth International Roundtable Conference on Computational
and Cognitive Models of Creative Design, Heron Island, Australia. 1.4

Massey, Douglas S. 2002. “A Brief History of Human Society: The Origin and Role of Emotion in
Social Life.” American Sociological Review 61:1-29. 2.3

McArthur, Victoria. 2008. “World of Warcraft as a Ludic Cyborg.” Pp. 264-265 in Pro-
ceedings of the 2008 Conference on Future Play: Research, Play, Share. New York: ACM.
DOI: 10.1145/1496984.1497046 9.1

McClelland, David C. 1961. The Achieving Society. Princeton, New Jersey: Van Nostrand. 5.2

McCrae, Robert R., and Paul T. Costa. 1989. “Reinterpreting the Myers-Briggs Type Indicator
from the Perspective of the Five-Factor Model of Personality,” Journal of Personality 57:17-40.
DOI: 10.1111/j.1467-6494.1989.tb00759.x 5.3

McCrae, Robert R., and Paul T. Costa. 1996. “Toward a New Generation of Personality Theo-
ries: Theoretical Contexts for the Five-Factor Model.” Pp. 51-87 in The Five-Factor Model of
Personality: Theoretical Perspectives, edited by Jerry S. Wiggins. New York: Guilford Press. 5.1

McCubbin, Chris. 2005. The Matrix Online: Prima Official Game Guide. Roseville, California: Prima
Games. 1.2

McCubbin, Chris, David Ladyman, and Tuesday Frase (eds.). 2005. Star Wars Galaxies: The Total
Experience. Roseville, California: Prima Games. 7.1

McGregor, Georgia Leigh. 2006. “Architecture, Space and Gameplay in World of Warcraft and Battle
for Middle Earth 2. Pp. 69-76 in Proceedings of the 2006 International Conference on Game Research
and Development. Perth, Australia: Murdoch University. 3.3

Medler, Ben. 2008. “Views From Atop the Fence: Neutrality in Games.” Pp. 81-88 in Pro-
ceedings of the 2008 ACM SIGGRAPH Symposium on Video Games. New York: ACM.
DOI: 10.1145/1401843.1401860 9.3

Merrick, Kathryn, Mary Lou Maher. 2006. “Motivated Reinforcement Learning for Non-Player
Characters in Persistent Computer Game Worlds.” In Proceedings of the 2006 ACM SIGCHI
International Conference on Advances in computer Entertainment Technology. New York: ACM.
DOI: 10.1145/1178823.1178828 1.4
100 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
Metzen, Chris. 2002. “Of Blood and Honor.” Pp. 545-613 in Warcraft Archive. New York: Pocket
Books. 9.1

Moore, Robert J., Nicolas Ducheneaut, and Eric Nickell. 2007.“Doing Virtually Nothing: Awareness
and Accountability in Massively Multiplayer Online Worlds,” Computer Supported Cooperative
Work 16:265-305. DOI: 10.1007/s10606-006-9021-4 3.3

Moriarty, Christopher, and Avelino J. Gonzalez. 2009. “Learning Human Behavior from Observa-
tion for Gaming Applications.” Proceedings of the 22nd International Florida Artificial Intelligence
Research Society Conference (FLAIRS-2009), Sanibel Island, FL, May 19-21, 2009. 1.4

Murray, Henry A. 1981. Endeavors in Psychology: Selections from the Personology of Henry A. Murray.
New York: Harper and Row. 5.3

Mylonas, Eric. 2005. Dark Age of Camelot: Epic Edition. Roseville, California: Prima Games. 2.2

Nardi, Bonnie. 2010. My Life as a Night Elf Priest: An Anthropological Account of World of Warcraft.
Ann Arbor: University of Michigan Press. 1

Nardi, Bonnie, and Justin Harris. 2006. “Strangers and Friends: Collaborative Play in World of
Warcraft.” Pp. 149-158 in Proceedings of the 2006 20th Anniversary Conference on Computer Supported
Cooperative Work. New York: ACM. 8.1

Nardi, Bonnie A., Stella Ly, and Justin Harris. 2007.“Learning Conversations in World of Warcraft.”
In Proceedings of the 40th Annual Hawaii International Conference on System Sciences. Washington,
DC: IEEE Computer Society. DOI: 10.1109/HICSS.2007.321 8.1

Nolen, Chelsea Winter. 2007. “Virtual Pluralities: Cultural Pluralism in a Massively Multiplayer
Online Role-Playing Game.” Unpublished senior thesis. St. Louis, Missouri: Washington Uni-
versity. 9.2

Osgood, Charles E., George J. Suci, and Percy H. Tannenbaum. 1957. The Measurement of Meaning.
Urbana: University of Illinois Press. 5.4

Osgood, Charles E., William H. May, and Murray S. Miron. 1975. Cross-Cultural Universals of
Affective Meaning. Urbana: University of Illinois Press.

