Electronic performance and tracking systems are becoming a standard in many sports to automate data collection and gather more profound insights into performance and game dynamics. In large soccer clubs and federations, the problem is that different electronic performance and tracking systems report different kinematic parameters and performance indicators, which should be the same. Furthermore, a drawback in recent validation studies is the subdivision of speed and acceleration zones in validating the systems, as we show that the kinematic parameters are interdependent. We propose a new method to classify multidimensional validation outputs with a clustering approach. Additionally, we offer a data-driven strategy to reduce errors between distinct systems when data from different electronic performance and tracking systems are compared and show the method`s effectiveness with data collected in a validation study. This error reduction strategy can be applied to correct errors when no validation data is available.
© Copyright 2023 Sports Engineering. The Faculty of Health & Wellbeing, Sheffield Hallam University. All rights reserved.