The evolution of fairness in the coevolutionary ultimatum games
The evolution of fairness in the coevolutionary ultimatum games
复制标题
DOI：
10.1016/j.chaos.2013.05.007
复制
发表时间：
2013-11-01
影响因子：
7.8
通讯作者：
Kokubo, Satoshi
中科院分区：
文献类型：
--
作者：
Miyaji, Kohei;Wang, Zhen;Kokubo, Satoshi
在最后通牒博弈中，两个参与者需要平分一定数量的金钱。通过响应者接受提议者的提议，最终实现金钱的共享，建立公平。在这里，我们计算出五个共同进化协议，其中策略（提供率p和阈值接受报价q）和底层网络拓扑结构可以共同进化，以证明如何链路服务的情况下影响公平性的演变。我们表明，这些共同进化协议的游戏的平衡显着影响。确定性规则导致了与人类行为实验结果不一致的过于奢侈或过于慷慨的结果。然而，概率规则产生公平的划分，类似于现实的情况。此外，我们还引入了一个幅度参数B来验证假设的链路服务协议的可扩展性。通过提高B，我们分析表明，在游戏中可以获得更好的性能，因为代理的总量也增加了。最后，我们进一步支持我们的结论，显示在这些共同进化的情况下，所谓的不切实际的切断事件。因此，我们提出了一个可行的方式来理解无处不在的自然公平，并希望它将激发进一步的研究，以解决社会分工。(C)2013爱思唯尔有限公司保留所有权利。
In the ultimatum games, two players are required to split a certain sum of money. Through the responder accepting the offer of proposer, the money will be shared and the fairness is built finally. Here, we figure out five coevolutionary protocols, where strategy (offering rate p and threshold for accepting an offer q) and underlying network topology can coevolve, to demonstrate how the link severing scenario affects the evolution of fairness. We show that the equilibrium of the games is significantly influenced by these coevolutionary protocols. The deterministic rules lead to overly lavish or overly generous result that is inconsistent with the outcome of human behavior experiment. However, the probabilistic rules produce fair division, similar to the realistic case. Moreover, we also introduce an amplitude parameter b to verify the plausibility of assumed link severing protocols. By means of enhancing b we analytically exhibit that preferable performance can be obtained in the game, since the total amount of agents increases as well. Last, we further support our conclusion by showing the so-called unrealistic severing events under these coevolution scenarios. We thus present a viable way of understanding the ubiquitous fairness in nature and hope that it will inspire further studies to resolve social division. (C) 2013 Elsevier Ltd. All rights reserved.