Meta Level Equilibrium Selection for Two Illustrative Noncooperative Games.
Meta Level Equilibrium Selection for Two Illustrative Noncooperative Games.
复制标题
两个示例性非合作博弈的元水平均衡选择。
DOI：
10.1007/978-3-031-07996-2_5
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
A.B.
中科院分区：
文献类型：
--
作者：
Suzuki;T.;Leoneti;A.B.
各种均衡概念，如纳什、一般元理性（GMR）、顺序稳定性（SEQ）等，有助于预测冲突的不同结果。例如，在囚徒困境（PD）博弈中，一些参与者可以选择不合作（正如纳什均衡所预期的那样），也可以选择合作；也许这是因为他们害怕别人的反击（正如GMR所预料的那样）。在任何情况下，实际玩家的选择是选择一种理性模式，如纳什，GMR，甚至其他类型的理性，自己。本文的目标是提出一个新的模型来解释这种合理性是如何在游戏中内生选择的（当这种选择本身产生另一种游戏时，因为结果也取决于其他玩家的选择）。将程序选择的收敛理论应用于这种均衡选择的背景下，我们首先建立了均衡的元层次选择模型（均衡参与者站在哪个均衡参与者站在哪个均衡参与者站在哪个均衡参与者选择他们的行动）。然后，我们将该模型应用于两种典型游戏：PD游戏（合作问题）和Hi-Lo游戏（协调问题）。在这两种情况下，我们对为什么在冲突中可能进行合作/协调给出了新的解释。
Various equilibrium concepts such as Nash, general meta-rationality (GMR), sequential stability (SEQ), and so on, help anticipate different results of a conflict. In the Prisoner Dilemma (PD) game, for example, some players can choose not to cooperate (just as the Nash equilibrium anticipates) or they can choose to cooperate; perhaps this is because they fear others’ counteraction (just as GMR anticipates). In any case, the actual players’ choice is to select one of mode of rationality, such as Nash, GMR, or even other types of rationalities, on their own. The objective of this paper is to propose a new model that explains how such rationality is selected endogenously within the game (when such selection itself makes another game because the result depends also on other players’ choice). Applying the convergence theory of procedural choice into the context of such equilibrium selection, we first model the meta-level selection of equilibria (which equilibrium players stand on to choose which equilibrium players stand on to choose their actions). Then, we apply the model to two typical games: PD game (cooperation matters) and Hi-Lo game (coordination matters). In both cases, we give a new explanation of why cooperation/coordination is possible within the conflicts.