Você está na página 1de 21

Game Theory and its

Applications

SARANI SAHABHATTACHARYA, HSS


ARNAB BHATTACHARYA, CSE

07 JAN, 2009
Prisoner’s Dilemma
2

Two suspects arrested for a crime


Prisoners decide whether to confess or not to confess
If both confess, both sentenced to 3 months of jail
If both do not confess, then both will be sentenced to
1 month of jail
If one confesses and the other does not, then the
confessor gets freed (0 months of jail) and the non-
confessor sentenced to 9 months of jail
What should each prisoner do?

Game Theory Jan 07, 2009


Battle of Sexes
3

A couple deciding how to spend the evening


Wife would like to go for a movie
Husband would like to go for a cricket match
Both however want to spend the time together
Scope for strategic interaction

Game Theory Jan 07, 2009


Games
4

Normal Form representation – Payoff Matrix

Prisoner 2

Confess Not Confess


Confess -3,-3 0,-9
Prisoner 1
Not Confess -9,0 -1,-1

Husband
Movie Cricket
Wife
Movie 2,1 0,0
Cricket 0,0 1,2

Game Theory Jan 07, 2009


Nash equilibrium
5

Each player’s predicted strategy is the best response


to the predicted strategies of other players
No incentive to deviate unilaterally
Strategically stable or self-enforcing

Prisoner 2

Confess Not Confess


Confess -3,-3 0,-9
Prisoner 1
Not Confess -9,0 -1,-1

Game Theory Jan 07, 2009


Mixed strategies
6

A probability distribution over the pure strategies of


the game
Rock-paper-scissors game
 Each player simultaneously forms his or her hand into the
shape of either a rock, a piece of paper, or a pair of scissors
 Rule: rock beats (breaks) scissors, scissors beats (cuts) paper,
and paper beats (covers) rock
No pure strategy Nash equilibrium
One mixed strategy Nash equilibrium – each player
plays rock, paper and scissors each with 1/3
probability
Game Theory Jan 07, 2009
Nash’s Theorem
7

Existence
 Any finite game will have at least one Nash equilibrium
possibly involving mixed strategies
Finding a Nash equilibrium is not easy
 Not efficient from an algorithmic point of view

Game Theory Jan 07, 2009


Dynamic games
8

Sequential moves
 One player moves
 Second player observes and then moves
Examples
 Industrial Organization – a new entering firm in the market
versus an incumbent firm; a leader-follower game in quantity
competition
 Sequential bargaining game - two players bargain over the
division of a pie of size 1 ; the players alternate in making offers
 Game Tree

Game Theory Jan 07, 2009


Game tree example: Bargaining
Period 2:
B offers x2.
(x1,1-x1) (x3,1-x3)
A responds.

1 Y 1 1 Y

x1 x3
N
B B (0,0)
N
A B x2 A A
N

Y
0 0 0
Period 1: Period 3:
A offers x1. (x2,1-x2) A offers x3.
B responds. B responds.
Economic applications of game theory

The study of oligopolies (industries containing only


a few firms)
The study of cartels, e.g., OPEC
The study of externalities, e.g., using a common
resource such as a fishery
The study of military strategies
The study of international negotiations
Bargaining
Auctions
11

Games of incomplete information


First Price Sealed Bid Auction
 Buyers simultaneously submit their bids
 Buyers’ valuations of the good unknown to each other
 Highest Bidder wins and gets the good at the amount he bid
 Nash Equilibrium: Each person would bid less than what the good
is worth to you
Second Price Sealed Bid Auction
 Same rules
 Exception – Winner pays the second highest bid and gets the good
 Nash equilibrium: Each person exactly bids the good’s valuation
Game Theory Jan 07, 2009
Second-price auction
12

Suppose you value an item at 100


You should bid 100 for the item
If you bid 90
 Someone bids more than 100: you lose anyway
 Someone bids less than 90: you win anyway and pay second-price
 Someone bids 95: you lose; you could have won by paying 95
If you bid 110
 Someone bids more than 11o: you lose anyway
 Someone bids less than 100: you win anyway and pay second-price
 Someone bids 105: you win; but you pay 105, i.e., 5 more than what
you value

Game Theory Jan 07, 2009


Mechanism design
13

How to set up a game to achieve a certain outcome?


 Structure of the game
 Payoffs
 Players may have private information
Example
 To design an efficient trade, i.e., an item is sold only when buyer
values it as least as seller
 Second-price (or second-bid) auction
Arrow’s impossibility theorem
 No social choice mechanism is desirable
Akin to algorithms in computer science

Game Theory Jan 07, 2009


Inefficiency of Nash equilibrium
14

Can we quantify the inefficiency?


Does restriction of player behaviors help?
Distributed systems
 Does centralized servers help much?
Price of anarchy
 Ratio of payoff of optimal outcome to that of worst possible
Nash equilibrium
In the Prisoner’s Dilemma example, it is 3

Game Theory Jan 07, 2009


Network example
15

C(x) = 1

C(x) = x
Simple network from s to t with two links
 Delay (or cost) of transmission is C(x)
Total amount of data to be transmitted is 1
Optimal: ½ is sent through lower link
 Total cost = 3/4
Game theory solution (selfish routing)
 Each bit will be transmitted using the lower link
 Not optimal: total cost = 1
Price of anarchy is, therefore, 4/3

Game Theory Jan 07, 2009


Do high-speed links always help?
16

C(x) = x C(x) = 1 C(x) = x C(x) = 1


C(x) = 0
C(x) = 1 C(x) = x C(x) = 1 C(x) = x

½ of the data will take route s-u-t, and ½ s-v-t


Total delay is 3/2
Add another zero-delay link from u to v
All data will now switch to s-u-v-t route
Total delay now becomes 2
Adding the link actually makes situation worse

Game Theory Jan 07, 2009


Other computer science applications
17

Internet
Routing
Job scheduling
Competition in client-server systems
Peer-to-peer systems
Cryptology
Network security
Sensor networks
Game programming

Game Theory Jan 07, 2009


Bidding up to 50
18
Two-person game
Start with a number from 1-4
You can add 1-4 to your opponent’s number and bid
that
The first person to bid 50 (or more) wins
Example
 3, 5, 8, 12, 15, 19, 22, 25, 27, 30, 33, 34, 38, 40, 41, 43, 46, 50
Game theory tells us that person 2 always has a winning
strategy
 Bid 5, 10, 15, …, 50
Easy to train a computer to win

Game Theory Jan 07, 2009


Game programming
19

Counting game does not depend on opponent’s choice


Tic-tac-toe, chess, etc. depend on opponent’s moves
You want a move that has the best chance of winning
However, chances of winning depend on opponent’s
subsequent moves
You choose a move where the worst-case winning
chance (opponent’s best play) is the best: “max-min”
Minmax principle says that this strategy is equal to
opponent’s min-max strategy
 The worse your opponent’s best move is, the better is your move

Game Theory Jan 07, 2009


Chess programming
20

How to find the max-min move?


Evaluate all possible scenarios
For chess, number of such possibilities is enormous
 Beyond the reach of computers
How to even systematically track all such moves?
 Game tree
How to evaluate a move?
 Are two pawns better than a knight?
Heuristics
 Approximate but reasonable answers
 Too much deep analysis may lead to defeat

Game Theory Jan 07, 2009


Conclusions
21

Mimics most real-life situations well


Solving may not be efficient
Applications are in almost all fields
Big assumption: players being rational
 Can you think of “unrational” game theory?
Thank you!
Discussion

Game Theory Jan 07, 2009

Você também pode gostar