Challenges

The digitization of human activities has recently extended to sports, making large amounts of data available to researchers. This data is useful to help refereeing, improve performance and monitoring of athletes, for instance for injury prevention, or provide strategic analysis, but also improve the fan experience.

While sports data analysis has in the past decades been more of a craft (practiced by coaches or sports journalists) than a science, this situation has recently changed. More and more mathematical and computational techniques have entered the field and are being used to support the work of movement and sport science experts, coaches and managers of professional teams, but also health actors and private individuals who want to improve their own performance or protect their health.

Most of the major teams in American sports and European soccer already hire data analysts; data analysis has become a flourishing industry and the CNRS has recently launched a GDR "Sport and physical activity " with the aim of facilitating interactions between researchers in different fields, and giving the French Olympic delegation an edge during the 2024 Olympic games.

Today, large amounts of data are regularly generated, not only by professional athletes, but also, with the advent of low-cost, high-quality sensors (e.g. in smartphones) by private individuals. The more the practice spreads and the more sensors are available and/or installed, the more complex such data become, turning high-dimensional, fusing different data types, and involving temporal and spatial aspects.

Despite the existence of these data sources, techniques to analyze them and the interest to do so, the three groups involved - data owners, analysts and experts/users - are not necessarily in contact or in collaboration.

This summer school is supported by the CNRS and organized by François Rioult and Albrecht Zimmermann.

Online user: 2 Privacy
Loading...