European Football Leagues Project
For this project, I work with 2021-2022 season data for the top 5 European football leagues. The data was acquired from FBref and spans passing, shooting, defense, shot creating actions, and possession statistics.
The goal of this project was to model professional footballer’s play styles using their recorded, on-field actions, rather than just their positions. This project would aid with team building by being able to find replacement players who can fill the same role as the player they replace. Additionally, It can identify undervalued players who play valuable roles for their team.
I was able to identify different play styles using clustering and compare and contrast the different play styles by looking at the different cluster centroids. I also made a tool to find the most similar players given a player player name as the input. This tool used cosine similarity to return the top 10 most similar players for a given player. This was used to identify lesser-known players who compare favorable to established, elite players.
For this project I used python
.