Reproducible Data and Code Package for Pareto-Efficient Performance–Loading Profiling of High-Velocity Fastball Pitching Using OpenBiomechanics Data

This repository contains the reproducible data, analysis code, statistical outputs, tables, and figures associated with the manuscript “A Pareto-Efficient Performance–Loading Framework for High-Velocity Fastball Pitching Using OpenBiomechanics Data”. The study is a secondary analysis of publicly available baseball pitching biomechanics data from the OpenBiomechanics Project. The analysis developed a Pareto-efficient performance–loading framework for identifying high-velocity fastballs with favorable velocity-adjusted shoulder–elbow loading profiles. Pitch velocity was treated as the performance objective, whereas velocity-adjusted elbow varus moment and shoulder internal rotation moment residuals were treated as loading objectives. This Zenodo package includes the raw OpenBiomechanics input files used in the analysis, processed analysis datasets, complete Python scripts, statistical output files, supplementary tables, manuscript figures, supplementary figures, and documentation required to reproduce the reported results. The scripts cover data screening, loading-residual construction, Pareto-efficient profile identification, velocity-threshold sensitivity analyses, candidate biomechanical feature comparisons, pitcher-level cluster bootstrap analysis, continuous association models, leave-one-pitcher-out influence diagnostics, table generation, and figure generation. The materials are provided to support transparency, peer-review reproducibility, and reuse of the analysis workflow. The findings derived from these materials should be interpreted as cross-sectional, exploratory, and hypothesis-generating. They do not establish causal mechanisms, predict injury risk, or provide direct technical prescriptions for pitchers. The original OpenBiomechanics Project data are publicly available from the OpenBiomechanics Project and Driveline Baseball. This archive contains the processed materials and reproducible analysis package generated for the present secondary analysis.