Impact of multi-valued gossip computational methods on cooperation

Gossip is considered to be an effective mechanism for promoting cooperation, most previous simulation models about gossip usually assume gossip content is expressed in a binary form(good or bad), which ignores the significance of that gossip content is multi-valued. In this paper, we consider the form of gossiping image score to study multi-valued gossip computational methods. Specifically, during the gossip session, two gossipers first gossip to each other their private image scores to everyone, then private image score matrix are updated together in accordance with the average, maximum and minimum of private image scores between two gossipers. Individual-based simulation and analytical results indicated that the three gossip computational methods will lead to the promotion of cooperation compared to the traditional image score model. Among them, when gossip times are greater than a certain threshold, the promotion of cooperation under the minimum method is highest, followed by the maximum method, the average method is worst, and we introduce the index of average image score, intuitively explained the conclusion. Thus, we can conclude that current results further enrich the form of gossip and explores the impact of different multi - valued gossip computational methods for promoting cooperation.