{"title":"网络重复博弈中的主导战略","authors":"Xiaochen Wang, Aming Li","doi":"arxiv-2409.04696","DOIUrl":null,"url":null,"abstract":"Direct reciprocity, stemming from repeated interactions among players, is one\nof the fundamental mechanisms for understanding the evolution of cooperation.\nHowever, canonical strategies for the repeated prisoner's dilemma, such as\nWin-Stay-Lose-Shift and Tit-for-Tat, fail to consistently dominate alternative\nstrategies during evolution. This complexity intensifies with the introduction\nof spatial structure or network behind individual interactions, where nodes\nrepresent players and edges represent their interactions. Here, we propose a\nnew strategy, ``Cooperate-Stay-Defect-Tolerate\" (CSDT), which can dominate\nother strategies within networked populations by adhering to three essential\ncharacteristics. This strategy maintains current behaviour when the opponent\ncooperates and tolerates defection to a limited extent when the opponent\ndefects. We demonstrate that the limit of tolerance of CSDT can vary with the\nnetwork structure, evolutionary dynamics, and game payoffs. Furthermore, we\nfind that incorporating the Always Defect strategy (ALLD) can enhance the\nevolution of CSDT and eliminate strategies that are vulnerable to defection in\nthe population, providing a new interpretation of the role of ALLD in direct\nreciprocity. Our findings offer a novel perspective on how cooperative strategy\nevolves on networked populations.","PeriodicalId":501043,"journal":{"name":"arXiv - PHYS - Physics and Society","volume":"185 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Dominant strategy in repeated games on networks\",\"authors\":\"Xiaochen Wang, Aming Li\",\"doi\":\"arxiv-2409.04696\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Direct reciprocity, stemming from repeated interactions among players, is one\\nof the fundamental mechanisms for understanding the evolution of cooperation.\\nHowever, canonical strategies for the repeated prisoner's dilemma, such as\\nWin-Stay-Lose-Shift and Tit-for-Tat, fail to consistently dominate alternative\\nstrategies during evolution. This complexity intensifies with the introduction\\nof spatial structure or network behind individual interactions, where nodes\\nrepresent players and edges represent their interactions. Here, we propose a\\nnew strategy, ``Cooperate-Stay-Defect-Tolerate\\\" (CSDT), which can dominate\\nother strategies within networked populations by adhering to three essential\\ncharacteristics. This strategy maintains current behaviour when the opponent\\ncooperates and tolerates defection to a limited extent when the opponent\\ndefects. We demonstrate that the limit of tolerance of CSDT can vary with the\\nnetwork structure, evolutionary dynamics, and game payoffs. Furthermore, we\\nfind that incorporating the Always Defect strategy (ALLD) can enhance the\\nevolution of CSDT and eliminate strategies that are vulnerable to defection in\\nthe population, providing a new interpretation of the role of ALLD in direct\\nreciprocity. Our findings offer a novel perspective on how cooperative strategy\\nevolves on networked populations.\",\"PeriodicalId\":501043,\"journal\":{\"name\":\"arXiv - PHYS - Physics and Society\",\"volume\":\"185 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-07\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Physics and Society\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.04696\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Physics and Society","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.04696","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Direct reciprocity, stemming from repeated interactions among players, is one
of the fundamental mechanisms for understanding the evolution of cooperation.
However, canonical strategies for the repeated prisoner's dilemma, such as
Win-Stay-Lose-Shift and Tit-for-Tat, fail to consistently dominate alternative
strategies during evolution. This complexity intensifies with the introduction
of spatial structure or network behind individual interactions, where nodes
represent players and edges represent their interactions. Here, we propose a
new strategy, ``Cooperate-Stay-Defect-Tolerate" (CSDT), which can dominate
other strategies within networked populations by adhering to three essential
characteristics. This strategy maintains current behaviour when the opponent
cooperates and tolerates defection to a limited extent when the opponent
defects. We demonstrate that the limit of tolerance of CSDT can vary with the
network structure, evolutionary dynamics, and game payoffs. Furthermore, we
find that incorporating the Always Defect strategy (ALLD) can enhance the
evolution of CSDT and eliminate strategies that are vulnerable to defection in
the population, providing a new interpretation of the role of ALLD in direct
reciprocity. Our findings offer a novel perspective on how cooperative strategy
evolves on networked populations.