Statistical Analysis of Premier League Match Statistics Using a Regression Analysis in R

This thesis analyzes the correlation between a team’s statistics and the success of their performances, and develops a predictive model that can be used to forecast final season results for that team. Data from the 2017-2018 Premier League season is to be gathered and broken down within R to highlight what factors and variables are largely contributing to the success or downfall of a team. A multiple linear regression model is then used to take out any factors that are believed to have little or no significance in match results. The predictions about the 17-18 season results based on the model proved to be satisfactory. The model saw an accuracy percentage very near to perfect and allowed for a correct prediction of table standings. This allowed for the next step in the experiment to be conducted which was to analyze and compare the findings with recent seasons effected by Covid-19. The breakdown of a season not effected and a season fully effected allows an opportunity to see what has changed in the game we see today