Individuality and reproducibility in high-speed motion of volleyball spike jumps by phase-matching and averaging.

For analysis of knee injuries in volleyball, we developed a system to obtain kinematic waveforms about the high-speed motion of volleyball spike jumps that could examine their individuality and reproducibility, with phase-matching and averaging. The form of six female players was recorded in two sessions of ten jumps each, with a video motion-analysis system, VICON, at the sampling frequency of 240 Hz. To identify individual jump characteristics despite differences from jump to jump, we averaged up to ten data sets with phases matched. For such matching, we defined and used a statistic, epsilon, as the arithmetic mean of all absolute differences in each of the angles at each time sampled, based on the least squares method. Statistical evaluation with adjusted coefficients of multiple determination and epsilon showed that intra-individual variation between the two sessions was significantly smaller than inter-individual variation; their means differed by more than 2 standard deviations (SD) and 3SD, respectively. Regression analysis showed that this system was reliable (p < 0.001). We concluded that the averaged data sets obtained by the system could embody the individuality of spike jumps, with satisfactory reproducibility.