Transcript slides - Courses

Computer-Mediated
Communication
Collective Action and CMC: Game
Theory Approaches and Applications
Coye Cheshire & Andrew Fiore
//
March 21, 2012
Why Game Theory for mediated
communication?
4/9/2015
Computer-Mediated Communication
1
Game Theory (definition)
“Game theory is the systematic study of
interdependent rational choice. It may be
used to explain, to predict, and to evaluate
human behavior in contexts where the
outcome of action depends on what several
agents choose to do and where their choices
depend on what others choose to do.”
http://plato.stanford.edu/entries/game-ethics/
4/9/2015
Computer-Mediated Communication
2
Game Theory and Core Concepts
 Analytical vs. Behavioral
Game Theory
 Cooperative and NonCooperative Games
 Zero and Non-Zero Sum
Games
 One-Shot vs. Repeated
(example for cooperative game)
 Equilibria
 (i.e., Nash Equilibrium)
4/9/2015
Computer-Mediated Communication
3
Types of Social Dilemmas
Different social dilemma games
make different use of the payouts:
T>R>P>S Prisoner’s Dilemma
But also…
T>R>S>P Chicken
T>P>R>S Deadlock
R>T>P>S Stag Hunt
Reward
Temptation
Sucker
Punishment
4/9/2015
Coop.
Coop.
A
Defect
B
Defect
3 (R)
5 (T)
0 (S)
3 (R)
0 (S)
1 (P)
5 (T)
1 (P)
Computer-Mediated Communication
4
Example: Chicken Game
T>R>S>P Chicken
Coop = Swerve
Defect = Do Not Swerve
Reward
Temptation
Sucker
Punishment
Swerve
A
No Swerve
4/9/2015
Swerve
B
No Swerve
3 (R)
5 (T)
1 (S)
3 (R)
1 (S)
-1 (P)
5 (T)
-1 (P)
Computer-Mediated Communication
5
 2-person repeated PD  N-person PD
Public Good
4/9/2015
Computer-Mediated Communication
6
The N-person PD
“No one wants to pay taxes because the
benefits are so diffuse and the costs are so
direct. But everyone may be better off if
each person has to pay so that each can
share the benefits”
cf. Schelling 1973; Axelrod 1984
http://www.flickr.com/photos/cogdog/4353774/
4/9/2015
Computer-Mediated Communication
7
In-Class Tournaments
4/9/2015
Computer-Mediated Communication
8
Small Group Discussion #1
 Get into a small group (2-3 students, three groups total)

Imagine you are going to play in a series of separate tournaments, each
against a single strategy for exactly 10 rounds. Your goal is to make the most
points.
 What strategy would you use if you were going to play against AllCooperate?
 What strategy would you use if you play against All-Defect?
 What strategy would you use if you play against Random (i.e.,
cooperation and defection are always randomly chosen)?
4/9/2015
Computer-Mediated Communication
9
The Evolution of Cooperation
 Axelrod’s famous (1984)
tournament allowed individuals
to submit any strategy.
 All strategies played each
other in the tournament.
 The winner was one of the
shortest submissions, about 4
lines of code.
4/9/2015
Computer-Mediated Communication
10
The Simple Effectiveness of the Tit-for-Tat Strategy
 Tit-for-Tat: begin
with ‘cooperate’
and then do
whatever the
opponent did on
the last turn.
4/9/2015
Computer-Mediated Communication
11
Lessons from Tit-for-Tat
 Be nice
 It starts by cooperating. Most topscoring strategies do this.
 Be forgiving
 It quickly and happily returns to
cooperation without holding a grudge.
 Be able to retaliate
 It never allows defection to go
unpunished.
 Be clear
 It is predictable and easy to understand.
It pays to be predictable in non-zero sum
games.
4/9/2015
Computer-Mediated Communication
12
Considering the “Shadow of the
Future”
4/9/2015
Computer-Mediated Communication
13
How is tit-for-tat different in the two
types of situations?
 2-person repeated PD
 N-person PD
Public Good
4/9/2015
Computer-Mediated Communication
14
Some common complaints…
 “A theoretical tool
cannot explain real life,
right?”
 “Hey, isnt this rational
choice?”
4/9/2015
(Picture courtesy vismod.media.mit.edu)
Computer-Mediated Communication
15
The Value Fallacy: Individuals and Collectives
4/9/2015
Computer-Mediated Communication
16