|
Abstract:
|
Starting pitchers in Major League Baseball are constantly evolving with the help of innovative technology to become more successful on the mound. Pitchers are always seeking improvement, and one way they may go about improving is by adding another pitch to their arsenal. In addition, teams are also looking to improve their pitching staff by adding both depth and breadth to a potential starting rotation. With the ever growing information available about starting pitchers, one question has not yet been explored: is there an ideal combination of pitching profiles within a starting rotation that maximizes a team’s success against opponents? To investigate this question, we gathered data on the pitch distributions for starting pitchers in 2020 from Statcast. Using clustering techniques, we create clusters of pitchers based on their pitch distributions in order to define unique pitching profiles. Then, we computed a measure of variation in pitching profiles for each teams’ starting pitchers. Lastly, we compared team performance as a function of variation in pitching profiles to identify an optimal starting rotation of pitchers.
|
