New

Why Clutch Qbs Don’t Exist (and Why It Looks Like They Do)

Nov 27, 2025; Arlington, Texas, USA; Kansas City Chiefs quarterback Patrick Mahomes (15) and Dallas Cowboys quarterback Dak Prescott (4) greet each other after the game at AT&T Stadium. Mandatory Credit: Kevin Jairaj-Imagn Images | Kevin Jairaj-Imagn Images

I have a dry-cleaner who loves giving me health advice. Every week, she tells something new to try, a water fast, linseed oil, all sorts of homeopathic stuff, tai chi walking, cupping therapy, you get the drift. I always nod politely and change our small talk to a different topic. The other day, she looked me in the eye and says,

“You’re not doing any of the stuff I’m telling you, right?“

“No.”

“So, you’re one of those ‘1+1=2’ guys, yeah?“

I cannot deny this. But I now understand that there must be ‘1+1=3’ people or even ‘1+1=4’ folks for whom logic, math, or scientific data plays no role in their lives. I suspect these may be the people who believe in the idea of clutch performance in sports.

I used to reflexively deny the existence of clutch. In fact, clutch, as it is commonly used, runs counter to any and all data-driven arguments. Most statisticians politely turn around and break wind in your direction when you mention clutch to them. I, for one, don’t buy the idea that a player has some kind of athletic superpower that he can simply switch on when needed.

The concept of a ‘clutch gene,’ the idea that certain players possess a repeatable, inherent ability to elevate their baseline skills strictly when the pressure mounts, is largely a statistical mirage. However, the narrative of clutchness is a powerful engine driven by media economics and cultural values.

There is such a thing a “clutch performance”, except it’s not what most people think it is.

What they tell you is a ‘clutch gene’ on television is almost always the result of tangible, repeatable advantages in physical preparation, specialized skill sets, or schematic manipulation. There are many situations that can make a player appear to be ‘clutch’ when in fact he isn’t.

Conditioning: Better physical conditioning may allow you to perform at your peak longer than your opponent. This may give you an advantage in the fourth quarter and give the impression of clutch performance, when in fact you’re simply more consistent than your opponent. A quarterback who suddenly finds clean pockets at the end of the game isn’t magically ‘wanting it more’ – he may be benefiting from a fatigued pass rush that can no longer generate pressure.

Aggression Metrics: Teams trailing late throw deep downfield far more often out of necessity, raising the standard yards per attempt metric, which is heavily weighted in the official NFL passer rating formula.

Specific end-game skills: Specific end-game skills can make a player excel in late-game situations without making him inherently ‘clutch.’ Like a baseball closer or a quarterback who thrives in a two-minute offense, some athletes are simply better at the specific demands of endgame play. In the NFL, early-game success often depends on scripted execution, while the fourth quarter rewards improvisation and adaptability. Players who excel in that chaos are often praised for ‘finding that extra gear,’ when in reality they just possess skills that are especially suited to those situations.

Schematic Superiority (Math Over Magic): For years, the Eagles have been unstoppable on 4th-and-1. Commentators praise Jalen Hurts’ ‘grit’ and ‘will to win.’ In reality, the tush push is a masterclass in physics and leverage. The Eagles use an elite offensive line, low pad level, and legal extra pushers to create an almost mathematical certainty of gaining one yard. It is a mechanical cheat code, not some heroic effort. But that’s just one very specific play, schematic superiority extends much further:

The Prevent Defense. Defenses often shift to soft prevent coverages late in games when protecting a lead. This surrenders quick, high-percentage underneath completions that naturally inflate a quarterback’s completion percentage and rating.
Exploiting the rule book. The NFL rule book changes dramatically in the final two minutes of a half (clock stoppages on out-of-bounds, specific replay review rules, defensive pass interference thresholds). Many so-called clutch performances are actually examples of exploiting the rule book rather than extraordinary execution. Elite teams and quarterbacks understand late-game rules and intentionally create situations, such as underthrown deep passes that draw defensive pass interference, to gain an advantage, even if the outcome is later framed as a clutch drive.
No-Huddle: Moving to a high-tempo, no-huddle offense limits complex defensive substitutions. Elite processors can decipher defensive looks much faster under these constraints.

