Direct arm angle versus release-based proxies in Major League Baseball pitchers: a large-scale validation study.

Abstract

Arm slot has been implicated in pitching biomechanics and injury risk, yet large-scale studies relied on indirect release-based proxies, yielding contradictory findings. With Statcast's Hawk-Eye system providing an anatomically defined measure of arm angle, this study evaluated how well surrogate metrics track fluctuations in true arm orientation. We analysed Major League Baseball pitchers using two approaches: (1) a year-over-year comparison of 336 pitchers who threw ≥ 100 four-seam fastballs in both 2023 and 2024, and (2) a pitch-level analysis of 63,824 four-seam fastballs from the 10 highest-volume pitchers between 2020 and 2025. In the season-over-season analysis, vertical release position changes correlated most strongly with arm angle changes (r = 0.74, ρ = 0.68; p < 0.001), whereas a geometric proxy showed moderate correlation (r = 0.47, ρ = 0.51; p < 0.001), and horizontal release was inversely correlated (r = -0.38, ρ = -0.41; p < 0.001). Pitch-level analyses revealed similar patterns, with vertical release consistently tracking short-term arm angle drift (r = 0.35-0.71). These findings show release-based proxies only partially track true arm angle. While vertical release is the strongest proxy, fidelity varies across pitchers and timeframes. This variability complicates the interpretation of prior injury-risk literature, highlighting the importance of incorporating direct arm angle into future biomechanical research to better investigate pitching mechanics and injury mechanisms.