TESTING THE EFFICIENT MARKET HYPOTHESIS IN FANTASY FOOTBALL AUCTION LEAGUES

Open Access
Author:
Armington, Keith Daniel
Area of Honors:
Finance
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
  • James Alan Miles, Thesis Supervisor
  • Chris Muscarella, Faculty Reader
  • James Alan Miles, Honors Advisor
Keywords:
  • Fantasy Football
  • Predictive Models
Abstract:
When people think of the Efficient Market Hypothesis, they’re likely to think of capital markets and the pricing of securities. What they probably won’t think of, however, is the market for National Football League players within fantasy football. Piggybacking on the success of the National Football League and the proliferation of the internet, fantasy football – in which participants attempt to predict which football players will perform best on Sundays – has created a whole new market. Such markets for football players are referred to as “auction-style fantasy football” and assigns cash values to players. Better performing players demand higher prices, and participants usually engage in bidding wars to buy the best players. The market for fantasy football is huge, and now with Las Vegas getting involved, payoffs for successful participants can be enormous. My thesis will focus on these auction-style markets and determine if they are “efficient.” If not, obviously, the inefficiencies could be exploited for personal gain. To determine if fantasy football auction-style markets are efficient, I focused on top quarterbacks from the past three years and attempted to produce a predicative model based largely on historical production and prices. Then, I took the best models and applied them to the 2010 football season to determine their accuracy in predicting the price of players. Finally, I compared the statistical differences in production of the quarterbacks with the difference between my model and the actual prices to see if any quarterbacks were incorrectly valued relative to their production. The result would show me if the fantasy football markets were indeed efficient or not, and if not, which players would be incorrectly valued. Hopefully, someone could use my models to determine a more accurate price for players and play the market accordingly.