Evaluating the Impact of Torpedo Bats in Major League Baseball Using JMP (2025-US-PO-2409)

Torpedo bats, which diverge from the traditional tapering bat shape, have drawn widespread attention following a surge in power production from players adopting them. While the spike in home runs has sparked public interest, this project investigates whether the unique bat design corresponds with measurable changes in player performance.

Using MLB Statcast data accessed via PyBaseball and Baseball Savant with Python code written by LLMs, this analysis begins with exploratory evaluation of success metrics such as launch speed, square-ups, and blasts. These indicators revealed promising patterns for some players, prompting deeper investigation through data mining and modeling. Stratified sampling by launch speed was used to balance torpedo and non-torpedo bat data, and tree-based methods uncovered consistent differences in swing length, bat speed, and attack angle, suggesting the bat design alters swing mechanics in meaningful ways.

JMP tools such as Graph Builder, Multiple File Import, and Workflow Builder supported the analysis, while JMP 19’s Data Table Tags enabled nested variable groupings. Though limitations in torpedo bat usage reporting remain, the combination of modern player tracking data and statistical modeling offers insight into how torpedo bats interact with swing mechanics and how these design changes may be shaping the future of offensive performance in baseball.