Who's on First: Simulating the Canadian Football League Regular Season

• Activity Number: 617
• Type: Contributed
• Date/Time: Thursday, August 7, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Sports
• Author(s): Keith Willoughby*+
• Companies: University of Saskatchewan
• Keywords: Football ; Simulation ; Spreadsheet modeling

Abstract

The Canadian Football League (CFL) features an 18-game regular season played between late June and early November. Teams are divided into two geographical divisions. Generally, the top three teams in each division qualify for post-season competition. The second and third place squads compete in a divisional semi-final, with the winner traveling to the first place team's stadium for the divisional final. The winners of each divisional final compete in the league's championship game. The eventual victor receives the CFL's championship trophy (the Grey Cup).

Given that the first place teams receive a bye into the divisional final, there is considerable incentive for teams to finish atop their respective divisions.

Using a spreadsheet model, we develop an approach to simulate the league's regular season. On a weekly basis, the model considers each team's current record and remaining opponents in the season. A team's probability of victory in any game is determined through a logistic regression model. This model incorporates a "power index" representing the proficiency of each team.