Acceleration Profiles in Elite Australian Soccer

Response to M. Wilkinson’s and R. Akenhead’s Letter to the Editor:

Dear Editor,

We would like to thank Dr. Wilkinson and Mr. Akenhead for their feedback [4] regarding our recent study describing the activity profiles of elite Australian soccer players [3]. Their comments on accounting for repeated measures (multiple matches) for each player by using a linear-mixed model and covarying for player are correct and we wish to clarify our reasoning for the statistical approach used in our study. We acknowledge that the suggested approach of mixed method modelling, covarying for player would eliminate one source of bias within the data. However, this approach assumes that each player’s activity is most dependent on the player themselves. In team sport a player’s activity is inherently associated with their direct opponent’s activity. Whilst obviously the player is an important variable to consider in an analysis, we suggest that the activity of a player is influenced by the tactics of their team, the opposition team and most importantly their direct opponent. Thus the activity profile of a given individual will vary greatly from match-to-match. For example in our study, match-to-match variation in accelerations on average for each player ranged from 38 to 84 for midfielders, 46 to 74 for central defenders and 47 to 71 for forwards. Further investigation of the data also reveals that high performing players one week did not necessarily perform better than their team-mates in following weeks (in essence their ranking in the team was also variable).

It is with this reasoning that we favour the approach of considering each player as an independent case. This method allowed us to consider all data and make comparisons to other research in the field where this method is commonly used [1] [2]. We acknowledge that this is not a perfect method and would advocate that future research may need to consider not only the player, but also their opposing player. Clearly in elite team sport research, where we are using data with a potential to provide an advantage to opposing teams it will be difficult to have multiple clubs share data over the course of a season to achieve consistent player pairings. This issue of covariance according to the player is certainly something worth noting, but by considering positional differences and retaining the majority of data collected we feel our results are still valuable to the sport science community.

• References

• 1 Bradley PS, Sheldon W, Wooster B, Olsen P, Boanas P, Krustrup P. High-intensity running in English FA Premier League soccer matches. J Sports Sci 2009; 27: 159-168
• 2 Di Salvo V, Pigozzi F, González-Haro C, Laughlin MS, De Witt JK. Match performance comparison in top English soccer leagues. Int J Sports Med 2012; Epub 2012/11/26. DOI: 10.1055/s-0032-1327660
• 3 Varley MC, Aughey RJ. Acceleration profiles in elite Australian soccer. Int J Sports Med 2012; Epub 2012/08/17. DOI: 10.1055/s-0032-1316315
• 4 Wilkinson M, Akenhead R. Violation of statistical assumptions in a recent publication?. Int J Sports Med 2013; 34: 281