Spillover modes in multiplex games: double-edged effects on cooperation and their coevolution

Abstract

In recent years, there has been growing interest in studying games on multiplex networks that account for interactions across linked social contexts. However, little is known about how potential cross-context interference, or spillover, of individual behavioural strategy impact overall cooperation. We consider three plausible spillover modes, quantifying and comparing their effects on the evolution of cooperation. In our model, social interactions take place on two network layers: repeated interactions with close neighbours in a lattice, and one-shot interactions with random individuals. Spillover can occur during the learning process with accidental cross-layer strategy transfer, or during social interactions with errors in implementation. Our analytical results, using extended pair approximation, are in good agreement with extensive simulations. We find double-edged effects of spillover: increasing the intensity of spillover can promote cooperation provided cooperation is favoured in one layer, but too much spillover is detrimental. We also discover a bistability phenomenon: spillover hinders or promotes cooperation depending on initial frequencies of cooperation in each layer. Furthermore, comparing strategy combinations emerging in each spillover mode provides good indication of their co-evolutionary dynamics with cooperation. Our results make testable predictions that inspire future research, and sheds light on human cooperation across social domains.

Assortativity provides a narrow margin for enhanced cooperation on multilayer networks

“…While the bulk of research concerning evolutionary games on networks has been devoted to single and isolated networks, such as small-world [20][21][22][23][24][25][26], scale-free [27][28][29][30][31][32][33][34][35][36][37][38][39][40][41], coevolving [42][43][44][45][46][47][48][49][50], hierarchical [51,52] and community [53] networks, the focus has in recent years been shifting towards multilayer and interdependent networks [54][55][56][57][58][59][60][61][62][63][64][65][66][67][68]. This research has revealed interdependent network reciproci...…”

Reciprocity-based cooperative phalanx maintained by overconfident players

“…In this work we will not assume any sophisticated environmental feedback mechanism [28][29][30] or demanding cognitive skill about players [31][32][33][34][35][36], but only explore the plain consequence of perception error collectively made by group members. To reveal the interaction between strategies and individual skills we consider a coevolutionary model [37][38][39][40][41] where players may not only imitate a more successful strategy but also adopt the way to consider individual achievement when making decisions. In particular, besides individuals who are performing unbiased estimation of their payoff values we also assume the initial presence of over-and underconfident players and monitor the coevolutionary process.…”

Diverse strategic identities induce dynamical states in evolutionary games

“…Today, evolutionary games are the workhorse for studying frequency dependent selection in biological systems, and they are used prolifically to investigate cooperation and competition in economic, social, and technological systems [3,4]. In general, evolutionary games allow modelers to make the interactions within a system measurable and interpretable, which pairs particularly well with network science [5][6][7][8][9][10].…”