Evolution of tag-mediated altruistic behavior in one-shot encounters on large-scale complex networks

An agent-based evolutionary model of tag-mediated altruism is studied on large-scale complex networks addressing multiplayer one-shot Prisoner's Dilemma-like games with four competing strategies. Contrary to standard theoretical predictions, but in line with recent empirical findings, we observed that altruistic acts in non-repeated interactions can emerge as a natural consequence of recognition of heritable phenotypic traits such as visual tags, which enable the discrimination between potentially beneficial and unproductive encounters. Moreover, we identified topological regimes in which cooperation always prevails at short times, but where unconditional cooperators are favored over conditional tag-based helpers, even though the latter initially have a slight reproductive advantage. After very long times, we found that all four strategies appeared about equally often, meaning that only one quarter of agents refused cooperation egoistically. However, our study suggests that intra-tag generosity can quickly evolve to dominate over other strategies in spatially structured environments that are otherwise detrimental to cooperative behavior.