Excess payoff dynamics in games
提出了一类博弈中的超额收益动态，其中策略的增长仅取决于其当前比例和超额收益，并揭示了其与复制者动态、布朗-冯·诺伊曼-纳什动态等经典模型的关系。
We present the family of Excess Payoff Dynamics for normal-form games, where the growth of a strategy depends only on its current proportion and the excess payoff, i.e., the payoff advantage of the strategy over the average population payoff. Requiring dependence only on the own excess payoff and a natural sign-preserving condition, the class essentially reduces to aggregate monotonic dynamics, a functional generalization of the Replicator Dynamics. However, Excess Payoff Dynamics also include a different subclass which contains the Replicator Dynamics, the Brown-von Neumann-Nash Dynamics, and other interesting examples as, e.g., satisficing dynamics. We also clarify the relation to excess demand dynamics from microeconomics.