Psychological games and sequential rationality

In psychological games the payoff to each player depends not only on what every player does but also on what he thinks every player believes, and on what he thinks they believe others believe, and so on. In equilibrium, beliefs are assumed to correspond to reality. Yet psychological games and psychological equilibria allow one to model belief-dependent emotions such as anger and surprise that are problematic for conventional game theory. We are particularly interested in issues of sequential rationality for psychological games. We show that although backward induction cannot be applied, and "perfect" psychological equilibria may not exist, subgame perfect and sequential equilibria always do exist.