An Empirical Analysis of Demand in Collegiate Football

Open Access
Pistner, Michelle Anne
Area of Honors:
Interdisciplinary in Economics and Statistics
Bachelor of Science
Document Type:
Thesis Supervisors:
  • Peter William Newberry, Thesis Supervisor
  • Russell Paul Chuderewicz, Honors Advisor
  • Murali Haran, Honors Advisor
  • Statistics
  • football
  • sports
The Pennsylvania State University is home to a strong athletic program marked by fan loyalty. However, as with many sports programs, a natural problem arises as to how to predict purchasing habits of individual buyers. This paper analyzes attendance and individual demand for the university’s football program. Linear regressions were used to model attendance on a game by game basis. This analysis concluded that opponent strength and past performance were the strongest predictors of attendance. Demand was further analyzed on the individual level. Purchasing habits for the year prior was found to be the strongest predictor of an individual’s purchases for the next year. However, the usefulness of this information is limited. In order to determine new buyers and returning buyers, logistic regression models and decision trees were constructed. Their results yield a simple framework relying on past behavior and account information to predict between different buying patterns.