Pangburn, Weaver. 1922.“The Worker’s Leisure and His Individuality,” American Journal of Sociology,
27: 433-441. DOI: 10.1086/213375 6

Pantel, Lothar, and Lars C. Wolf. 2002.“On the Impact of Delay on Real-Time Multiplayer Games.”
Pp. 23-29 in Proceedings of the 12th International Workshop on Network and Operating Systems
Support for Digital Audio and Video. New York: ACM. DOI: 10.1145/507670.507674 3.1
 9.3. RESEARCH OPPORTUNITIES 101
Papargyris, Anthony, and Angeliki Poulymenakou. 2004. “Learning to Fly in Persistent Digital
Worlds: The Case of Massively Multiplayer Online Role Playing Games,” SIGGROUP Bulletin
25:41-49. DOI: 10.1145/1067699.1067706 8.1

Pearce, Celia. 2010. “The Diasporic Game Community: Trans-Ludic Cultures and Latitudinal
Research Across Multiple Games and Virtual Worlds.” In Online Worlds, edited by William Sims
Bainbridge. Guildford, Surrey, England: Springer. 1.2

Peter, Christian, and Russell Beale (eds.). 2008. Affect and Emotion in Human-Computer Interaction.
Berlin: Springer. DOI: 10.1007/978-3-540-85099-1 2.3

Pittman, Daniel, and Chris Gauthier Dickey. 2007. “A Measurement Study of Virtual Populations in
Massively Multiplayer Online Games.” Pp. 25-30 in Proceedings of the 6th ACM SIGCOMM Work-
shop on Network and System Support for Games. New York:ACM. DOI: 10.1145/1326257.1326262
8.3

Picard, Rosalind W. 1997. Affective Computing. Cambridge, Massachusetts: MIT Press. 2.3

Porges, Irwin. 1975. Edgar Rice Burroughs: The Man who Created Tarzan. Provo, Utah: Brigham
Young University Press. 2.5

Prensky, Marc. 2001. Digital Game-based Learning. New York:McGraw-Hill. 1.3

Rankin, Yolanda A., McKenzie McNeal, Marcus W. Shute, and Bruce Gooch. 2008. “User Centered
Game Design: Evaluating Massive Multiplayer Online Role Playing Games for Second Language
Acquisition.” Pp. 43-49 in Proceedings of the 2008 ACM SIGGRAPH Symposium on Video Games.
New York: ACM. DOI: 10.1145/1401843.1401851 9.2

Robinett, Warren. 2003. “Foreword.” Pp. vi-xix in The Video Game Theory Reader, edited by Mark
J.P. Wolf and Bernard Perron. New York: Routledge. 4.2

Rossi, Peter H., and James D. Wright. 1984. “Evaluation Research: An Assessment,” Annual Review
of Sociology 10:331-352. DOI: 10.1146/annurev.so.10.080184.001555 9.3

Rettberg, Scott. 2008. “Corporate Ideology in World of Warcraft.” Pp. 19-38 in Digital Culture,
Play, and Identity: A World of Warcraft Reader, edited by Hilde G. Corneliussen and Jill Walker
Rettberg. Cambridge, Massachusetts: MIT Press. 9.2

Rossi, Peter H., Richard A. Berk, and Kenneth J. Lenihan. 1980. Money, Work, and Crime. New
York: Academic Press. 9.3

Rossi, Peter H., Richard A. Berk, and Kenneth J. Lenihan. 1982. “Saying it Wrong with Figures,”
American Journal of Sociology 88:390-393. DOI: 10.1086/227677 9.3

Rouse, Richard. 2001. Game Design: Theory and Practice. Plano, Texas: Woodware. 1.3
102 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
Ryan, Marie-Laure. 2001. “Beyond Myth and Metaphor: The Case of Narrative in Digital Media.”
Game Studies 1 ( July);
http://www.gamestudies.org/0101/ryan/ 6.1

Rymaszewski, Michael, Wagner James Au, Mark Wallace, Catherine Winters, Cory Ondrejka, and
Benjamin Batstone-Cunningham. 2007. Second Life: The Official Guide. Hoboken, New Jersey:
Wiley. 1.1

