Quantifiably clutch: Patrick Mahomes and the Chiefs are an exception to every rule

An examination of what Mahomes and the Chiefs do when trailing shows why they're so dangerous in close games.

I’ll get right to it. We’ve stumbled upon yet another “the NFL has never seen anything like Patrick Mahomes” stat. And this one might be the clincher for any holdouts. It’s so absurd that it got double-checked, because it seems like it can’t possibly be true (how’s THAT for a lede?).

Before I get started, a huge thanks to Ethan Douglas and Lyndon Reid for hunting down and charting numbers for this project. You should follow them both on Twitter, they’re doing really fun stuff with numbers that matter. All right, let’s talk Mahomes, and how to gauge football greatness.

If you’ve followed the NFL, particularly the analytics field, for any period of time, you’ve probably heard that the results of close games are mostly random, or “noise.” There are so many random events that can swing a close game (a fumble recovery, bad penalty, drops, etc) that winning close games is more affected by those events than by any one thing the team does.

That’s why it’s been seen as concerning for the Chiefs that on Sunday they became the first team in league history to win six straight games by one score or less. The basic premise is that winning close games isn’t sustainable because it’s random. Let’s put a pin and that and come back to it.

Related to the idea of close games being random is the idea that being “clutch” isn’t necessarily a real thing. The idea is that great players are just great more often than non-great players, and therefore appear to step up in big moments more often (I’m not saying I necessarily agree with this idea, but that’s the thought process), while average or bad players are more likely to be bad or average in big moments.

Let me ask you a question related to that subject… have you ever felt like the moment the Chiefs are trailing in a game, Patrick Mahomes starts to play at an even higher level than his absurd standard?

Take the Saints game for example. Early in the third quarter New Orleans scored to take a 15-14 lead over Kansas City. The Chiefs’ offense had mostly struggled throughout the day, but on the ensuing drive Mahomes completed 5/5 passes for 51 yards, including the go-ahead touchdown for Kansas City. That drive, of course, included absurd plays like this one.

In short, the moment the Chiefs were trailing Mahomes unlocked another level of play and brought them back to the right side of the scoreboard. And be honest… were you surprised? Or did it feel somewhat expected?

The idea that Mahomes suddenly “ramped up” to something even better than he usually is goes against what we know about the clutch gene, randomness, and basically every rule we have about football. But I started thinking anecdotally about the close games the Chiefs have played in, and even their losses, and my memory told me that nearly every time the Chiefs started to trail or needed a score, Mahomes was able to provide it at least once. Even in games they didn’t ultimately prevail.

But that just has to be my imagination, right? There’s no way that Mahomes actually pushes himself to even higher levels when trailing, is there?

Well… that’s where the numbers come into play. And fair warning… this is going to look made up.

Remember EPA per play? “Expected Points Added,” which basically means whether a play got you closer to scoring (for example, going from the 50-yard-line on 3rd down to the 20-yard-line and 1st down makes you much more likely to score), and by how much. Obviously, the bigger the number the better.

The chart above looks at EPA per play for dropbacks (the vertical) and per rush (the horizontal, which includes quarterback scrambles and so is very related to Mahomes being one of the best scrambling QB’s in the NFL). Obviously up is better for the passing game, and to the right is better for the run game and QB scrambles.

Find the Chiefs on this chart. It might take a moment.

The reason it takes a moment is because the Chiefs are hanging out on an island all by themselves at the uppermost top right. They’re so far above everyone else that the chart is actually larger because of their existence.

Are you a subscriber yet? If not, you’re missing a lot of stuff like this that’s aimed at giving you a better understanding of the Patrick Mahomes, the Chiefs, and football in general. It’s 12 bucks a year forever (60% off) to subscribe now using the link below, and that will be your price forever. Do it. You know you want to. All right, back to Mahomes being historically great.

When Lyndon discovered that stat, he and I initially thought that this has to be an aberration due to small sample size. After all, the Chiefs haven’t been trailing that often in 2020, so just a few big plays will make a massive difference. So Lyndon expanded the scope to go back to the beginning of 2018, when Mahomes first took over as a starter.

And, well…

A few teams manage to close the “rushing” gap (which makes sense if Mahomes’ scrambling is a big reason why the Chiefs are there, as it’s a smaller sampling of plays), but look at that gap in EPA per Dropback. It’s even higher than the 2020 chart.

What does this mean? Well, in a nutshell, it means that the Chiefs’ passing offense, since 2018, has been a little over 1.5x better than the next best offense in the league when trailing. They’re farther ahead of the next-best team than that team is from the 16th-best team. It’s utterly unbelievable.

