Introduction

Game theory is a branch of so-called Bayesian rational choice theory (RCT). It has two distinct forms of application:

1. (i) explaining individuals' behaviour in social settings by their motives and reasons;

2. (ii) as an abstract model for the analysis of social structure, within the paradigm of methodological individualism (MI).

Game theory is explanatorily useful only to the extent that it models individuals' motives and reasons appropriately. Modelling, by contrast, aims not at replicating the world, but at artificially isolating features in order to study their potential or dynamics. An explanatory approach fails if it cannot explain observable real-life behaviour. An abstract model, by contrast, can be a very fruitful analytical tool exactly when it fails if it is precise enough to tell us why it fails, and how the model can be enriched, changed or modified. Insights achieved from abstract modelling do not themselves explain phenomena but can be used in the development of explanatory hypotheses or even concept-formation; but these hypotheses then have to be tested independently.

The first section of this chapter clarifies the basic concepts and assumptions of RCT: rational choice, preference, expected utility and the structure of modern utility theory. The subsequent section turns to game theory proper and remarks on its relationship to the broader concept of RCT.