The use of relative speed thresholds in team sports: Applications for GPS analysis

Data from global positioning system (GPS) technology are typically presented as the distances covered in specific locomotor categories (e.g., walking, jogging, striding, sprinting). Differences are found when categorisations are made using either pre-defined absolute thresholds or thresholds relative to maximum speed. However, there are two distinct methods of using relative speed thresholds currently employed in the literature, although no study has attempted to compare them. Therefore, the purpose of this study was to compare the differences in data when analysing the same GPS files relative to speed, using either a maximum velocity sprint (Vmax) or a maximum velocity achieved during match-play (Vpeak). Following institutional ethics approval, 99 GPS files were analysed from rugby union match-play and split between forwards (n=59) and backs (n=40). The male participants involved were part of a Regional Academy and had the following characteristics (age: 17.5 ± 0.7 years; stature: 183.6 ± 6.6 cm; body mass: 90.6 ± 10.6 kg). Vmax was established by players performing a maximum 40 m sprint, whilst Vpeak was defined as the maximum velocity achieved during each match. The locomotor categories were defined as walking 0-20%, jogging 20-50%, striding 50-80% and sprinting 80-100% (Duthie et al., 2006) of either Vmax or Vpeak. Data were analysed using magnitude based inferences. The mean Vmax and Vpeak for all players were 8.7 ± 0.6 m.s-1 and 7.2 ± 0.9 m.s-1, respectively. There were almost certain differences in walking (2088 ± 298 vs. 1611 ± 435 m), striding (670 ± 244 vs. 1197 ± 375 m) and sprinting (28 ± 29 vs. 145 ± 73 m) between Vmax and Vpeak, for forwards. There was also a likely difference in jogging (2674 ± 313 vs. 2502 ± 301 m). Very likely differences were found for walking (2414 ± 288 vs. 2177 ± 347 m) and striding (708 ± 159 vs. 927 ± 347 m) for backs. There was also an almost certain difference in sprinting (66 ± 41 vs. 151 ± 49 m) whilst an unclear difference was found for striding (2409 ± 433 vs. 2338 ± 352 m) for backs. The use of relative thresholds using Vpeak seems to overestimate the distance covered in striding and sprinting whilst underestimating walking and jogging. Practitioners should look to use Vmax for relative speed thresholds as Vpeak from match-play is likely to change from match-to-match and consequently misrepresent the movement demands that players are exposed to.