Impact of sensor/reference position on player tracking variables: Center of scapulae vs center of pelvis

Electronic performance and tracking systems (EPTS) traditionally rely on one of two body positions as the ultimate representative for the entire body in space: the upper torso between the scapulae (GPS- and radar-based systems) or the body's estimated center (optical and some radar-based systems). The aim of this study was to quantify the impact of the respective reference point upon the resulting kinematic tracking variables. We present a marker-based method comparing center of pelvis (COP) derived tracking variables with center of scapulae (COS) derived tracking variables in a 30 × 30 m (900 m²) VICON measurement area. Fourteen male soccer players completed a running circuit with prescribed team-sport specific movements. Results showed that differences between COP and COS heavily depend on the underlying movement characteristic. Low-speed running showed the lowest deviations whereas accelerated movements and movements with sharp changes in direction lead to a significant increase in the observed differences. Results further showed that COS sprinting distance was on average -44.65% (p < 0.001) lower in comparison to COP. Similarly, maximum speed obtained from COS was -2.94% (p = 0.001) lower in comparison to COP. On the contrary, maximum acceleration values of COS were on average 16.15% (p = 0.02) higher compared to COP. Our work illustrates that the anatomical reference point used to represent the entire body in space needs to be carefully considered in the interpretation of tracking variables delivered by different EPTS.