The report presents a data‑driven scheme for classifying Major League Baseball hitters based on their ground‑reaction‑force (GRF) sequence and timing. Using the Baseball Performance Lab’s GRF Master Database (80 swings across three stance variants) and 12 engineered sequence‑and‑timing features, Ward hierarchical clustering was applied. Validation across k = 2‑7 showed a silhouette peak at k = 4, yielding four biomechanically distinct hitter buckets that each exceed the n ≥ 5 confidence threshold. The buckets are: (1) Ground‑Up Rotators (44% of the sample) – a classic “twist‑then‑shift” pattern where back‑foot free moment fires first; (2) Lateral Drivers (38%) – a “shift‑then‑twist” pattern with an early lateral push and a longer P1→P2 gather; (3) Slow Rotation Phase (11%) – standard front‑half sequence but a prolonged bat‑vertical‑to‑bat‑parallel rotation (≈ 67 ms vs 55 ms average); and (4) Delayed Vertical (8%) – back‑foot vertical load peaks second, co‑peaking with lateral and free‑moment forces, producing a compressed simultaneous ground load. For each bucket the report details the order of force peaks, phase‑duration breakdowns, exemplar hitters closest to the centroid, and stance mix. A summary page juxtaposes the buckets, highlighting that the two dominant patterns cover 82% of hitters while the smaller buckets serve as diagnostic categories. The document concludes with practical implications for scouting and coaching, emphasizing that bucket membership is diagnostic rather than deterministic, with ~15% of hitters expected to sit near cluster boundaries and shift across captures.
Topics
Methodology: GRF Master Database and Ward Hierarchical Clustering
Construction of the dataset (80 swings, three stance variants), selection of 12 sequence‑and‑timing features, application of Ward hierarchical clustering, exploration of k = 2‑7, silhouette analysis, and criteria for choosing k = 4.
Four Biomechanical Hitter Buckets – Overview and Population Distribution
Presentation of the four clusters (Ground‑Up Rotators, Lateral Drivers, Slow Rotation Phase, Delayed Vertical) with percentages, sample sizes, and the rationale for treating them as diagnostic categories.
Ground‑Up Rotators – Twist‑Then‑Shift Kinetic Chain Pattern
Detailed description of the classic pattern where back‑foot free moment fires first, followed by lateral push; includes force‑peak order table, phase‑duration values (635 ms, 420 ms, 53 ms), exemplar hitters, and stance mix.
Lateral Drivers – Shift‑Then‑Twist Pattern with Extended Gather
Characterization of the alternative pattern where back‑foot lateral force peaks early, followed by rotation; highlights the longer P1→P2 gather (~687 ms), force‑peak order, phase‑duration breakdown, exemplar hitters, and stance mix.
Slow Rotation Phase – Prolonged Bat‑Vertical to Bat‑Parallel Transition
Explanation of the bucket where the bat spends more time rotating from vertical to parallel (≈ 67 ms vs 55 ms average), with a quicker initial gather, force‑peak order, total swing time (~989 ms), and exemplar hitters.
Delayed Vertical – Simultaneous Trail‑Side Ground Load
Analysis of the rare pattern where back‑foot vertical load peaks second, co‑peaking with lateral and free‑moment forces, resulting in a compressed ground‑load sequence; includes force‑peak order, phase durations, exemplar hitters, and stance mix.
Practical Implications for Scouting and Coaching
Synthesis of findings: two dominant patterns cover 82% of hitters, the remaining 18% are diagnostic categories; discussion of bucket stability, boundary cases (~15% of hitters), and how the scheme can inform player evaluation and development.