And yet, when one watches Mahomes, it’s believable. Take the Chiefs’ comeback win over the Raiders for example. The Chiefs got the ball back on their own 25-yard-line trailing by three points with 1:43 remaining and only one timeout left. But did the result feel in doubt at that point? Mahomes proceeded to complete 6/7 passes for all 75 yards and the game-winning touchdown, and it looked frighteningly easy while he did it.

And this isn’t a new thing. As the numbers show, Mahomes has been doing this since he became the starter. Even in games the Chiefs ultimately lost, he consistently played so well when trailing that they had an opportunity to win.

One example is the no-longer-painful-because-they’re-champs 2018 AFC Championship game. The Chiefs fell behind by a field goal with a mere 32 seconds remaining and only one timeout left. It’s a situation that’s almost impossible by normal standards. Two throws later, the Chiefs were in field goal range. And those throws were, much like the situation, impossible by normal standards.

You might think I’m done talking about how different Mahomes is, but I’m not. Because we haven’t finished yet with mind-melting numbers and showing why the Chiefs are an exception to the rules that typically govern football.

Ethan Douglas @EthanCDouglas

Patrick Mahomes has never lost an NFL game by more than a possession. He's one of a kind, and that's not hyperbole. No matter how many big names and future hall of famers I cram on this chart, he rises above. He is the comeback King and no one else is even royalty.

This chart is another way of showing how Mahomes isn’t like other quarterbacks, even elite ones. You see, other quarterbacks have their EPA per dropback (the vertical) improve as their likelihood of winning increases (the horizontal). This makes sense, as defenses are forced to press and quarterbacks are able to pick and choose their spots more. It’s worth noting that on the right side of the graph (high win probability) Mahomes’ EPA per play is right there with the best quarterbacks in the league over the last few decades.

But notice that Mahomes’ line goes UP as it goes to the left, as opposed to the rest of the quarterbacks on the chart? Weird, right?

What that means is that as the Chiefs are LESS likely to win (win probability getting lower), Mahomes EPA per play gets higher. Much higher. To the point that it’s more than twice as high as anyone else on a list that includes Tom Brady, Peyton Manning, Aaron Rodgers, Russ Wilson, and every other quarterback someone might mention as being as good as Mahomes when trailing. Statistically, they aren’t, and it’s not even remotely close.

The short version? Mahomes is better than Brady, Manning, and everyone else when his team’s win probability starts to drop. Demonstrably, quantifiably better. And it’s a big reason why he’s never lost a game by more than a single score, and why they can win one-score games at a higher rate than normal.

And make no mistake, it’s hard to be better as a quarterback when your team is trailing. Opponents know you have to throw the ball, so they’re able to pin their ears back and rush the passer while utilizing coverages less worried about the run. The thing is, Mahomes is special in those situations too. We’ve got one last ridiculous stat to go.

Ethan Douglas @EthanCDouglas

Related chart that explains my issue with "balance", at least as it relates to the Chiefs. The x-axis here is now "pass probability" - how likely a team is to pass given play-level context (score, down, distance, time, etc). Mahomes has been best when teams KNOW he is passing.

This chart is a little more complicated. The “pass probability” horizontal axis is measuring how likely a team is to throw based on down, distance, and score. So the farther to the right you go, the more likely a team is to throw (3rd and long, trailing by a lot, etc). Once again, we’ve got a list that includes the best quarterbacks of the last few decades. Notice how some of them get worse the more likely they are to pass, while some of them get only slightly better? That’s because, again, it’s hard to pass the ball on obvious passing downs when the defense knows it’s coming.

Then there’s Mahomes. The more the team has to throw (and the more the defense knows they have to do so), the better he gets by a landslide. As is the situation with his ability to outperform everyone else when trailing, he’s so much better than everyone else that it’s practically a different chart because of him.

Look, if someone doesn’t believe Mahomes is the best quarterback in the league at this point, they might never. But here’s a simple list of the quantifiable facts we currently have:

• Mahomes and the Chiefs are demonstrably, significantly better on offense when trailing than anyone else has been since he took over. It’s not even close.

• Mahomes and the Chiefs are demonstrably, significantly better when trailing than Brady, Manning, or any of the other best quarterbacks of the last 20 years. It’s not even close.

• Mahomes and the Chiefs are demonstrably, significantly better when in obvious passing downs than Brady, Manning, or any of the other best quarterbacks of the last 20 years. And, yet again, it’s not even close.

The normal rules of football don’t necessarily apply to the Chiefs. This doesn’t mean they’re unbeatable. Everyone is beatable in the NFL. It just means that in order to do so, teams have to overcome a quarterback who is playing at a level the league hasn’t seen before.

It’s not your imagination. It’s not a dream. Patrick Mahomes is breaking the NFL, and is doing things no one else has done. And we get to watch it every single week.