OSW

SIGNATURE WORK
CONFERENCE & EXHIBITION 2024

Statistical Method for Soccer Match Prediction Based on Historical Data of the English Premier League and Model Improvement

Name

Yuan Shan

Major

Data Science

Class

2024

About

Statistical Method for Soccer Match Prediction Based on Historical Data of the English Premier League and Model Improvement.

Signature Work Project Overview

This Signature Work aims to apply statistical methods to predict soccer matches. I selected the Bradley-Terry model as the initial framework and expanded it specifically for the soccer domain, incorporating home-field advantage and draw predictions to enhance prediction accuracy. A highlight of this work is the proposal of an original model that integrates club-level factors into the statistical model for match predictions, thereby enhancing the original model from a statistical theory perspective.
I used the English Premier League as the testing ground for my model. Through data analysis during the development of the statistical model for match prediction, I gained insights into the Premier League, as well as explored the impact of Covid-19 on the soccer world. This aspect provided domain-specific insights for our statistical model development and added practical significance and interest to this work.

Signature Work Presentation Video