But clutch extends far beyond the measurable and quantifiable. Here are four more examples:

Reputational Clutch: ESPN says a player is clutch. Therefore he is. Every time he wins it’s because he’s clutch; when he loses, it was bad luck. The Choker is the exact opposite. Every time he wins it’s because he got lucky, when he loses it’s because he’s a choker. In 2015, ESPN published a ranking of the “most clutch QBs.” The top four QBs at the time: Cam Newton, Marcus Mariota, Carson Palmer, and Teddy Bridgewater. A decade later, I can only think, “what a preposterous ranking.”

Anti-Clutch: This is a guy who simply plays badly for the first three quarters and then finds a way to elevate his game to halfway normal levels. ESPN loves this kind of guy. They used to call him Tim Tebow.

The absence of choking: Players who maintain their performance despite increased pressure will appear to be clutch, especially if their opponent cannot maintain his performance under pressure. Yet all they’re doing is being consistent. No supernatural clutchness to be found anywhere; nothing for Sam and Dean to investigate.

The Variance Illusion: Because human brains seek patterns, we remember the one spectacular red zone interception and completely forget the five times that same player threw a touchdown in the red zone.

The ‘clutch quarterback’ is not a formally defined term (because clutch doesn’t exist, of course), but in common usage it refers to a quarterback who makes important positive plays for his team at key moments and in the tightest situations. Typically, these would be:

Fourth quarter comebacks – This one is most commonly associated with clutch, and while not perfect, should be included in any analysis.
Improvement in the fourth quarter over the first three quarters – As mentioned above, the fourth quarter is tied to the term clutch, and the level of play of a quarterback in the fourth compared to the first three should be looked at.
Performance on third down – How a player performs on a third down is key, and is one of the most pressure-filled situations in the game.
Playoff numbers compared to regular season – This one is important because each game in the playoffs is high pressure, and improvement under these circumstances is a good sign.

All of this of course assumes that there are quarterbacks who are able to elevate their level of play in specific situations. Why they would choose to lower their performance at other times is beyond me of course, but let’s walk through each point nevertheless.

Fourth quarter comebacks (4QC)

When most fans think of clutch, it’s almost always tied to the fourth quarter. Quarterbacks who play well and win a lot of games in the fourth are known as clutch, quarterbacks who don’t are often labeled as chokers.

And typically we get these clutch QBs shoved in our face in forms like this chart from Foxsports from last year:

This type of ‘ranking,’ without any context, may give you clicks and eyeballs, but raises more questions than it provides answers: How many games did the QB play, and how many opportunities for a comeback did he have in the first place? We are here to answer those questions.

The next chart highlights the top seven active QBs with at least 35 starts and with

the highest number for fourth quarter comebacks (the updated Foxsports ranking)
the number of 4QCs as a percentage of total games started
the number of 4QCs as a percentage of fourth quarter opportunities (games where the QB’s team was behind by -1 to -7 points at some point in the fourth quarter or overtime).

body .sbnu-legacy-content-table td, body .sbnu-legacy-content-table th, body .sbnu-legacy-content-table { border: 1px solid #000 !important; border-collapse: collapse !important; } body .sbnu-legacy-content-table td, body .sbnu-legacy-content-table th { padding: 0px 6px !important; }