Scacchi, Walt. 2004. “Free and Open Source Development Practices in the Game Community,”
IEEE Software 21(1):59-66. DOI: 10.1109/MS.2004.1259221 9.1

Schultze, Ulrike, and Matthew Michael Leahy. 2009. “The Avatar-Self Relationship: Enacting
Presence in Second Life.” In Proceedings of the Thirtieth International Conference on Information
Systems. Atlanta, Georgia: Association for Information Systems. 6.2

Schütz, Alfred. 1971. “On Multiple Realities.” Pp. 207-259 in Collected Papers. The Hague, Nether-
lands: Nijhoff. 5.3

Sears, Andrew, Jonathan Lazar, Ant Ozok, and Gabriele Meiselwitz. 2008.“Human-Centered Com-
puting: Defining a Research Agenda.” International Journal of Human-Computer Interaction 24:2-
16. DOI: 10.1145/1394427.1394430 1.3

Sheldon, Nathan, Eric Girard, Seth Borg, Mark Claypool, and Emmanuel Agu. 2003. “The Effect
of Latency on User Performance in Warcraft III.” Pp. 3-14 in Proceedings of the 2nd Workshop on
Network and System Support for Games. New York: ACM. DOI: 10.1145/963900.963901 3.1

Sherif, Muzafer, O.J. Harvey, B. Jack White, William R. Hood, and Carolyn W. Sherif. 1988.
Intergroup Conflict and Cooperation: The Robbers Cave Experiment. New York: Harper and Row.
9.3

Smith, M. Brewster, Jerome S. Bruner, and Robert W. White. 1956. Opinions and Personality. New
York, Wiley. 5.3

Stanton, Howard, Kurt W. Back, and Eugene Litwak. 1956. “Role-Playing in Survey Research,”
American Journal of Sociology, 62: 172-176. DOI: 10.1086/221958 6

Stark, Rodney, and William Sims Bainbridge. 1985. The Future of Religion. Berkeley: University of
California Press. 2.4

Stark, Rodney, and William Sims Bainbridge. 1987. A Theory of Religion. New York: Toronto/Lang.
2.4

Steinkuehler, Constance. 2008. “Massively Multiplayer Online Games as an Educational Technol-

ogy: An Outline for Research,” Educational Technology 48(1):10-21. 9.2
 9.3. RESEARCH OPPORTUNITIES 103
Steinkuehler, Constance, and Marjee Chmiel. 2006. “Fostering Scientific Habits of Mind in the
Context of Online Play.” Pp. 723-729 in Proceedings of the 7th International Conference on Learning
Sciences. Bloomington, Indiana: International Society of the Learning Sciences. 9.2

Su, Wen-Poh, Binh Pham, and Aster Wardhani. 2007. “Personality and Emotion-Based High-Level
Control of Affective Story Characters,” IEEE Transactions on Visualization and Computer Graphics
13: 281-293. DOI: 10.1109/TVCG.2007.44 5.4

Suznjevic, Mirko, Maja Matijasevic, and Ognjen Dobrijevic. 2008. “Action Specific Massive Multi-
player Online Role Playing Games Traffic Analysis: Case Study of World of Warcraft.” Pp. 106-107
in Proceedings of the 7th ACM SIGCOMM Workshop on Network and System Support for
Games. New York: ACM. DOI: 10.1145/1517494.1517519 3.1

Sweetser, Penelope, and Peta Wyeth. 2005. “GameFlow: A Model for Evaluating Player Enjoyment
in Games.” Computers in Entertainment, 3(3). 5.3

Tarng, Pin-Yun, Kuan-Ta Chen, and Polly Huang. 2008. “An Analysis of WoW Players’ Game
Hours.” In Proceedings of the 7th ACM SIGCOMM Workshop on Network and System Support
for Games. New York: ACM. DOI: 10.1145/1517494.1517504 8.3

Tørnquist, Ragnar. 2001. Prophet Without Honour.

www.anarchy-online.com/anarchy/frontend/files/CONTENT/download/
documents/prophet_without_honour.pdf 9.1

Tychsen, Anders. 2006. “Role Playing Games: Comparative Analysis Across Two Media Platforms.”
Pp. 75-82 in Proceedings of the 3rd Australasian Conference on Interactive Entertainment. Perth,
Australia: Murdoch University. 4.3

