Tatistical Data Useful for the Operation of a Baseball Team

Baseball is a game well suited for operations analysis, and is already well provided with statistical records of past performances of individuals. For the making of operational decisions before and during a game, the manager must use the existing records of past performances with caution, but he could derive considerable assistance from types of statistics not commonly collected. One such type would show the average performance of a large number of batters in selected circumstances. An example is given which shows that an advantage of 0.032 applies to the batter whose “handedness” is opposite to that of the pitcher. The sampling error for a season's batting average is too great to permit much useful analysis of the proficiency of one individual in selected circumstances, although special records might be kept in batting practice. The sampling error should be allowed for when judging a new player before he has had several hundred times at bat. Aside from individual performance, data on the chances of a team in selected circumstances would be useful for deciding tactics and changes in the line-up. As an example, an analysis is given of the expectation of runs after a team has filled the bases, and its application to the choice of a defensive strategy.