# of 4QC			4QC in % of Games played				4QC in % of 4QC Opportunities
Player	Games	4QC	Player	Games	4QC	4QC%	Player	4QC Opportunities	4QC	4QC%
Matthew Stafford	239	39	Jordan Love	48	8	17%	Patrick Mahomes	50	20	40%
Kirk Cousins	167	26	Matthew Stafford	239	39	16%	Jordan Love	21	8	38%
Aaron Rodgers	257	25	Bryce Young	44	7	16%	Jalen Hurts	27	10	37%
Andy Dalton	169	25	Patrick Mahomes	126	20	16%	Dak Prescott	52	19	37%
Joe Flacco	201	23	Justin Herbert	95	15	16%	Matthew Stafford	112	39	35%
Patrick Mahomes	126	20	Kirk Cousins	167	26	16%	Jared Goff	53	18	34%
Dak Prescott	139	19	Andy Dalton	169	25	15%	Josh Allen	51	17	33%

The names in each cluster are already quite instructive: A ranking of ‘clutch’ QBs that features every Joe, Kirk, and Andy can’t mean much. And in fact, in our sample of 33 QBs with more than 35 starts, by far the biggest determinant of how many 4QCs a QB has is the number of games played. Here’s a graphic visualizing that:

By and large, 4QCs are a function of games played. An R² of 0.85 means 85% of the fourth quarter comebacks are directly explained by the number of games played. As such, the absolute number of 4QCs (and the related 4QCs as a % of total games) is utterly useless.

Much more interesting is the right side of the table above, which shows 4QCs as a percentage of 4QC opportunities. For one thing, it feels like it has the ‘right’ names at the top, and secondly, it feels intuitively correct: you can only stage a fourth quarter comeback when you are actually behind in the fourth quarter, right?

Here’s the full list of all currently active QBs, with a separate section for QBs with fewer than two seasons worth of starts simply because their numbers tend to fluctuate a lot early on and tend to settle down once they play more games.

4th Quarter Comebacks
Player	Games	4QC Opportunities	4QC	4QC in % 4Q Opportunities
Patrick Mahomes	126	50	20	40%
Jordan Love	48	21	8	38%
Jalen Hurts	82	27	10	37%
Dak Prescott	139	52	19	37%
Matthew Stafford	239	112	39	35%
Jared Goff	151	53	18	34%
Josh Allen	127	51	17	33%
Tua Tagovailoa	76	27	9	33%
Kirk Cousins	167	79	26	33%
Andy Dalton	169	76	25	33%
Justin Herbert	95	49	15	31%
Carson Wentz	99	43	13	30%
Lamar Jackson	107	37	11	30%
Marcus Mariota	82	41	12	29%
Bryce Young	44	24	7	29%
Brock Purdy	45	14	4	29%
Deshaun Watson	72	32	9	28%
Baker Mayfield	120	50	14	28%
Geno Smith	98	47	13	28%
Aaron Rodgers	257	91	25	27%
Kyler Murray	87	42	11	26%
Joe Flacco	201	89	23	26%
Daniel Jones	82	39	10	26%
Trevor Lawrence	77	33	8	24%
Sam Darnold	90	34	8	24%
Teddy Bridgewater	65	22	5	23%
Jameis Winston	89	50	10	20%
Joe Burrow	77	30	6	20%
Mitchell Trubisky	57	26	5	19%
Mac Jones	57	26	4	15%
Jacoby Brissett	65	34	5	15%
C.J. Stroud	46	16	2	13%
Justin Fields	53	25	3	12%
Players with < 35 starts
Jayden Daniels	24	7	4	57%
Bo Nix	34	15	8	53%
Caleb Williams	34	19	8	42%
Tyler Shough	9	5	2	40%
Shedeur Sanders	7	4	1	25%
J.J. McCarthy	10	3	1	33%
Cam Ward	17	5	1	20%
Drake Maye	29	9	1	11%
Jaxson Dart	12	3	0	0%
Michael Penix	12	6	0	0%

Fourth quarter improvement

In football analytics, Expected Points Added (EPA) is the holy grail metric. It measures how much value a quarterback adds to his team’s scoring probability on every single play, which is why we’ll look at EPA per play in this section. On average, an elite NFL quarterback produces around +0.20 to +0.30 EPA per play, keep this in mind as we walk through the numbers.

