game theory by james crissey luis mendez james reid
TRANSCRIPT
Game Theory
by James Crissey
Luis Mendez
James Reid
Complex Adaptive Systems and the Evolution of Reciprocation
by Karl Sigmund
University of Vienna
Human Societies
cooperation is ubiquitous few individual inequalities in
reproductive potential social rules level reproductive
opportunities and prevent establishment of global controller
Human cooperation
not based on relatedness Based on reciprocation based on economic exchanges human cohesiveness is explained
based on economic exchanges
Game theory
studies the economics of interacting egoists - non-cooperative game theory
players cooperate, not by the command of a controller or by deference to the group
cooperation based on rules
Game theory cont . . .
Trivers– suggested reciprocation as mutual
assistance– introduced the prisoner’s dilemma
Expansion of Trivers Findings
1981 Axelrod and Hamilton applied the evolutionary game theory to players engaged in many rounds of the prisoners dilemma game against randomly chosen co-players.
Axelrod adapted the genetic algorithms of John Holland to simulate the effects of evolutionary trial and error.
Expansion of Trivers finding cont. First application of genetic algorithms to
a genuine evolutionary problem rather than a technical optimization problem
Indirect reciprocity
Alexander (1987) the biological basis for human moral systems
an act of assistance may be returned not to the donor but a third party
Direct Reciprocation
Prisoner’s dilemma is a two-player game where both players have the same two strategies and the same payoff
Direct Reciprocation
Prisoner’s dilemma– two-player game in which both player’s
have the same two strategies as well as the same two payoffs
– can to cooperate or defect– rewarded based on what they choose
Direct Reciprocation cont . . .
PD game player’s choose simultaneously whether they want to cooperate or defect
assumes that if player starts off by cooperating and then does what ever the other person does
Direct Reciprocation cont . . .
Generous Tit for Tat (GTFT)– person retaliates sometimes after players
defects– but always cooperates after they cooperate – Pavlov rule, suggests that people will
cooperate if they use the same rule in the first round- assumes that players will repeat a move if it leads to high payoff
Indirect Reciprocation
the donor does not obtain a return from the recipient
donor receives return from third party cost of an altruistic act is offset by a
raised ‘score’ or status, which increases chance to subsequently become the recipient of an altruistic act
Indirect Reciprocation cont . . .
Indirect reciprocity involves reputation and status, and results in everyone in the group continually being assessed and re-assessed (Alexander, 1987)
Nowak and Sigmund (1998)– model of indirect reciprocation
Indirect Reciprocation cont...
in donor who provides there is a cost however, the donor’s score increases if player refuses to play, score decreases if the frequency of less discriminating
reaches a certain threshold, then defectors can take over, with the result that cooperation disappears in the population
Indirect discrimination cont . . .
random drift can subvert populations of discriminate altruist by indiscriminate altruist
once there frequency is too large, defectors can invade; but as soon as defectors reduce the proportion of indiscriminate altruist, the discriminate altruists fight back and eliminate the defectors
Indirect Reciprocation
this leads to a cooperative population which is proof against defectors, but not against indiscriminate altruists
chances of two players meeting again is low must know the score of the co-player interaction between two individuals is only
observed by a small subset of the population
Indirect reciprocation cont . . .
only the few on-lookers will update score of donor
For larger groups it is difficult to establish cooperation
for indirect reciprocation to function– compute minimal amount of discriminators– minimal amounts of rounds per generation– maximal size of society
Discussion/Critical Analysis
underestimates kinship and genetic relatedness