带有不对称参与者的多维竞赛一般化:均衡与最优奖品设计
Generalized Multidimensional Contests with Asymmetric Players: Equilibrium and Optimal Prize Design
AI总结:
本文研究了异质努力成本下多维竞赛的均衡与最优奖品设计,提出在特定条件下,奖品分配规则能确保唯一纯策略均衡,并刻画了努力最大化规则,发现跨战斗成本外部性不影响最优分配。
摘要:
我们研究了两个具有异质努力成本的玩家之间的n维竞赛,其中每个维度(战斗)被建模为Tullock竞赛。奖品分配规则是身份无关、预算平衡且随胜利数递增的。玩家的成本可以跨战斗分离或表现出跨战斗外部性。我们识别了一个充分条件,保证在所有可接受的奖品分配规则和所有玩家不对称程度下存在唯一的均衡,并且均衡是纯策略的。在此条件下,我们刻画了努力最大化奖品分配规则:如果一个玩家胜场比对手多至少预设边界的,其获得全部奖品;否则均分。在对称玩家情况下,当n为奇数时多数规则最优。有趣的是,在我们的设定中,跨战斗成本外部性不影响最优奖品分配规则。
英文摘要:
We study $n$-dimensional Tullock contests between two players with heterogeneous effort costs. We establish sharp bounds on discriminatory power $r$ ensuring that equilibrium is unique and pure across prize rules and cost asymmetries: $r\leq1/\lceil n/2\rceil$ when prizes depend only on the number of victories across all dimensions, and $r\leq1/n$ when they may also depend on player identity. An effort-maximizing rule splits the prize unless one player wins by a sufficient margin in the first case, and gives the weaker player a possibly zero headstart in the second. When both bounds hold, restricting prizes to victory counts alone can be costless despite unequal costs, but the percentage loss in total effort weakly increases with asymmetry.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。