ABSTRACT
Game theory models can be used in a predictive fashion. When they clearly mirror an actual situation, they can be used to predict which actions will ensue. Experimental social psychology also places a primary emphasis on representative realism--adequately representing variables which provide the framework for study--rather than descriptive realism of any natural occurrences. A model never gives the entire situation found in the real world. Some models represent more closely than others the detailed features of reality, but game theory models seldom have great ambitions in this direction. They aspire only to be highly summarized pictures of one aspect of the situation--that aspect which shows how the values of decision makers may impinge on one another's actions. Game theory models are therefore seldom used in a truly predictive manner. Typically they occur rather in after-the-fact reconstructions which help us understand why a particular result followed from a given situation.