RBSDM.com is an excellent resource for just this type of analysis, and they even provide ready-to-publish charts to look at. Here’s the first chart that covers the 1st-3rd quarter only from 2021-2025. As you can see at the top of the chart, I’ve set a few other filters as well (min 1,000 plays, active QBs only, regular season only) that give us 30 quarterbacks.

If you look at the top right quadrant, you’ll see some of the best QBs in the league today, and you may or may not be surprised that outside of Matthew Stafford, all six of the best 4QC quarterbacks are in the top right corner and above the line: Patrick Mahomes, Jalen Hurts, Jordan Love, Dak Prescott, Jared Goff, and Josh Allen. But will their performance hold up in the fourth quarter?

Here’s the fourth quarter chart with one important additional filter: The Win Probability Threshold is set at 20%, which excludes plays where one team has less than a 20% or more than 80% chance to win, which filters out “garbage time” – lopsided, late-game situations where play-calling and Expected Points Added (EPA) become heavily skewed. The 20% threshold is primarily used to evaluate quarterback or play-caller execution in competitive, neutral-leverage situations. All other filters can be seen at the top of the chart.

Matthew Staffords rockets up the chart, Hurts, Prescott, Mahomes and Love are also above or close to the 0.20 threshold, Herbert stays below 0.20 and Allen drop below the 0.20 line. C.J. Stroud and Joe Burrow jump the shark if you will and also show up above the 0.20 threshold. All in all, that gives us the following ranking of QBs and how much they improved their EPA/Play in the fourth quarter/overtime.

Quarters 1-3 vs 4th Quarter / OT
QB	Quarters 1-3	4th Quarter / OT	Change
Matthew Stafford	0.13	0.35	0.22
Jayden Daniels	0.10	0.25	0.16
Jalen Hurts	0.14	0.29	0.15
CJ Stroud	0.10	0.24	0.15
Joe Burrow	0.17	0.25	0.08
Geno Smith	0.04	0.11	0.07
Bryce Young	-0.08	-0.01	0.07
Mac Jones	0.00	0.06	0.05
Dak Prescott	0.18	0.22	0.04
Patrick Mahomes	0.20	0.23	0.02
Kirk Cousins	0.09	0.10	0.01
Sam Darnold	0.10	0.11	0.01
Caleb Williams	0.00	0.01	0.01
Daniel Jones	0.07	0.07	0.00
Aaron Rodgers	0.11	0.11	0.00
Jaxson Dart	0.15	0.15	-0.01
Justin Herbert	0.13	0.12	-0.01
Jordan Love	0.20	0.19	-0.01
Kyler Murray	0.10	0.08	-0.02
Tua Tagovailoa	0.16	0.13	-0.03
Bo Nix	0.08	0.04	-0.04
Jacoby Brissett	0.02	-0.02	-0.04
Justin Fields	0.02	-0.02	-0.05
Lamar Jackson	0.17	0.09	-0.08
Baker Mayfield	0.09	0.00	-0.09
Josh Allen	0.20	0.11	-0.10
Brock Purdy	0.27	0.14	-0.13
Jared Goff	0.16	0.01	-0.14
Trevor Lawrence	0.11	-0.08	-0.19

This chart looks simple enough, but isn’t as easy to read as it looks. If you’re a fan of the QBs at the top of the table, you can rejoice. “Yay, my QB is clutch in the fourth quarter!!!” But what’s the point of improving in the fourth quarter when, like Bryce Young, you’re the worst of 29 QBs in quarters 1-3 and ‘improve’ to fourth-worst in the fourth quarter?

You’ll notice that the cells with EPA>0.20 are marked in green, and you’ll also notice that only one QB has two green cells to his name: Patrick Mahomes once again proves that he’s a planet unto himself. Dak Prescott, Jordan Love, and Joe Burrow narrowly miss out on having two ‘greens’ as well. Stafford, Jayden Daniels, Jalen Hurts, and C.J. Stroud show strong fourth quarter improvements, and it would be interesting to understand what holds them back in the first three quarters, just as the magnitude of the drop off for Josh Allen, Brock Purdy, and Jared Goff would warrant an extra look.

