Team sport is a diverse and interesting mix of athleticism, creativity, competition, and collaboration. Some of the most popular team sports include basketball, football, hockey, cricket, tennis, and baseball.
Tracking technology can be used to profile team sport athletes during training and competition, enabling practitioners to sync vision with external load data and examine tactical or collective behaviour. However, it is important that tracking systems and derived metrics are appropriate to the particular sport or playing position.
Some sports, such as basketball and netball, feature intermittent high-intensity action, characterised by rapid changes in direction or short distances between locomotor demands (e.g., standing, running, sprinting), and contact actions such as post-ups or screens [118]. These characteristics may be less relevant to a basketball player than a football player, given the smaller court dimensions and limited number of players.
Metrics for analysis in these sports must be considered with a critical thinking process, and a healthy dose of scepticism and awareness of appropriate theoretical frameworks. Speed thresholds have been determined using data mining techniques in team sports such as soccer, rugby and football [125], with Gaussian curves or k-means clustering methods used to categorise instantaneous velocities into speed zones.
In addition, moving averages and time series segmentation have been used to analyse a single athlete’s movement during a match. The speed of the approach and the ease with which data can be accessed allows time series analysis to detect how an athlete’s physical output changes as a function of time during matches. The resulting trace can be visualised by encoding thousands of data points into time-series segments.