An individual approach to monitoring locomotive training load in English Premier League academy soccer players

Research that examines the activity profiles of players in team sports (specifically football) has traditionally used parameters to describe activity that are based around the velocity of movement. ¹ This approach is probably based on the theoretical assumption that such classifications can help potentially provide information on both the energy system contribution and the overall energy expenditure associated with the activity(s). The classification of activities in the majority of the research is based on arbitrary thresholds that have unclear origins. Most commonly, these thresholds seem to be more of a product of commercial companies who specialise in providing tracking data for practical athlete monitoring than any scientifically agreed criteria.

Concerns with the generalizability of such thresholds to a range of different individuals have resulted in attempts to create more “bespoke” threshold ranges. This process is common in the applied practice of elite football organisations and in a number of published peer-reviewed research projects. ² While such approaches make “intuitive” sense, the difficulties in creating thresholds that generate theoretically better “insights” into the demands of training/games may mean that such attempts do little more than “muddle” our understanding. Issues that could impact the thresholds generated to individualise data include:

The choice of the “correct” speed that is used to “scale” the thresholds (e.g. maximal sprint speed, maximal speed achieved in a match/training, maximal aerobic speed, etc.). The frequency with which these maximal speeds (whichever is chosen) needs re-evaluation to ensure physiological relevance of the scaling (weekly, monthly, etc.)

The ability of thresholds obtained from incremental exercise protocols that use steady rate stages (e.g. threshold type protocols) to inform physiological responses in unpredictable rapidly changing intermittent exercise tasks.

These methodological points may represent only a small number of the potential factors that could impact the appropriateness of the scaling of velocity-based thresholds. Unfortunately, little current research seems focussed on understanding the potential methodological and theoretical issues associated with individualising thresholds as the majority of papers seem to simply describe the impact of different approaches to classification. It is clear that the use of a generalised threshold requires the acceptance of a number of assumptions that may limit the accuracy of the interpretation of the activity data. Such assumptions are, however, simpler conceptually than the layers of assumptions that may accompany an approach that is recent in its inception and based on limited data (and therefore not well understood). Such new ideas that are not yet grounded in a strong theoretical/methodological evidence base may simply multiply error in ways that are difficult to understand. Metaphorically, “chasing a rabbit down a hole” to respond to intuition too quickly has the potential to move us further away from obtaining a better understanding of the demands of the sport; an outcome that is in direct contrast to that originally intended. It therefore seems useful to spend more effort understanding the methodological/theoretical and practical impact of different approaches to scaling activity data in future research rather than simply demonstrating that another different scaling factors simply leads to another set of outcomes in the classification of activities. This type of research would build upon the insights created by the recent paper from Scott and Lovell ³ that casts doubt on the potential of individualised approaches to add additional practical insight to more traditional approaches.