Abstract
The NFL Scouting Combine provides prospective professional American football players the opportunity to demonstrate physical performance abilities and position-specific skills in a series of standardized assessments, though the current understanding of the influence of these tests on draft status and position remains unclear. PURPOSE: This study evaluates the NFL Combine's physical tests for explaining variance in performance within positions and predicting relative draft position (RDP) and draft status (DS). METHODS: Players (n=1234) were categorized into four position groups (PG) using k-means clustering: quarterbacks (QB), skill, linemen, and mid-size. Predictors included results of the six NFL Combine tests, height, weight, power, and momentum. Only players who completed all combine assessments were included (n=1234). Principal Component (PC) analysis with oblimin rotation and Horn’s parallel analysis reduced data dimensionality for each PG. Linear mixed models and binary logistic regression were used to predict the impact of PCs on RDP and DS for each position. RESULTS: Two to three meaningful PCs (adjusted eigenvector ≥ 1) were identified for each PG with the first PC universally representing momentum metrics, explaining 33-45% of intra-PG variance. PCs significantly associated with RDP were identified for all PGs but defensive backs and running backs, while PCs significantly associated with DS were found for all PGs except QBs and power running backs (p0.05), and 10/26 PCs showed no significant relationship with DS (p>0.05), suggesting that physical performance alone does not account for most variability in RDP once underlying physical thresholds to be drafted are met. CONCLUSION: The NFL Combine physical performance tests inconsistently predict DS and RDP but should not be overlooked during draft preparation. Prospects should prioritize position-specific skills and game performance over optimizing these test results. Future research should explore the influence of these components on NFL career success and longevity.
Recommended Citation
Doyle, Benjamin P.; Stanelle, Sean; Riechman, Steven E.; and Mann, J Bryan
(2025)
"AI-Driven Insights from the NFL Scouting Combine: Principal Component Analysis, K-Means Clustering, and Regression for Predicting Draft Position and Status,"
International Journal of Exercise Science: Conference Proceedings: Vol. 2:
Iss.
17, Article 106.
Available at:
https://digitalcommons.wku.edu/ijesab/vol2/iss17/106
