A new approach to quantify angles and time of changes-of-direction during soccer matches

Soccer players frequently perform change-of-directions (CODs) at various speeds during matches. However, tracking systems have shown limitations to measure these efforts. Therefore, the aim of the present study was to propose a new approach to measure CODs using a local positioning system (LPS), and clarify position-related difference in profile of CODs by using the approach.

The x- and y-coordinate data for each soccer player were measured with a local positioning system. Speed, acceleration, jerk, and direction of speed were derived from the coordinate data. Based on accelerations of above 2 m/s2, the onsets and ends of CODs derived from jerk were identified (COD duration). Changes of direction of speed (θCOD) were determined for the corresponding period. Six collegiate male soccer players performed CODs according to 13 set angles (0-180°; every 15°) so that differences between θCOD and set angle could be determined (Exp. 1). Relative frequency distributions of θCOD and number of CODs were determined in 79 collegiate and amateur male soccer players during 9 soccer matches (Exp. 2).

In Exp. 1, θCOD was positively related to set angle (r = 0.99). Each θCOD was smaller than the corresponding set angle, and the difference became greater with increasing COD angle. In Exp. 2, The number of CODs in a match was 183 ± 39 across all positions. There were no significant position-related differences in the number of CODs. The duration of a COD was 0.89 ± 0.49 s across all positions. The relative frequency distribution of θCOD revealed that the number of CODs at 0-15° and 105-135° tended to be higher than those at other angles during soccer matches. Further, θCOD was affected by the speed at the onset of COD during soccer matches (Exp. 2).

The current findings demonstrate that θCOD derived from direction of speed and jerk may be a new indicator for evaluating COD during soccer matches.