Predicting Ticket Holder No-Shows: Examining Differences Between Reported and Actual Attendance at College Football Games

Abstract

Reported attendance for most sport events is based on tickets disseminated, not actual number of spectators who physically enter the venue. Yet nearly all live sport event demand studies are based on reported attendance rather than the actual attendance. The current study examines multiple measures of home game attendance for NCAA Division I college football programs as reported from both game box scores and post-event scanned ticket audits provided to The Wall Street Journal. Regression models are utilized to examine factors that have a statistically significant relationship with three different measures of attendance: (a) reported attendance, (b) actual attendance, and (c) total number of ticket holder no-shows. Several independent variables, included demographic factors, measures of game attractiveness, and residual preferences, demonstrated such relationships with each measure of attendance when examining ticket usage data from 595 game dates during the 2017 season.