“…This is less related to our use of the notion of "sampling dynamics," which is in line with the literature cited above. Salant and Cherry (2020) (see also Sawa and Wu, 2023) generalized sampling dynamics by allowing new agents to use various procedures to infer from their samples the aggregate behavior of the opponents (in addition to allowing for payoff heterogeneity in the population). Salant and Cherry paid special attention to unbiased inference procedures in which the agent's expected belief about the share of opponents who play an action coincides with the sample mean.…”