The Chiefs once again lost to the Bills in the regular season, something that they’ve grown accustomed to despite having no issue dismissing Buffalo from the playoffs in recent years. Even as they were sitting a score or two behind for much of the game, the specter of Mahomes Magic loomed, but it didn’t quite manifest in this contest.

In a world where Bo Nix seems to post a big fourth quarter (after a questionable first three) every week, I thought it’d be helpful to talk about quarterback clutch performance and put Patrick Mahomes into some perspective with others in the league.

First off, how might we operationalize clutch? I thought of it two ways, in both cases using our player value metric Total Points to draw the comparison. 

The simple one is performance late in games, comparing fourth quarter production to first-three-quarters production. 

The more complex one is comparing high-leverage situations to low-leverage ones, building in an understanding that a fourth quarter touchdown when you’re down by 28 points isn’t very clutch. For this we borrow methodology from the baseball world, where we can label a situation’s “leverage index” by comparing how much a team’s win probability can swing in a given situation compared to the average. 

Leverage can get really high in some situations (like 10x a typical play), but values above 2 are pretty rare in football, so samples get small. So for this purpose I am considering high leverage to be 1.5x and low leverage to be 0.5x an average situation.

To put that in slightly more concrete terms, here are three 1st-and-10 situations around midfield that have different leverage indices by our calculations.

Average leverage: 1st quarter, 4 minutes left, 7 point lead

High leverage (1.5x average): 4th quarter, 4 minutes left, tie game

Low leverage (0.5x average): 2nd quarter, 4 minutes left, 13 point lead

I took each quarterback’s relative success in fourth quarters and high-leverage situations (compared to the rest of their plays) and simply averaged them together to come up with a “clutch composite”, if you will. And if we look at current quarterbacks with at least 1,500 evaluated plays in their career so far (roughly two-plus full seasons), it’s not surprising who we find at the top.

Most clutch current quarterbacks, minimum 1,500 career dropbacks / carries

* The average of the gap between fourth-quarter and other-quarter performance and high-leverage and all-other-leverage performance

Would you look at that: Mahomes sits above the rest of his peers in terms of how well he rises to the situation. And he’s done so across a much greater sample size. Josh Allen has had similar overall production, but it’s been more balanced between the low-leverage and high-leverage spots.

And lest we think that early-career Mahomes is coloring the picture, his most clutch seasons by this measure are the three most recent.

An aside on other players

In looking at this research, I figured I’d poke around to see other interesting trends. 

Bo Nix has had a lot of “no no no yes” games this year, but his performance from a Total Points perspective hasn’t been appreciably better in big-time situations. The Broncos have averaged fourth-quarter scoring that’s almost a touchdown better than the first three quarters, but Nix hasn’t shown that kind of productivity jump on his own.

Justin Fields has been one the least clutch quarterbacks this year. He’s basically become less clutch every year he’s been in the league. In general, it’s fair to assume that a player’s clutch performance has an outsized impact on the vibes surrounding his season, and that feels particularly relevant when a player gets benched. 

Drake Maye and Sam Darnold are obvious counterexamples to that last point, as players who have had really strong seasons so far. Maye has been excellent in low-to-medium leverage and quite poor in high leverage. Darnold has been elite in the first three quarters but the worst quarterback in the league in the fourth quarter.

Of the players who have been the most clutch by this composite measure this year, Michael Penix Jr. is the only one who has been worse in high-leverage spots (and therefore very good in fourth quarters). He’s run up the score or turned it on in garbage time but has not risen to the big moments.