This paper develops a stochastic learning model in which agents choose a limited number of partners to play size-dependent minimum-effort games. The payoff for each agent in this interaction depends on the minimum of the efforts of her partners and herself, and is increasing with the size of her...