The paradox of backward induction, like the puzzles raised by equilibrium refinement, is mainly a problem for those who view game theory as contributing to a normative theory of rationality (specifically, as contributing to that larger theory the theory of strategic rationality). The non-psychological game theorist can give a different sort of account of apparently “irrational” play and the prudence it encourages. This involves appeal to the empirical fact that actual agents, including people, m