Whereas orthodox game theory relies on the unrealistic assumption of (commonly known) perfect rationality, participants in game playing experiments are at best boundedly rational. This makes it necessary to supplement orthodox game theory by a behavioral theory of game playing. We first point out that this applies also to (one person-) decision theory. After reviewing the influential experiments based on repeated games and the ultimatum game the typical reactions to the striking experimental results are categorized. Further sections are devoted to alternating offer bargaining and characteristic function experiments.