Kin selection and direct reciprocity are two most basic mechanisms for promoting cooperation in human society. Generalizing the standard models of the multi-player Prisoner's Dilemma and the Public Goods games to allow for asymmetric cost-benefit ratios across the players, we study the effects of genetic relatedness on cooperation in the context of repeated interactions. Two sets of interrelated results are established: a set of analytical results focusing on the subgame perfect equilibrium and a set of agent-based simulation results based on an evolutionary game model. We show that in both cases increasing genetic relatedness does not always facilitate cooperation. Specifically, kinship can hinder the effectiveness of reciprocity in two ways. First, the condition for sustaining cooperation through direct reciprocity is harder to satisfy when relatedness increases in an intermediate range. Second, full cooperation is impossible to sustain for a medium-high range of relatedness values. Moreover, individuals with low cost-benefit ratios can end up with lower payoffs than their groupmates with high cost-benefit ratios. Our results point to the importance of explicitly accounting for within-population heterogeneity when studying the evolution of cooperation.

亲缘选择和直接互惠是人类社会促进合作的两种最基本的机制。我们推广了多人囚徒困境和公共物品博弈的标准模型，以允许参与者之间存在不对称的成本效益比，我们研究了重复互动背景下遗传相关性对合作的影响。建立了两组相互关联的结果：一组关注子博弈完美均衡的分析结果和一组基于演化博弈模型的基于主体的模拟结果。我们表明，在这两种情况下，增加遗传相关性并不总是有利于合作。具体来说，亲属关系可以通过两种方式阻碍互惠的有效性。首先，当相关性在中等范围内增加时，通过直接互惠维持合作的条件就更难满足。其次，对于中高范围的相关性值来说，全面合作是不可能维持的。此外，成本效益比较低的个人最终获得的回报可能低于成本效益比较高的群体。我们的结果指出了在研究合作演化时明确考虑群体内异质性的重要性。