Valuation of Pitcher's Skill Using Two-dimensional Kernel Density in Baseball

In this study, we evaluate pitchers in the 2019 Major League Baseball season from the viewpoint of swinging by the batter. However, we assessed the batters' swing from the run expectancy-value, and it was found that the pitcher has less risk when they swing than when they did not. Therefore, the pitcher that swung by the batter is a better pitcher than other pitchers. Based on this, we propose an evaluation method. The method used probability estimation based on the generalized additive model. Finally, we evaluate pitchers in 2019 and thus consider the applicability of the proposed model.