Rigorous or tolerant: The effect of different reputation attitudes in complex networks

In the collaborative network environment, the ubiquitous cooperation phenomenon among autonomous and selfish individuals has drawn extensive attentions during the last years. The evolution of spatial public goods game has provided an important theoretical framework to understand the emergence of cooperation in this field. Along this line, reputation-based mechanisms have been verified to promote in public goods games by taking into consideration the historical behaviors of players. In most existing studies, individuals are assumed to participate in the favorable games (with good reputation partners) to get more payoffs, and they try to avoid interacting with individuals with bad reputation. However, the effect of individuals' reputation tolerance on the evolution of cooperation in a complex network is still ambiguous. In this paper, we focus on the evolution of cooperation in public goods games on complex networks where individuals may have various reputation tolerances. Specifically, an individual decides whether or not to participate in a public goods game by considering the reputation of its neighbors. The results show that the reputation tolerance based scheme does help to prevent defectors' free-riding behavior, and enhances the formation of cooperative clusters. Moreover, Fermi function performs worse than Best Imitation and Roulette in regular networks and random networks, but better in scale-free networks.