Assessing maximal sprinting speed in soccer - criterion validity of commonly used devices

This study aimed to investigate the criterion validity of commonly used devices to assess maximal sprinting speed (MSS) in soccer. Thirty elite youth soccer players completed three trials of a 30-m sprint test to assess MSS. All sprints were simultaneously captured via a radar gun (Stalker ATS II), timing gates (Smartspeed Pro, Fusion Sport), a magnetic timing system (Humotion SmarTracks) and a global navigation satellite system (GNSS) (Kinexon Perform GPS Pro). The radar gun and the GNSS recorded sprinting speed continuously, while the fastest 5-m split during the 30-m sprint was used for the timing gates and the magnetic system. The best trial of the radar gun (i.e. criterion measure) and corresponding values of the other devices were analyzed. Equivalence testing was performed to assess the statistical equivalence of MSS between the radar gun and the three other devices against a difference value of ± 0.36 km/h and Bland & Altman's 95% limits of agreement (LoA) were computed to investigate the agreement between MSS results. Differences between GNSS versus radar gun suggested a lack of systematic bias (-0.01 km/h, 95% confidence interval [CI] -0.15 to 0.15 km/h), whereas timing gates-based MSS assessments were prone to larger uncertainty compared to the criterion method (-0.19 km/h, 95% CI: -0.37 to 0.00 km/h) given the pre-defined region of equivalence. The magnetic system (-0.54 km/h; -0.71 to -0.37 km/h) overestimated MSS compared to the radar gun with mean differences being non-equivalent. Based on the practically important difference bounds of ± 0.36 km/h, the width of the 95% LoA was broad enough to suggest a lack of reasonable agreement for MSS assessment regardless of device of interest (GNSS: -0.79 to 0.78 km/h, timing gates: -0.79 to 1.16 km/h, magnetic system: -0.24 to 1.32 km/h). While our results suggested a lack of systematic bias for the investigated GNSS and the timing gates when compared against the radar gun for MSS assessment over 30 m in elite youth soccer players on a team level, the width of the 95% LoAs did not indicate reasonable measurement interchangeability on an individual level. Based on the present results, we do not recommend using the magnetic system for both group and individual analyses in this population.