Introduction to Game Theory

Sanjay Singh

1
Department of Information and Communication Technology
Manipal Institute of Technology, MAHE,Manipal-576104, INDIA
sanjay.singh@manipal.edu
2
Centre for Artificial and Machine Intelligence (CAMI)
MAHE, Manipal-576104, INDIA

March 2, 2021

Sanjay Singh Introduction to Game Theory

Game theory (GT) and mechanism design deal with

interaction among strategic agents
Game theory is concerned with analysis of games
Mechanism design involves designing games with
desirable outcomes
The term game used in phrase game theory corresponds
to an interaction involving decision makers or players who
are rational and intelligent
Informally, rationality of a player implies that the player
chooses his strategies so as to maximize a well defined
individualistic payoff
Intelligence means that players are capable enough to
compute their best strategies

Sanjay Singh Introduction to Game Theory

 GT is a tool for logical and mathematical analysis that
models conflict and cooperation between decision makers
It provides a principled way of predicting the result of the
interaction among the players using equilibrium analysis
Traditional game such as chess is straightforward in nature
Games that GT deals with are more generic and can be
viewed as abstraction and extension of the traditional
games
Abstraction and extension are powerful enough to include
all complexities and characteristics of social interactions

Sanjay Singh Introduction to Game Theory

Example
Consider two students (call them A and B), belong to MIT,
Manipal, who are close friends. The students derive utility by
spending time together either studying (in MITL) or going to
student plaza (SP). Thus to spend time they have two options
(or strategies):MITL and SP. If both of them are in MITL, each
one gets a payoff of 100. If both go to SP, each gets a payoff of
only 10. If one of them remains in MITL and other goes to SP,
the payoff is 0 for each.

B
A MITL SP
MITL 100,100 0,0
SP 0,0 10,10

Sanjay Singh Introduction to Game Theory

Suppose the two friends have to choose their strategies
simultaneously and independently of each other
Being rational and intelligent, each one would like to select
the best possible strategy
It is clear that both opting for MITL is the best strategy and
both going to SP is also fine though worse than going to
MITL
The worst happens when they choose different options
since each ends up with zero utility
Game theory helps us with a principled way of predicting
the options that would be chosen by both students
In this case, the outcome of both opting for MITL or SP can
be shown to be what called as Nash equilibria
Nash equilibria are strategy profiles in which no player is
better off by unilateral deviating from his equilibrium
strategy

Sanjay Singh Introduction to Game Theory

Figure 1: Legends of game theory and mechanism design

Sanjay Singh Introduction to Game Theory

 Key Notions in Game Theory

Sanjay Singh

1
Department of Information and Communication Technology
Manipal Institute of Technology, Manipal University
Manipal-576104, INDIA
sanjay.singh@manipal.edu
2
Centre for Artificial and Machine Intelligence (CAMI)
Manipal University, Manipal-576104, INDIA

March 2, 2021

Sanjay Singh Key Notions in Game Theory

Module Outline

Coordination game
Key notions in GT:
preferences
utilities or payoffs
rationality
intelligence and
common knowledge

Sanjay Singh Key Notions in Game Theory

Symbols and Notations
Strategic Form Games

hN, (Si )i∈N , (ui )i∈N i or hN, (Si ), (ui )i

Γ
n Number of players in the game
A set of players, {1, 2, , . . . , n}
N
Set of actions or pure strategies of player i
Si
Short form for S1 , . . . , Sn
(Si )i∈N
Short form of (Si )i∈N (when context is clear)
(Si )
Set of all pure strategy profiles=S1 × . . . × Sn
S
s A strategy profile,s = (s1 , . . . , sn ) ∈ S
All players other than i
−i
s−i A profile of strategies of agents except i
Another representation for profile (s1 , . . . , sn )
(si , s−i )
Set of all strategy profiles of all agents except i
S−i

