An analysis of football data to improve corner kick strategies.
This analysis uses the public and free StatsBomb football dataset to make recommendations to a Set-Piece coach. These recommendations refer to both defensive and offensive strategies for corner kicks.
An example of this can be found in the following illustration:
The data is provided as JSON files exported from the StatsBomb Data API, in the following structure:
- Competition and seasons stored in
competitions.json. - Matches for each competition and season, stored in
matches. Each folder within is named for a competition ID, each file is named for a season ID within that competition. - Events and lineups for each match, stored in
eventsandlineupsrespectively. Each file is named for a match ID. - StatsBomb 360 data for selected matches, stored in
three-sixty. Each file is named for a match ID.
The data set contains information about the competitions, season, individual matches with their results, up to the line-up of both teams.
Furthermore, the data set is continuously maintained in the area of a 360 data record and contains detailed information on which events in a game, such as the tackling of a player, through to the position data of a player and what activity he was involved in during an event.
The analysis is in the file: Football_analysis.ipynb
Der Gleiche Inhalt mit Bereitstellung der interaktiven Visualisierung findet sich in der Football_analysis.html.
The conda environment with which this analysis was performed can be generated with:
create --name <env> --file requirements.txtAll necessary dependencies are listed in requirements.txt.