Tychsen, Anders, Michael Hitchens, Thea Brolund, and Manolya Kavakli. 2005. “The Game Mas-
ter.” Pp. 215-222 in Proceedings of the Second Australasian Conference on Interactive Entertainment.
Sydney, Australia: Creativity and Cognition Studios Press. 4.3

Vallerand, Robert J., Céline Blanchard, Geneviève A. Mageau, Richard Koestner, Catherine Ratelle,
Maude Léonard, Marylène Gagné, and Josée Marsolais. 2003. “Les Passions de l’Âme: On
Obsessive and Harmonious Passion,” Journal of Personality and Social Psychology, 85:756–767.
DOI: 10.1037/0022-3514.85.4.756 5.3

Von Neumann, John, and Oskar Morgenstern. 1944. Theory of Games and Economic Behavior. Prince-
ton, New Jersey: Princeton University Press. 9

Walther, Bo Kampmann. 2005. “Atomic Actions - Molecular Experience: Theory of Pervasive Gam-
ing.” Computers in Entertainment 3(3). 1.1
104 CHAPTER 9. IMPLICATIONS FOR EXTERNAL SOCIETY
Wang, Chih-Chien, and Yi-Shiu Chu. 2007.“Harmonious Passion and Obsessive Passion in Playing
Online Games,” Social Behavior and Personality, 35:997-1006. DOI: 10.2224/sbp.2007.35.7.997
5.3

Webb, Steven Daniel, and Sieteng Soh. 2007. “Cheating in Networked Computer Games: A Re-
view.” Pp. 105-112 in Proceedings of the 2nd International Conference on Digital Interactive Media
in Entertainment and Arts. New York: ACM. DOI: 10.1145/1306813.1306839 4.2

Whitney-Robinson, Voronica. 2004. Star Wars Galaxies: The Ruins of Dantooine. New York: Ballan-
tine. 9.1

Williams, Dmitri. 2006. “Virtual Cultivation: Online Worlds, Offline Perceptions,” Journal of Com-
munications 56:69-87. DOI: 10.1111/j.1460-2466.2006.00004.x 9.3

Williams, Dmitri, Nicolas Duchenaut, Li Xiong, Yuanyuan Zhang, Nick Yee, and Eric Nickell.
2006. “From Tree House to Barracks: The Social Life of Guilds in World of Warcraft,” Games
and Culture 1: 338-361. DOI: 10.1177/1555412006292616 8.3

Winnicott, D.W. 1971. Playing and Reality. New York: Basic Books. 9

Yang, Kuo-shu, and Michael Harris Bond. 1990. “Exploring Implicit Personality Theories with
Indigenous or Imported Constructs:The Chinese Case,” Journal of Personality and Social Psychology
58: 1087-1095. DOI: 10.1037/0022-3514.58.6.1087 5.1

Yee, Nick. 2006. “The Demographics, Motivations and Derived Experiences of Users of Massively-
Multiuser Online Graphical Environments.” PRESENCE: Teleoperators and Virtual Environments
15: 309-329. DOI: 10.1162/pres.15.3.309 5.2

Yee, Nick. 2009. “Befriending Ogres and Wood-Elves: Relationship Formation and The Social
Architecture of Norrath.” Game Studies 9(1);
http://gamestudies.org/0901/articles/yee 5.2

Zeisel, Hans. 1982a. “Disagreement over the Evaluation of a Controlled Experiment,” American
Journal of Sociology 88:378-389. DOI: 10.1086/227676 9.3

Zeisel, Hans. 1982b. “Hans Zeisel Concludes the Debate,” American Journal of Sociology 88:394-396.
DOI: 10.1086/227678 9.3
 105

Author’s Biography

WILLIAM SIMS BAINBRIDGE

William Sims Bainbridge is the author of 18 books and about 200 articles in the areas of sociology
of technology, social movements, and research methodologies. For his recent book, God from the
Machine, he programmed a neural network multi-agent system to simulate religious cognition and
conversion in a large community. His book about World of Warcraft, The Warcraft Civilization, has
recently been published by MIT Press. After organizing the first large scientific meeting inside World
of Warcraft in May 2008, he edited a book growing out of the proceedings, Online Worlds, published
by Springer. He spends well over a thousand hours each year doing observational research inside
virtual worlds, and he is currently writing a pair of books on how gameworlds explore both the future
and the past of humanity. For the last seventeen years, he has served as a program officer managing
review of grant proposals in the social science and computer science directorates of the National
Science Foundation. He has extensive experience editing publications on the societal implications
of nanotechnology, converging technologies, and human-computer interaction.