Broncos coach Sean Payton thinks AI is coming to play calling

Sean Payton, like the rest of us football enjoyers, was pretty impressed with Amazon Prime Vision’s Defensive Alerts feature, shown during the Denver Broncos last matchup with the Kansas City Chiefs.

The Broncos lost 19-8, Denver’s 16th straight against their I-70 rivals but in watching the game back Payton saw something that caught his eye. It was that artificial inelegance model trained to identify potential blitzes before the snap, it highlights players it believes have a high probability of pressuring with a red circle under them kinda like a video game.

“You guys watched a little of this Amazon, this AI predicting when pressures are coming, and inelegance-building scenarios to help play-callers,” Payton said Wednesday “All of that stuff is coming. I think part of it you take some of it in and it means a lot and the other ahh I’m good, you trust your gut.”

Payton was in a larger discussion at his press conference about advanced stats, when to go for two-point plays after a touchdown, and how the Brocnos implement some of the numbers. The Broncos first-year coach used the opportunity to share just how aware he is of how many took in the game two weeks ago and how it could impact actual NFL decisions in the future.

“In the box, on the sideline, no tech, no smartwatches, none of that. But as I saw it, I thought, this is all coming and I’ll probably be on a golf course somewhere when it does,” Payton said. “Coming but from the standpoint of when simulated decision-making relative to pressure, just begin to dabble and begin to think of where it can go, the people calling plays 20 years from now will be better. It continues to advance.”

Payton’s partnership with Drew Brees for the New Orleans Saints produced one of the best offenses in NFL history. The now Broncos head coach is credited with his schemes, play design and gutsy calls.

According to a great story breaking down the technology in The Athletic, this is how it works:

The model is trained not to identify the usual four down linemen that typically rush the passer. It’s trained to identify unique players who rush the passer on 60 percent or less of snaps. It’s being fed tracking data from Next Gen Stats, which is derived from RFID chips in every player’s shoulder pads. The data includes the players’ acceleration, their orientation and where they are facing. From all that data, the machine starts to understand familiar movement patterns from the defense as a whole, which helps it predict which player is going to blitz.

The story states that the goal is to get viewers to see the game like a quarterback. On the play below, the Broncos ran a Cover 0 blitz and the model highlighted the obvious and not-so-clear blitzes on the play. Before the ball was snapped but after the audience saw what was happening, Patrick Mahomes identified the same thing as the AI did but it was too late and the Broncos were still able to bring down the KC QB.

ALERT! ALERT! 🚨

The new Defensive Alert feature available exclusively on @TNFPrimeVision w/@NextGenStats identified Patrick Surtain II as a defender to watch off the edge before the snap.

📺: https://t.co/1OMRlplD2C pic.twitter.com/gzPe21ilNH

— Next Gen Stats (@NextGenStats) October 13, 2023

Here’s another play where the AI showed off what was going to happen before it did.

Nice look at @NFLonPrime's Prime Vision with @NextGenStats new Defensive Alerts feature. 🔥🏈👀

AI tracks defensive movements prior to the snap and highlights players likely to blitz the QB. #PrimeVision pic.twitter.com/lUm9lcgwUQ

— PrimeVideoSportsPR (@PrimeSportsPR) October 13, 2023

***