Scribd
Upload
15 views16 pages

Lecture1 1

Regulatory Economy

Uploaded by

Munkherdene
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
15 views16 pages

Lecture1 1

Regulatory Economy

Uploaded by

Munkherdene
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

ECN 307E: Game Theory

Lecture 1: Introduction

Istanbul Technical University


Fall 2023

1 / 16
Individual Decision Problems vs. Strategic Interactions

Recall from your Microeconomics class.


Consumer’s utility maximization problem:

max u(x1 , x2 )
x1 ,x2 ≥0

s.t. p1 x1 + p2 x2 ≤ w.

Competitive firm’s profit maximization problem:

max py − c(y).
y≥0

These two problems are examples of individual decision


problems: decision maker’s utility/payoff is completely
determined by his own choice.

2 / 16
Individual Decision Problems vs. Strategic Interactions

In many situations, decision makers interact with each other:


chess
oligopoly competition
auction
political campaigning
...

In these situations, one participant’s utility/payoff not only


depend on his own choice but also on others’.
We call these kinds of situations as strategic interactions.

3 / 16
Individual Decision Problems vs. Strategic Interactions

Individual decision problems are easy to solve.


Calculus, first and second order conditions.
Constrained optimization, Lagrangian.

Strategic interactions are much more difficult to analyze.


Consider the famous game: Rock-Paper-Scissors:
If I predict that you will choose “Rock”, I will choose “Paper”.
But if you predict my prediction, you will choose “Scissors”.
If I predict that you predict my prediction, I will choose “Rock”.
If you predict that I predict that you predict my prediction · · · ·
Not a simple optimization problem.

4 / 16
Historical Background

John von Neumann Oskar Morgenstern

In 1944, John von Neumann and Oskar Morgenstern published


the book “Theory of Games and Economic Behavior” which
established the game theory as a field.

5 / 16
Game Theory

Theory of Games and Economic Behavior on strategic


interactions:

“[I]t is unlikely that a mere repetition of the tricks which served us so


well in physics will do for the social phenomena too.”

“[T]he process of mathematization is not at all obvious. ... We shall


find it necessary to draw upon techniques of mathematics which have
not been used heretofore in mathematical economics, and ... further
study may result in the future in the creation of new mathematical
disciplines.”

−→ The establishment of Game theory as a field.

6 / 16
Game Theory

Game theory provides a language to describe a strategic


interaction.
We will learn how to write down a formal model.
Game theory also offers a prediction about what rational
players will do in a given strategic interaction.
We will also learn how to solve the model.

7 / 16
Setting The Stage

Before kicking off · · · ·


Throughout the semester, we will be regularly examining
situations where there is an underlying uncertainty.

Understanding the choice under uncertainty is crucial.

Will briefly review expected utility theory to refresh your


knowledge.

8 / 16
Review: Expected Utility Theory

We always assume that agents are rational.

Rational agents aim to maximize their utilities/payoffs when


making decisions.

More straightforward when an agent faces certain outcomes.

What if there is uncertainty?

Expected utility theory provides an answer.

9 / 16
Review: Expected Utility Theory

First define the main apparatus for modeling uncertainty.

Let X = {x1 , x2 , . . . , xn } be set of possible outcomes.

The notion of simple lottery to represent risk (uncertainty).

Definition
A simple lottery L over X = {x1 , x2 , . . . , xn } is a probability
distribution
N
X
L = (p1 , . . . , pN ), with pn ≥ 0, and pn = 1,
n=1

where pn is the probability of outcome xn occurring.

10 / 16
Review: Expected Utility Theory

Consider literally a lottery that pays $0 with probability 99%


and $100 with probability 1%:
X = {0, 100}
prob. distribution p with p(0) = 0.99, p(100) = 0.01.

Tomorrow’s weather can be either sunny, cloudy or rainy with


probabilities 80%, 15% and 5%; respectively:
X = {s, c, r}
prob, distribution p with p(s) = 0.8, p(c) = 0.15, p(r) = 0.05.

11 / 16
Review: Expected Utility Theory

Let L(X) be the set of all lotteries defined over X.


How would you define an agent’s utility over the lotteries?

Definition
Let u(x) be an agent’s payoff function over X = {x1 , . . . , xn }.
Then the agent’s expected payoff from a lottery L ∈ L is
n
X
U (L) ≡ E[u(x)] = p1 u(x1 ) + . . . + pn u(xn ) = pk u(xk ).
k=1

U (·) is called the expected utility function.


u(·) is called as vonNeumann–Morgenstern utility function.

12 / 16
Review: Expected Utility Theory

Expected payoff is the expectation of the payoffs from the


outcomes with respect to the lottery p.
In other words, it is the weighted average of the payoffs from
the outcomes.

13 / 16
Review: Expected Utility Theory

Consider the previous weather example.


Suppose that an agent has two options: carry an umbrella
and not carry an umbrella.
His/her payoffs (von Neumann–Morgenstern utility function):

−1 if it does not rain and s/he carries an umbrella;


10 if it rains and s/he carries an umbrella;
0 if it does not rain and s/he does not carry an umbrella;
10 if it rains but s/he does not carry an umbrella.

What should s/he do?

14 / 16
Review: Expected Utility Theory

His/her expected payoff from carrying an umbrella is

0.8 × (−1) + 0.15 × (−1) + 0.05 × 10 = −0.45,

His/her expected payoff from not carrying an umbrella is

0.8 × 0 + 0.15 × 0 + 0.05 × (−10) = −0.5.

=⇒ Carrying an umbrella is optimal.

15 / 16
Technical Appendix
Formal Statement of Expected Utility Theorem

Theorem (Expected Utility Theorem.)


When preferences over L(X ) satisfy continuity and
independence axioms then these preferences can be represented
in expected utility form.

16 / 16

You might also like