Reliability and Validity of Quantitative Video Analysis of Baseball Pitching Motion

Abstract

Video recordings are used to quantitatively analyze pitchers’ techniques. However, reliability and validity of such analysis is unknown. The purpose of the study was to investigate the reliability and validity of joint and segment angles identified during a pitching motion using video analysis. Thirty high school baseball pitchers participated. The pitching motion was captured using 2 high-speed video cameras and a motion capture system. Two raters reviewed the videos to digitize the body segments to calculate 2-dimensional angles. The corresponding 3-dimensional angles were calculated from the motion capture data. Intrarater reliability, interrater reliability, and validity of the 2-dimensional angles were determined. The intrarater and interrater reliability of the 2-dimensional angles were high for most variables. The trunk contralateral flexion at maximum external rotation was the only variable with high validity. Trunk contralateral flexion at ball release, trunk forward flexion at foot contact and ball release, shoulder elevation angle at foot contact, and maximum shoulder external rotation had moderate validity. Two-dimensional angles at the shoulder, elbow, and trunk could be measured with high reliability. However, the angles are not necessarily anatomically correct, and thus use of quantitative video analysis should be limited to angles that can be measured with good validity.