Roles in football cannot be still considered as mere positions. Thus different clustering techniques are used in this repo in order to explore the different football roles under a new perspective in which performance metrics, in addition to the position, contribute to the definition of a role.
The second part of the project is the development of different xG (Expected Goals) models in order to estimate the best one given the variables used in this approach. In addition to distance and angle, several others dummy variables were uses in order to give information about the action before the shot and the body part which the shot was took with.
Data used are taken from:
-
Pappalardo, L., Cintia, P., Rossi, A. et al. A public data set of spatio-temporal match events in soccer competitions. Sci Data 6, 236 (2019). https://doi.org/10.1038/s41597-019-0247-7 and are also available at: https://figshare.com/collections/Soccer_match_event_dataset/4415000/5. In particular, events data, which are large .json files, are not stored in this repository and can be downloaded from https://figshare.com/articles/dataset/Events/7770599?backTo=/collections/Soccer_match_event_dataset/4415000.