Svoboda Jakub, Chatterjee Krishnendu
Institute of Science and Technology Austria, Klosterneuburg 3400, Austria.
Proc Natl Acad Sci U S A. 2025 Dec 23;122(51):e2524109122. doi: 10.1073/pnas.2524109122. Epub 2025 Dec 15.
Evolutionary games provide a flexible mathematical framework for many problems in biology and social evolution. Prisoners' dilemma, and in particular, the important special case of donation games, represents social dilemmas where cooperation is mutually beneficial, yet defection is preferred by selfish agents. In evolutionary games on networks, the agents interact over a population structure. The existence of population structures that promote cooperative behavior is a fascinating and active research topic. Previous research establishes structures promoting cooperation in the limit of weak selection where the benefit-to-cost ratio exceeds 1.5. The existence of such structures for medium and strong selection for [Formula: see text] and for weak selection for [Formula: see text] has been a long-standing open question. First, we answer the open questions in the affirmative: For every selection strength and every [Formula: see text], we construct networks promoting cooperation. Second, we present a robustness result with respect to and selection strength: Our structures promote cooperation for a range of these parameter values rather than specific parameter values. Finally, we supplement our theoretical results with simulation results on small population structures that show the effectiveness of our construction over well-studied population structures.
进化博弈为生物学和社会进化中的许多问题提供了一个灵活的数学框架。囚徒困境,尤其是捐赠博弈这一重要的特殊情况,代表了社会困境,即合作对双方都有益,但自私的个体更倾向于背叛。在网络上的进化博弈中,个体在种群结构上进行互动。促进合作行为的种群结构的存在是一个引人入胜且活跃的研究课题。先前的研究确定了在弱选择极限下(收益成本比超过1.5)促进合作的结构。对于中等和强选择下的[公式:见原文]以及弱选择下的[公式:见原文],这种结构的存在一直是一个长期悬而未决的问题。首先,我们肯定地回答了这些开放性问题:对于每一种选择强度和每一个[公式:见原文],我们构建了促进合作的网络。其次,我们给出了关于[公式:见原文]和选择强度的稳健性结果:我们的结构在一系列这些参数值范围内促进合作,而不是特定的参数值。最后,我们用关于小种群结构的模拟结果补充我们的理论结果,这些结果表明我们构建的结构比经过充分研究的种群结构更有效。