Sanjay Singh Key Notions in Game Theory

Symbols and Notations-contd

Strategic Form Games

ui Utility function of player i; ui : S → R

Short form for u1 , . . . , un
(ui )i∈N
Short form for (ui )i∈N (when context is clear)
(ui )
An equilibrium strategy of player i
s∗i
An equilibrium strategy profile
s∗ =
(s1 , . . . , s∗n )
∗

Best response correspondence of player i, bi : S−i → 2Si

bi

Sanjay Singh Key Notions in Game Theory

Key terms

A game refers to any social situation involving tow or more

individuals
The individuals involved in a game is called the player or
sometime also called agent
Basic assumptions that game theorist make about players:
players are rational
players are intelligent
A player is rational if he makes decisions consistently in
pursuit of his own objectives
A player is intelligent if he knows everything that we know
about the game, and he can make inference about the
situation that we can make

Sanjay Singh Key Notions in Game Theory

Game theory is defined as the study of mathematical

models of interaction between rational, intelligent decision
makers
Decision makers: players or agents
Interaction may involve: conflict or cooperation
Game theory provides general mathematical techniques
for analyzing situations in which two or players make
decision that influence one another’s welfare
A game can be considered as a mathematical model of a
situation where every player strives to obtain his best
possible outcome, knowing well that all other players are
also striving to obtain their best possible outcomes

Sanjay Singh Key Notions in Game Theory

Strategic Form Games (SFG)
Also known as normal form game

Example
Consider the example of the student coordination problem. For
the sake of convenience,lets rename students as 1, 2, and
MITL and SP as A and B. We have two players, namely, student
1 and 2. Each of them can choose any action or strategy from
the set {A, B}. They choose their individual action
simultaneously, independent of each other. Depending upon
the strategies chosen, the two players obtain payoffs as shown
in Table 1.
2
1 A B
A 10,10 0,0
B 0,0 1,1

Sanjay Singh Key Notions in Game Theory

Definition (Strategic Form Game)

A strategic form game Γ is a tuple hN, (Si )i∈N , (ui )i∈N i, where
N = {1, 2, . . . , n} is a set of players
S1 , S2 , . . . , Sn are sets called the strategy sets of players
1, . . . , n; and
ui : S1 × S2 × . . . × Sn → R for i = 1, 2, . . . , n are mappings
called the utility functions or payoff functions

Example
For the example at left, it is clear that
2
1 A B N = {1, 2}; S1 = S2 = {A, B}
A 10,10 0,0 u1 (A, A) = 10; u1 (A, B) = 0; u1 (B, A) =
B 0,0 1,1 0; u1 (B, B) = 1
u2 (A, A) = 10; u2 (A, B) = 0; u2 (B, A) =
0; u2 (B, B) = 1

Sanjay Singh Key Notions in Game Theory

Strategies are also called as the actions or pure strategies
Strategy profile, S = S1 × S2 × . . . × Sn
Every profile of strategies corresponds to an outcome in
the game
Term strategy profile and outcome are used synonymously
The terms players, individuals, persons, decision makers,
and agents are also used synonymously

Sanjay Singh Key Notions in Game Theory

A SFG is a simultaneous move game that captures each

agent’s decision problem of choosing a strategy that will
counter the strategies adopted by the other agent
A player can be thought of as simultaneously choosing
their strategies from the set S1 , S2 , . . . , Sn
We can view the play of a SFG as follows: each player
simultaneously selects a strategy and informs this to a
neutral observer who then computes the outcome and the
utilities

Sanjay Singh Key Notions in Game Theory

Preference

Student coordination game has four preferences: (A,A),

(A,B), (B,A), and (B,B), which are also four strategy profiles
Each student has certain preferences over these outcomes
Here, student prefers (A,A) over (B,B), prefers (B,B) over
(A,B) or (B,A), and no preference between (A,B) and (B,A)
Preferences that a player has over outcomes can be
formalized as a preference relation over the set of
outcomes S
Preference relation of each player will be reflexive,
transitive, and complete (i.e.,every pair of outcome is
covered by the relation)
In general situation, the preference relation of different
players will be different

Sanjay Singh Key Notions in Game Theory

Utilities

Utility function or payoff function of a player is a real valued

function defined on the set of outcomes or strategy profiles
Utility function of each player maps strategy profiles into
real numbers to capture preferences
Utility of a player in an outcome depends not only on his
own strategy but also on the strategies of the rest of the
players
Whether it is possible to map preference profiles to real
numbers without losing any preference information?

