Spatial game with multiple interaction patterns in constrained interaction environment: A computational method based on opponent’s ability

In previous studies of spatial game, it is usually assumed that all individuals in a population play the game in the same interaction pattern. However, in real life, a variety of interaction patterns often emerge spontaneously, especially in a constrained interaction environment. Furthermore, some scenarios reflect the fact that individuals are generally more inclined to choose the capable partners for interactions. For these reasons, we propose a computational method on evolutionary game with multiple interaction patterns based on opponent’s ability and investigate the cooperative dynamics of spatial prisoner’s dilemma. For simplicity, in this work, we just consider the case of two interaction patterns, which are conditional interaction and stochastic interaction. For the individual who adopt the conditional interaction pattern, if the opponent is strong in ability, it will be perfectly willing to interact with opponent, while if the opponent is weak in ability, it will not do that. For the individual who adopt the stochastic interaction pattern, the willingness of interaction with other party varies with the opponent’s abilities, the more powerful the opponent, the stronger the willingness of interaction. Notably, for paired two individuals, regardless of the combined configuration of interaction patterns, the final interaction intensity between them depends on the result of information fusion of their unilateral interaction willingness. Experimental results show that the mechanism with multiple interaction patterns can effectively propagate cooperative behavior in networked population, and the moderate estimation level can lead to the flourish of cooperation. Besides, we find that the more the straightforward individuals (who adopt the conditional interaction pattern) in the population, the more prosperous the propagation of cooperative behavior, while the more the euphemistic individuals (who adopt the stochastic interaction pattern), the easier the emergence of selfish behavior. Considering such a real environment with multiple interaction patterns can provide a brand new perspective for further study on the dynamics of evolutionary game in structured population.