Examining the concurrent validity of markerless motion capture in dual-athlete team sports movements

Markerless motion capture systems could help collect kinematic data in team sports. However, the validity of these systems when measuring multiple individuals simultaneously has yet to be determined. The current study aims to assess the concurrent validity of full-body kinematics during dyadic sports tasks of a markerless compared to a marker-based motion capture system. Twelve couples performed three dyadic movements: the walk high-five, dribble simulation, and double sidestep. Movements were simultaneously recorded with both systems. For each movement, the segment orientations, joint angles and center of mass were compared using the standardized typical estimate of the error, correlation coefficient, root mean squared difference (RMSD), and Bland-Altman analysis. There was an acceptable agreement in all segment orientations (RMSD < 12.5°), except for the hand segment. The flexion/extension joint angles showed an acceptable similarity between the two systems, excluding the wrist. There was a close-to-perfect similarity and agreement in the center of mass overall movement tasks. Based on the study results, the markerless motion capture system can be used for kinematic analysis of dyads in team sports, similar to how it is used in single-person movement analysis.