Sanjay Singh Key Notions in Game Theory

Rationality
One of the key assumption in game theory is that the
players are rational
An agent is said to be rational if the agent always makes
decisions in pursuit of her own objectives
Particularly, it is assumed that each agent’s objective is to
maximize the expected value of her own payoff measured
in some utility scale
Rationality implies selfishness of the agent if her utility
function captures her self-interest
Self-interest does not mean that each player wants to harm
the other player
It also does not mean that the players only care about
themselves
Self-interest means that each player has certain individual
preferences over the outcomes and the player continuously
seeks to obtain these preferred outcomes
Sanjay Singh Key Notions in Game Theory

Maximizing expected utility is not necessarily the same as

maximizing expected monetary returns
In general, utility and money are non-linearly related
For example, a certain amount of money may provide
different utilities to different players depending upon how
endowed or desperate they are
When there are two or more players, it would be the case
that the solution to each player’s decision problem
depends on the others’ individual problems and vice-versa
When such rational agents interact, their decision
problems have to be analyzed together, like a system of
simultaneous equation
Game theory provides mathematical framework to deal
with such analysis

Sanjay Singh Key Notions in Game Theory

Intelligence

Intelligence means that each player in the game knows

everything about the game that a game theorist knows,
and the player is competent enough to make any
inferences about the game that a game theorist can make
An intelligent player is strategic, i.e., would fully take into
account his knowledge or expected behavior of other
agents in determining what his optimal response should be
Such a strategy is called best response strategy
Each player is assumed to have enough resources to carry
out the required computations involved in determining a
best response strategy

Sanjay Singh Key Notions in Game Theory

The assumption that all individuals are rational and

intelligent may not be satisfied in a typical real-world
situation
Any theory that is not consistent with the assumptions of
rationality and intelligence loses credibility on the following
account
If a theory predicts that some individuals will be
systematically fooled into making mistakes, then this
theory will lose credibility when individuals learn to better
understand the situations
Theory based on rationality and intelligence assumption
will be sustainable

Sanjay Singh Key Notions in Game Theory

Common Knowledge

Common knowledge is an important implication of

intelligence
Aumann defines common knowledge as follows:
A fact is common knowledge among the players if every
player knows it, every players knows that every player
knows it, and so on
If it happens that a fact is known to all the players, without
the requirement of all players knowing that all players know
it etc, then such a fact is called mutual knowledge
In game theory, analysis often requires that the assumption
of common knowledge to be true; however, sometimes, the
assumption of mutual knowledge suffice for the analysis
A player’s private information is any information that the
player has that is neither common knowledge nor mutual
knowledge among any of the players
Sanjay Singh Key Notions in Game Theory

Classification of Games

Non-cooperative Games and Cooperative Games

Static Games and Dynamic Games
Different Representational Forms
Games with Perfect Information and Games with Imperfect
Information
Complete Information and Incomplete Information Games

Sanjay Singh Key Notions in Game Theory

Summary

Preferences of a player specify qualitatively the player’s

ranking of different outcomes of the game
Utilities are real valued payoffs that players receive when
they play different actions
Rationality means that players always choose their
actions so as to maximize their expected utilities
Depending on how the utility function is defined, rationality
could mean self-interest, altruism, indifference etc
Intelligence means that the player are as knowledgeable
as game theorist and have enough computational power to
compute their best response action
Common knowledge means that all players know the
entire structure of the game, all players know that all
players know the game

Sanjay Singh Key Notions in Game Theory