Date of Graduation

5-2024

Document Type

Thesis

Degree Name

Bachelor of Science in Industrial Engineering

Degree Level

Undergraduate

Department

Industrial Engineering

Advisor/Mentor

Cassady, Richard

Committee Member/Reader

Rainwater, Chase

Abstract

Each NFL, NBA, and MLB season consists of a regular season, in which teams play a set number of scheduled games and a playoff, in which qualifying teams compete for a championship. At the conclusion of each season, teams are ranked based on their performance throughout the season. This study aims to investigate the ability of each league's season format to accurately rank teams using Monte Carlo simulation. Matches between two teams are simulated by using the team’s assigned strength ranks to calculate a winning probability for each team. The winning probabilities are simulated with different skill values, dictating how much impact the difference in the team's strength ranks has on the outcome of games. The outcome is determined randomly using these probabilities. For each league, season format features—including the scheduling and regular season, tournament qualification and seeding, and the playoff format—are modeled. The study concludes that the NFL's season format is least effective at ranking teams. The MLB's season format is more slightly more effective at ranking teams when the value of skill is low, but slightly less effective than the NBA's season format when skill is more of a factor. The study also finds that the inclusion of a playoff results in less accurate rankings compared to the regular season winning percentages alone, across all leagues and skill values.

Keywords

Simulation, Monte Carlo, tournament design, ranking efficacy, java

Download

Share

COinS