The Combine Lie

Every February, the NFL Combine dominates sports media. Forty-yard dash times are analyzed. Bench press reps are counted. Prospects rise and fall based on numbers collected in a controlled gym environment.

The Combine Lie asks: which of these numbers actually predict NFL performance?

The Setup

I pulled Combine data for every draft pick from 2000-2020 and matched it against career performance metrics (games started, Pro Bowl selections, WAR proxies where available) by position.

The goal was to find: for each position, which Combine metric has the highest predictive correlation with career success?

What the Data Shows

For most skill positions, the answer is surprising. Forty-yard dash time — the most-covered metric — has low predictive validity for wide receivers and running backs. Three-cone drill, which measures change-of-direction and rarely gets prime coverage, has significantly stronger correlation.

For offensive linemen, the best predictor of career performance has nothing to do with athleticism: it's arm length, combined with the height-weight-speed profile that signals how a prospect will develop physically.

For quarterbacks, no Combine metric has meaningful predictive validity. The Combine was designed to evaluate athletes, not decision-makers, and quarterback performance is driven almost entirely by cognitive and processing traits that no physical drill measures.

The Media Gap

The metrics that get coverage are the ones that are dramatic and easy to visualize — a 4.3 forty produces shareable content in a way that three-cone drill results don't.

The result is that draft coverage optimizes for engagement rather than prediction, which is why the consensus board is reliably wrong in predictable ways.