A passion project involving Sports Analytics, the following repository contains my attempt at using statistical and mathematical modelling to simulate matches between clubs, and simulate an entire season based on the match simulation model.
The project utilizes data from the following sources :
- FIFA 20 Ratings : https://www.kaggle.com/stefanoleone992/fifa-20-complete-player-dataset
- Squad Possession Statistics for Big 5 Leagues : https://fbref.com/en/comps/Big5/2019-2020/possession/squads/2019-2020-Big-5-European-Leagues-Stats
- Squad Shot Creation : https://fbref.com/en/comps/Big5/2019-2020/gca/squads/2019-2020-Big-5-European-Leagues-Stats
- Squad Shots and Shots on Target : https://fbref.com/en/comps/Big5/2019-2020/shooting/squads/2019-2020-Big-5-European-Leagues-Stats
- Squad Advanced Goalkeeping Statistics : https://fbref.com/en/comps/Big5/2019-2020/keepersadv/squads/2019-2020-Big-5-European-Leagues-Stats
- League Table : https://fbref.com/en/comps/9/Premier-League-Stats
The project aims to predict match results both on the basis of static probabilistic modelling, and dynamic random simulation. Building on these models the project aims to succesfully sim the English Premier League for the season 2020-21 Do keep in mind that due to the dynamic element no two function runs or sims will be the same. Enjoy and do share my work if it interested you!