The National Football League has announced that its annual Big Data Bowl competition will, for the first time, ask its thousands of data scientists to determine which pre-draft measurements actually correlate with success in the NFL. Which drills matter. Which numbers predict anything. It is the most honest announcement in modern sports, and it is also, in the specific, an admission delivered in a lab coat.

Consider the scene. Every February, in Indianapolis, the most physically gifted young men in America run a forty-yard dash in shorts, leap vertically off a mat, throw a football in a gym, and have their hands measured with a caliper by a man whose job title is closer to jeweler than coach. Teams then pay hundreds of millions of dollars based, in part, on those numbers. This has been the religion since 1987.

And now the religion has commissioned a study of itself.

This is what every institution does at the exact moment it stops believing in its own rituals but cannot yet afford to stop. The Catholic Church does not audit the saints. The NFL audits the shuttle run. One of these institutions is older, and the other one is the one with the stopwatch.

The truth the study will find, because the truth is already known, is that the combine measures the ability to perform the combine. A 4.3 forty in shorts in February predicts, with devastating accuracy, a 4.3 forty in pads in November, which is to say it predicts the past. What it cannot measure is the one thing scouts have always actually looked for: whether the man plays football when football is happening.

"We have spent forty years measuring the audition and calling it the movie. The data scientists have been hired to confirm the box office numbers of a film nobody has seen."

Some say the analytics will finally separate the athletes from the players. Others say the opposite, that the numbers will vindicate the old scouts with their stopwatches and their gut feelings. The truth, as always, lies somewhere in a third position I just invented: the data will prove that both groups were right about different players, which is the least profitable sentence in professional sports.

COMBINE METRICS: A PARTIAL AUDIT

  • 40-yard dash: Measures speed over a distance that occurs zero times in a football game.
  • Broad jump: Measures the ability to jump twice in the same direction, a play called "the two-hop," which does not exist.
  • Hand size: Measured with a caliper. No caliper has ever tackled anyone.
  • Bench press: 225 pounds, repeated. The heaviest object most linebackers will ever move in a game is another linebacker, who is actively resisting.
  • Shuttle run: Five yards, back. Football is eleven players trying to prevent exactly this.

None of which means the study is worthless. It means the study is late. The league that once drafted a quarterback for his “it factor” now wants a spreadsheet to tell it what the factor was, and the spreadsheet will say what every coach has said since the first game ended in a score: the factor was whether he played the game.

At press time, three teams had already used the study’s findings to draft a different man, which is what teams do with studies, and what every institution does with audits of its own religion: read them, nod, and keep the stopwatch.