EA FC Machine Learning Analysis

EA FC is the largest football game available on the market today. With data of over 16,000 players, this data can serve several purposes for models. In this case, creating a classification model for the players will be the goal. Through CEF, Additive Models, Tree based models, ID Generalization, OOD Generalization, and Covariate Shift Models, this project aimed to explore the different ways that models can predict a players position in EA FC. The goal was to use the most important features in a player, such as pace, shooting, passing, dribbling, defending and physical, to attempt to predict their specific position in the midfield. These midfield positions vary from Outside Midfielder, Central Midfielder, Central Attacking Midfielder, and Central Defending Midfielder. As a fan of the game myself, I wanted to see which of these variables had the largest impact in defining a players position, to better understand tactics when playing the game.