In any case, Dak Prescott ranks fifth overall in EPA in the first three quarters and seventh in the fourth quarter. Not bad for a QB many reflexively call un-clutch.

Third Down Performance

Teams that consistently move the sticks on third down tend be the teams that win consistently. It follows that quarterbacks that move the sticks consistently on third down are very good quarterbacks; quarterbacks that fail at this most basic of football tasks are usually not very good.

The simplest way to measure a QB’s ability to move the sticks is to add up his passing and rushing first downs and divide them by his number of pass attempts and runs on third down. Which is exactly what you see in the next table.

3rd Down Conversion Percentage
Player	Pass Att	1D	Rush Att	1D	3rd-down %
Brock Purdy	353	172	43	29	50.8%
Patrick Mahomes	1,066	512	126	77	49.4%
Joe Burrow	676	320	68	38	48.1%
Josh Allen	965	415	257	172	48.0%
Dak Prescott	1,121	521	130	71	47.3%
Jalen Hurts	574	235	217	132	46.4%
Justin Herbert	896	390	116	70	45.5%
Jameis Winston	838	373	91	47	45.2%
Aaron Rodgers	2,239	975	226	133	44.9%
Lamar Jackson	714	281	246	141	44.0%
Carson Wentz	892	374	101	60	43.7%
Jordan Love	413	175	35	19	43.3%
Top 34 QB Average (42.9%)
Jared Goff	1,302	559	65	28	42.9%
Tua Tagovailoa	616	269	55	19	42.9%
Baker Mayfield	1024	408	145	88	42.4%
Marcus Mariota	684	267	104	67	42.4%
Matthew Stafford	2,297	958	126	64	42.2%
Mitchell Trubisky	519	212	78	39	42.0%
Justin Fields	377	131	140	85	41.8%
C.J. Stroud	424	175	39	18	41.7%
Kirk Cousins	1,581	645	80	47	41.7%
Deshaun Watson	586	234	113	57	41.6%
Daniel Jones	688	274	115	58	41.3%
Trevor Lawrence	641	252	93	50	41.1%
Mac Jones	476	183	68	36	40.3%
Teddy Bridgewater	568	219	85	43	40.1%
Jacoby Brissett	619	235	90	49	40.1%
Geno Smith	838	326	99	49	40.0%
Kyler Murray	707	270	115	57	39.8%
2025 NFL Average (39.5%)
Joe Flacco	2,011	770	138	77	39.4%
Andy Dalton	1,540	593	142	67	39.2%
Sam Darnold	784	293	82	46	39.1%
Bryce Young	370	118	44	26	34.8%
Players with <35 starts
Drake Maye	209	89	30	17	44.4%
Jaxson Dart	95	36	27	17	43.4%
Jayden Daniels	178	66	40	26	42.2%
Bo Nix	341	130	53	33	41.4%
Tyler Shough	84	33	5	3	40.4%
2025 NFL Average (39.5%)
Michael Penix	96	37	6	2	38.2%
Caleb Williams	291	106	48	21	37.5%
Shedeur Sanders	59	21	8	4	37.3%
Cam Ward	143	46	11	6	33.8%
J.J. McCarthy	63	18	10	4	30.1%

The first thing to understand here is that while we are using 3D% as a QB stat, there’s a strong team component in here is well. Without proper pass protection, run blocking on designed run, and without a good receiving corps, even the best QBs will struggle to move the sticks.

And if you’re looking for ‘clutchness’ you’ll find it at the very top of this list, and you’ll be happy to see that Dak Prescott is right in that Planet Mahomes tier.

Playoffs versus regular season

We’ll use the EPA stats again for this final exercise. The chart below shows the regular season EPA/Play for 28 active QBs.

