Pitching strategy evaluation via stratified analysis using propensity score

Abstract

Recent measurement technologies enable us to analyze baseball at higher levels. There are, however, still many unclear points around the pitching strategy. The two elements make it difficult to measure the effect of pitching strategy. First, most public datasets do not include location data where the catcher demands a ball, which is essential information to obtain the battery's intent. Second, there are many confounders associated with pitching/batting results when evaluating pitching strategy. We here clarify the effect of pitching attempts to a specific location, e.g., inside or outside. We employ a causal inference framework called stratified analysis using a propensity score to evaluate the effects while removing the effect of disturbing factors. We used a pitch-by-pitch dataset of Japanese professional baseball games held in 2014-2019, which includes location data where the catcher demands a ball. The results reveal that an outside pitching attempt is more effective than an inside one to minimize allowed run on average. Besides, the stratified analysis shows that the outside pitching attempt was always effective despite the magnitude of the estimated batter's ability, and the ratio of pitched inside for pitcher/batter. Our analysis would provide practical insights into selecting a pitching strategy to minimize allowed runs.

Estimating the effect of hitting strategies in baseball using counterfactual virtual simulation with deep learning

“…Even if we obtain the batting ability when changing the strategy in the above approach, in real-world baseball data, there should be biases in the data with and without changing the strategy. In sports studies, to investigate the causal effect of some plays or timeouts in many sports, propensity score matching or other approaches were used in the previous work (Fujii et al, 2022;Gibbs, Elmore, & Fosdick, 2020;Nakahara et al, 2022b;Toumi & Lopez, 2019;Vock & Vock, 2018;Yam & Lopez, 2019). Such extension to reduce the biases of data in our setting would also be future work.…”