Biological movement variability during the sprint start: performance enhancement or hindrance?

In the current study, we quantified biological movement variability on the start and early acceleration phase of sprinting. Ten male athletes aged 17-23 years (100-m personal best: 10.87 ± 0.36 s) performed four 10-m sprints. Two 250-Hz cameras recorded the sagittal plane action to obtain the two-dimensional kinematics of the block start and initial strides from subsequent manually digitized APAS motion analysis. Infra-red timing lights (80 Hz) were used to measure the 10-m sprinting times. The coefficient of variation (CV%) calculation was adjusted to separate biological movement variability (BCV%) from estimates of variability induced by technological error (SEM%) for each individual sprinter and measure. Pearson's product-moment correlation and linear regression analysis were used to establish relationships between measures of BCV% and 10-m sprint start performance (best 10-m time) or 10-m sprint start performance consistency (10-m time BCV%) using SPSS version 12.0. Measurement error markedly inflated traditional measures of movement variability (CV%) by up to 72%. Variability in task outcome measures was considerably lower than that observed in joint rotation velocities. Consistent generation of high horizontal velocity out of the blocks led to more stable and faster starting strides.