On the toxic discourse around "reaches" and the consensus board

In yet another touchpoint for analytics vs. old school scouting, the discussion around the use of consensus boards consistently misses the point.

If you’re reading this, you likely already know that over draft weekend, a discussion lit up social media around the idea of reaches and the consensus board. The Jacksonville Jaguars took a tight end 150+ spots ahead of where he was projected to go, and the wunderkind GM defiantly defended the selection. John Lynch faced questions around his pick of De’Zhaun Stribling at 33 when he was projected to come off the board some 50 picks later based on consensus.

There’s that dirty little world again. Consensus. A certain group of media and football people, including many in the league, reject the notion that such a thing exists, and at least in the NFL, they’re right. And the consensus board, such as they exist on the internet, consists of dorks like me with a keyboard who find film by hook or by crook, write our own reports, and grade players.

Already, this creates an inherent tension between the professionals and the amateurs. How could anyone deign to know the secret, nearly unknowable truths of collegiate prospects better than the teams that invest millions of dollars and thousands of man-hours in getting these picks right?

Well, because teams are bad at this.

Jason Fitzgerald at Over the Cap wrote about this phenomenon in 2023, and Pro Football Focus studied it several years earlier, based on their grades. What they found was that the league does not do a materially better job of identifying and ordering talent than the noobs. Fitzgerald’s analysis centered around second contracts and found the consensus board performs on par, and even better in some bands of the draft, compared to the NFL, outside of the first round, in terms of average money paid to players.

How can this possibly be?

And the answer can be found in the way someone like Lynch responded to questions about his draft misses after “reaching” relative to consensus. The only consensus he said he cares about is in his building.

This would not fly in any other part of life. The 49ers’ track record of drafting skill players on Day 2 has been horrendous under Lynch and Kyle Shanahan, while consistently picking players well above the range where consensus put them. If a hedge fund or a restaurant or a t-shirt company had this sort of discrepancy, they would not say, “Well, we trust our process.”

They would analyze that process, dig into the history of what they do, and try to identify trends. This lack of intellectual curiosity, mixed with the hubris of being stamped a “ball knower,” leads to poor outcomes.

Teams ignore the well-worn truths of draft history because they believe in their evaluations of players despite all evidence to the contrary. Teams make mistakes precisely because they are overconfident in their own ability to differentiate WR3 vs WR6, or QB1 vs. QB4.

There’s a reason teams like the Baltimore Ravens and Philadelphia Eagles consistently stack talent: they make normie picks that tend to follow consensus. And every year, when the draft grades come out — a stupid exercise in any case — they’re regularly among the highest-graded because they just did the basic thing and took the player most people agree is the best.