Evaluating Hitting Skills of NPB Players with Logistic Regression Analysis

摘要

Luck or unluck is supposed to exist in baseball. In this paper, we consider unpredictable "lucky" or "unlucky" cases which are occurred on the ground in order to find out authentic ability of batters. If all cases are observable, hit probabilities can be calculated. However, it is impossible to gather all the information on the ground. Therefore, "luck" is defined as influence of information that cannot be observed. To provide more appropriate evaluation for NPB~1 players, this paper proposes a predicted BABIP using Logistic Regression and describes the predicted BABIP reducing the influence of luck on the assessment of hitting ability. Analysis results show that there is a possibility of appropriate evaluation on hitting abilities of NPB players using the predicted BABIP.