Social diversity promotes cooperation in spatial multigames

Social diversity is omnipresent in the modern world. Here we introduce this diversity into spatial multigames and study its impact on the evolution of cooperation. Multigames are characterized by two or more different social dilemmas being contested among players in the population. When a fraction of players plays the prisoner’s dilemma game while the remainder plays the snowdrift game cooperation becomes a difficult proposition. We show that social diversity, determined by the payoff scaling factors from the uniform, exponential or power-law distribution, significantly promotes cooperation. In particular, the stronger the social diversity, the more widespread cooperative behavior becomes. Monte Carlo simulations on the square lattice reveal that a power-law distribution of social diversity is in fact optimal for socially favorable states, thus resonating with findings previously reported for single social dilemmas. We also show that the same promotion mechanism works in time-varying environments, thus further generalizing the important role of social diversity for cooperation in social dilemmas. Copyright c © EPLA, 2017 Cooperative phenomenon exists widely in the real world, ranging from animal to human societies [1]. However, how to understand why selfish players are willing to donate to the collective income at individual cost remains unclear. This confusion is normally investigated using the evolutionary game theory [2–6] as a classical theoretical framework. Furthermore, the prisoner’s dilemma game and the snowdrift game as the exemplifications for addressing the cooperative phenomenon have received substantial attention [7–29]. For example, in a standard prisoner’s dilemma game played by two players, each should simultaneously decide whether to cooperate or to defect. Since a defecting player will acquire the maximum payoff if encountering a cooperative player, the emergence of cooperation faces an enormous challenge. In fact, the entire population resorts to defection within the prisoner’s dilemma game in well-mixed populations [4]. Therefore, much research has been devoted to exploring mechanisms that can bring about the promotion of the cooperative strategy among selfish players. In order to resolve social dilemmas, where selfish players pursuing short-term individual benefits might lead to the tragedy of the commons [30], an impressive amount of research has been carried out in this field over the past years [31–41]. In ref. [36], Nowak has discussed five classic mechanisms for the promotion of cooperation: kin selection, direct reciprocity, indirect reciprocity, network reciprocity, as well as group selection. In ref. [37], Nowak and May have first introduced the prisoner’s dilemma game on the square lattice. They have found that cooperators form compact clusters and so escape the exploitation of the defecting players, thus facilitating cooperative behavior. Following this significant breakthrough, network reciprocity attracted considerable attention [42–47]. In ref. [42], Santos and Pacheco have discovered that scale-free networks promote the emergence of cooperation both in the snowdrift game and the prisoner’s dilemma game far beyond the boundaries imposed by regular lattices. Moreover, coevolutionary rules where strategies of players and other properties simultaneously evolve have been investigated [48–54], further enriching the mechanisms for raising the degree of cooperation. In addition, several other approaches have been considered that may favorably influence the evolution of cooperation, like payoff noise [55–59], strategic complexity [60–67], inhomogeneous activity of players [68,69], populations of mobile individuals [70], as well as multilayer networks [71–76].