Faculty Advisor: Anantha Chandrakasan and Anette (Peko) Hosoi
Contact e-mail: firstname.lastname@example.org and email@example.com
The recent rise and success of analytic strategies in baseball management (popularized by Moneyball and other titles) prompts the question: Can similar strategies be applied to other sports? There are a number of reasons why one might expect baseball to be more amenable to these types of analysis than other sports, e.g. baseball positions are fairly "unspecialized" positions relative to other sports such as football. (In baseball, almost all players must take their turn at bat whereas football receivers are not expected to take their turn in the defensive line or play quarterback.) In addition, (perhaps most importantly), MLB has vast quantities of data available with approximately 2500 games played per season as opposed to approximately 500 per season in the NFL. For these reasons, football presents a new and exciting challenge in the field of sports analytics.
Alongside this growing interest in sports analytics there has been a concomitant rise in a new proving ground for analytic strategies: fantasy sports. Recent estimates suggest that 30-35 million people play fantasy football every fall. These players make use of a growing supply of online football statistics available at websites such as ESPN.com, Yahoo, NFL.com.
The aim of this SuperUROP project is to write an algorithm to draft and manage a fantasy football team. The project will involve aspects of machine learning to collect and synthesize available data, and the development of an app to make recommendations to fantasy football team managers.
Return to the SuperUROP site.