An inspection game model of the stolen base in baseball: A theory of theft

This paper models the stolen base play in baseball as a simple inspection game. The model offers equilibrium predictions relating the frequency with which a stolen base play is attempted, and the frequency with which it is successful. Using an extensive play-by-play dataset from 37 Major League Baseball seasons, qualitative and quantitative support is found for the predictions of the model. An exogenous change in the average number of runs scored per game during the period covered by the dataset provides a natural experiment; the equilbrium model predicts the change in the relationship between attempt and success frequencies observed in the data.