Predicting Arsenal Results with Elo Scores

How Premier League Elo scores can be used to predict the outcome of fixtures, using classification models in Sklearn

Introduction

I’ve previously written about Elo scores, and how they can be used in machine learning classification models. A team’s Elo score is represented by a number which either increases or decreases depending on the outcome of matches between other Elo-ranked teams. Following every game, the winning team takes points from the losing one. The difference between the ratings of the winner and loser determines the total number of points gained or lost after a game; the number of points both lost and obtained are weighted according to the quality of the team and their most recent form. This weighting is predicated on the Elo score the team had prior to the match. You can read more about Elo and its application in chess here. I have previously used Elo scores in machine learning classification models for match prediction in cricket ODI matches. The problem is that the Indian Premier League is over now (well done, Mumbai) and I wanted to try something new. I wanted to apply Elo modelling to my other passion: Arsenal.