Validity and reliability of a computer-vision-based smartphone app for measuring barbell trajectory during the snatch.

The purpose of this study was to investigate the validity of a smartphone app to measure biomechanical barbell parameters during the snatch. Ten collegiate NCAA division I athletes performed two repetitions each at 40, 50, 60, 70, and 80% of their 1-repetition maximum snatch. Barbell motions were simultaneously recorded with a motion capture system and the smartphone app. The motion capture system recorded the 3-D position of a reflective marker attached to the end of the barbell while the smartphone app was used to record sagittal plane video and track the shape of the weight plate from which the barbell center was derived. Peak forward (PFD) and backward (PBD) displacements and peak vertical displacement (PVD) and velocity (PVV) were calculated from both sets of data. Significant, strong to very strong Pearson's product-moment correlation coefficients between both systems were noted for all parameters (r = 0.729-0.902, all p < 0.001). Small significant biases between systems were observed for PVD (ES = 0.284, p < 0.001) and PFD (ES=0.340, p < 0.01), while trivial to small, non-significant biases were observed for PBD (ES = 0.143) and PVV (ES = -0.100). Collectively, the results suggest that the app can provide biomechanical data of barbell motions similar to a 3-D motion capture system.