And we again have the usual suspects at the top of the list. Planet Mahomes all by himself at the top and a second tier of EPA QBS with Aaron Rodgers, Josh Allen, Lamar Jackson, Joe Burrow, Dak Prescott, and Jalen Hurts.

Next up is the playoff EPA, again with the Win Probability Threshold set at 20% to exclude garbage time.

Most Cowboys fans will look at this chart with some level of concern. How can a QB like Dak Prescott (and Brock Purdy, Jared Goff, or Lamar Jackson) who was close to the top in all the ‘clutchness’ metrics we looked at previously, suddenly experience such a steep EPA drop in the playoffs? Did they suddenly lose all their abilities that made them stand out previously? Or is it some team or coaching issue that leads to this drop? Because Mr. Mahomes doesn’t seem to care whether it’s the regular or the post-season.

One (partial) reason can be found in the amount of snaps each QB played from behind in the playoffs. Dak Prescott, for example, played 72% of his playoff snaps from behind, Lamar Jackson is at 63%, Brock Purdy at 58% and Jared Goff at 51%. The chart below, featuring 11 starters with at least 150 playoff snaps, shows a solid R² of 0.53, meaning half of the drop in performance is due to the percentage of snaps played from behind.

Now, why these guys are playing from behind so much is an entirely different discussion.

The other point we’ve largely eliminated in this analysis is garbage time. And at least for Dak Prescott, looking at that garbage time is quite instructive. Here’s his EPA/Play by Win Probability Threshold.

Dak Prescott Career Playoff EPA/Play
	All Plays	5-95% Win Probability	10-90% Win Probability	20-80% Win Probability
Plays	331	250	203	149
EPA/Play	0.16	0.09	0.04	0.02

Prescott’s total playoff number looks okay with a 0.16 EPA/Play, but it decreases the more ‘garbage time’ you take away. At the end, when you only look at plays with a win probability of 20-80%, you’re taking away almost half of Prescott’s total playoff plays.

When you allow the 2016 Packers to jump out to a 21-3 lead early in the second quarter, or allow the 2018 Rams to take a 20-7 lead into the half, or trail the 2021 49ers by 23-7 halfway though the third quarter, or have the 2023 Packers go into the half with a 27-7 lead, you’re going to play from behind a lot. For Prescott, those four games are the rough equivalent of two full games of ‘garbage time.’ So maybe we’re overshooting here. Garbage time used to be something like ‘down by two scores in the fourth quarter with not enough time to score twice’ (or some variation of that) but here we’re taking about whole second halves as garbage time.

I’ve indulged the garbage-time argument here because it’s one that’s frequently brought up in discussions about Dak Prescott. But if it’s a valid argument for Prescott, it must be a valid argument for everyone else as well, right?

Would someone like to tell Tom Brady that most of his performance in Super Bowl LI against Atlanta ‘doesn’t count’ because he was playing in garbage time from the middle of the second quarter almost to the end of the fourth?

One-way arguments that don’t rebound, that cannot be universally applied, are bad arguments.

In any case, the Cowboys, and the Cowboys offense, and Dak Prescott need to play better and more consistently in big games and the playoffs. That’s something you can recognize even if you’re a big Prescott supporter, just as detractors can recognize that Prescott has mostly played at a really high level over the last ten years.

I’ll leave the last words to Prescott’s former head coach, Jason Garrett:

I’m smiling because I love the guy. I think he gives them a chance and he’s played so well. I do think he’s an underappreciated player. There’s so much scrutiny about the Cowboys and his position that he’s in, and with that comes a lot of emotions.
If you just pull back and watch how he’s played for 10 years now, he’s just played at a really high level. And I think he gives them a chance to have success.
The one thing that I think he would tell you is, as well as they’ve played on offense – the numbers were really gaudy – I think he would say, “Hey, we gotta play better in these big games that we need to win.” They need to play consistently well offensively throughout the year, I think they have a chance to do that.

Back to Listing

credit: