Fantasy Football Profitability: Using Machine Learning to Construct Optimal Teams and See Consistent Returns
Open Access
Author:
Baak, Andrew
Area of Honors:
Data Sciences
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
Francis C Wham, Jr., Thesis Supervisor John Yen, Thesis Honors Advisor
Keywords:
Machine Learning Data Science Sports Gambling Fantasy Sports
Abstract:
Rapid technological development in computing over the past three decades has led to a rise in applying machine learning applications and methods to solve a plethora of problems across a variety of industries. The creation and expansion of the internet has also led to a completely new way to instantly access, share, and download information. Additionally, the sports betting industry has revolutionized as a result of the internet, which has led to the development of popular online platforms to place bets on numerous different types of sporting events, including fantasy football. These three technological developments have led to a rise in the study of developing machine learning models from the vast quantities of data available online to try and outperform the field of contest entries. Machine learning models can be developed to predict the optimal teams to construct for weekly fantasy football contests to consistently beat the fields and see monetary returns, much in the same way that mathematical models are used in stock prediction in the financial industry. A well-constructed model can enable a user to consistently profit from playing daily fantasy football. While the model created for this project did not end up producing profitable results, it serves as a solid foundation for this type of work going forward, and with improvements can likely produce consistent profits in years to come.