This presentation focuses on the classification of pitchers' straight pitches. Straight pitches have different characteristics depending on vertical change (hop, drip) and horizontal change (shoot change, slider change), as well as ball speed and rotation speed. For Major League Baseball (MLB), this data can be easily collected on the Baseball Savant page. Therefore, for pitchers' straight pitches in the 2024 MLB regular season, we use a Gaussian Mixture Model (GMM) to classify these straight pitches into several clusters, and visualize the results using coordinates of vertical and horizontal change. In addition, we aggregate the outcomes (strikeout rate, play results, etc.) for each cluster and quantitatively evaluate the characteristics and performance of each cluster. We also clarify which cluster the straight pitches thrown by Japanese MLB pitchers (Kikuchi, Imanaga, Yamamoto, etc.) belong to, and compare them with the average outcome of that cluster to examine the superiority or inferiority of their performance.
