Collective behavior and kin selection in evolutionary IPD

Jiawei Li, Graham Kendall

Research output: Journal PublicationArticlepeer-review

4 Citations (Scopus)

Abstract

Some strategies can be evolutionarily stronger than others although no evolutionarily stable strategy exists in iterated prisoner's dilemma if the long-term payoff for each player is not insignificant. Li and Kendall (2009) introduced a so-called collective strategy for evolutionary iterated prisoner's dilemma which plays a sequence of predefined moves and then identifies the opponent according to the response. It only cooperates with kin members and defects against any other strategies.Agroup of collective strategies is especially strong in evolution. In this paper, we study a mixed strategy that assigns probabilities to the collective strategy and the strategy that always defects. Apopulation of mixed strategies has the advantage of expelling fake kin members so that other strategies do not have the chance to indirectly invade. Simulations show that it is evolutionarily strong in maintaining a homogeneous population. Kin selection favors collective behavior among group members which is not necessarily cooperation.We find that defection can also be a kin altruism and there is qualitative benefit as well as quantitative benefit from the altruistic behaviors.

Original languageEnglish
Pages (from-to)509-525
Number of pages17
JournalJournal of Multiple-Valued Logic and Soft Computing
Volume16
Issue number6
Publication statusPublished - 2010

Keywords

  • Collective behavior
  • Evolutionarily stable strategy
  • Kin selection
  • Prisoner's dilemma

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Logic

Fingerprint

Dive into the research topics of 'Collective behavior and kin selection in evolutionary IPD'. Together they form a unique fingerprint.

Cite this