Evolution of Extortion in the Social-Influenced Prisoner'S Dilemma

Zhipeng Wang,Miao Li,Dan Wang,Qinghe Chen

DOI: https://doi.org/10.1142/s0217984916500299

2016-01-01

Modern Physics Letters B

Abstract:The introduction of extortion strategy has attracted much attention since it dominates any evolutionary opponent in iterated prisoner’s dilemma games. Despite several studies argue that extortion is difficult to survive under strategy imitation and birth–death updating rules in well-mixed populations, it has recently been proven that a myopic best response rule facilitate the evolution of cooperation and extortion. However, such updating rules require a strong assumption of complete knowledge of all players, which is unlikely to hold in social networks in reality. To solve this problem, we introduce the concept of social influence into the model to limit players’ knowledge within their neighborhood. It turns out that this myopia initiated by social influence prevents players from observing superior strategies and therefore enables cooperators and extortioners to be evolutionarily stable. We also suggest that heterogeneous networks contribute to the evolution of cooperation and extortion under such social influence.