A Qualitative Approach to Predicting the Success of NFL Draftees
Open Access
Author:
Insley, Drew
Area of Honors:
Statistics
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
Andrew John Wiesner, Thesis Supervisor Matthew D Beckman, Thesis Honors Advisor
Keywords:
NFL Sports Analytics Statistics Machine Learning
Abstract:
As NFL teams continue to employ analytics to evaluate prospects, it is important to consider a holistic approach. Quantitative data have been the driving force behind the decision to draft certain players, as teams look at metrics like a player’s 40-yard dash, vertical jump, and bench press. This report focuses on the use of qualitative data to evaluate players based on the strengths and weaknesses in their scouting reports for four major positions – Quarterback, Wide Receiver, Cornerback, and Defensive End. By employing a k nearest neighbors classifier, a classification tree, and a logistic regression, this study determines that the strengths in a player’s scouting report is the most important factor in a player’s chance of receiving a second contract with their draft team.
