How to Analyze Team Sport Data
Millions of people across the globe play team sport – it’s a popular way to socialize, stay healthy and have fun. It also teaches people how to work well with their teammates, a skill that can translate into countless situations in life. Team sports also teach kids about commitment, training and setting goals – helping them understand that hard work pays off and that there are generally no shortcuts. It also teaches them that it’s important to celebrate success and share the burden of loss.
Regardless of the sport, most team sports are characterized by high level of spatial interaction and temporal dynamics between players. The data available for analysis is usually large and can be acquired through different methods ranging from simple optical recognition or local positioning systems to more advanced tracking devices. Almost all of these data sources are based on descriptive (statistical) data.
Characterizing these data requires a more holistic approach to movement and context than simply considering x- and y-coordinates of entities. Adding information about the environment where the entities move makes it possible to explore more complex patterns and gain insights into how they are motivated, such as using collective movement models or the research on group movement in animals (see [52] for an example). In addition, adding contextual information enables analysts to make sense of the observed patterns by linking them with their meaning. This is particularly important when exploring complex phenomena in team sports where interactions between multiple players can cause unexpected results.