Open Access
Miller, John Z
Area of Honors:
Actuarial Science
Bachelor of Science
Document Type:
Thesis Supervisors:
  • Andrew John Wiesner, Thesis Supervisor
  • Ron Gebhardtsbauer, Honors Advisor
  • NFL
  • height
  • QBR
  • binomial logistic regression
  • passer rating
  • completion percentage
  • NFL draft
  • early entry
  • special eligibility
  • scout
  • analytics
  • starting experience
  • prediction
  • regression model
With the advent of predictive analytics in everything from politics to professional sports, the most lucrative sport in North America demands a taste of what statistical analysis has to offer for the future of its brightest stars. When a young student athlete has the opportunity to decide between finishing their collegiate academic career or potentially becoming an instant-millionaire, how do they decide upon their future? If the student decides to stay in school for another year, are they enriching their future, or jeopardizing their chance at national stardom? How is an organization affected by the early departure of its best player? This paper focuses on college quarterbacks and their decisions to enter the NFL draft early. We discuss the important statistical variables in considering a quarterback’s probability of both entering the NFL draft early and being drafted by an NFL team. Through these variables, we demonstrate the power of a statistical model in projecting the decisions of these students with factors such as height, a player’s high school rating, their collegiate starting experience, and their college performance summarized in their Total Quarterback Rating (QBR). While these variables have been found to be extremely significant in predicting a player’s draft likelihood, we also assess the potential bias of NFL evaluators overweighting characteristics that are expected to predict success. These statistical models provide a lens through which NFL scouts can re-evaluate how quarterbacks are selected, as well as a lens for college quarterbacks to evaluate their future NFL potential in their decision to declare early.