Your chest tightens. Your heart races. Your palms sweat. You're sitting down, but only barely; you're on the edge of your seat and the balls of your feet. You're holding your hands out in front of you, balled up tightly into fists.
As you watch with eyes wide open, your team's quarterback flicks his wrist. As the ball sails through the air, you can't breathe. Your heart stops. Time stops. The whole game—the whole season—has come down to this moment.
This is why we watch sports.
Whether you're alone at home on your couch or in the stadium with 60,000 of your closest friends, nobody makes a sound. Everything hangs in the balance, everyone's muscles go taut with adrenaline, and the air itself crackles with potential energy.
When the ball comes down safely in the receiver's hands and two feet land in the painted turf?
You jump through the ceiling, both fists to the sky, screaming and whooping and hollering out every watt of that electrifying moment. In the stadium, strangers become friends and friends become family as everyone screams, cheers, laughs, hugs and high-fives.
When the ball sails past the receiver's outstretched arms and into the clutches of the deep-lying safety?
A long, slow exhalation—the release of pressure to the atmosphere—60,000 silent trudges up the stairs to the exits, and many more televisions flick off.
Whether that moment ends in explosion or deflation will always be a part of that quarterback's story. Whether he slammed down the plunger and detonated the stadium with joy or snipped the wires and sent everyone skulking back home will forever etch the following in the minds of everyone who watched: "clutch" or "choke."
Memory: The Most Convincing Liar
Over the years, whether or not any given player is "clutch" has been the subject of countless articles, blog posts, call-in-radio-show arguments and TV debates. Over the recent years, the din has grown even louder as we argue about what clutchness is and whether or not it even exists.
These moments, what statisticians call "high-leverage situations," stick with us.
We remember athletes as clutch performers when they're at their bests in the biggest moments or as choke artists when they can't get it done.
What we don't remember are the missed field goals in the first quarter. We don't remember the perfectly placed slant that converted a crucial third down on the touchdown drive before halftime. We don't remember the uncalled pass interference on the play before which would have put the home team within easy striking distance.
Our brains filter out all the little details that build up to that clutch moment, the details which made the situation high-leverage. Over the course of a career, these few massive moments and the impressions they leave on our minds can outweigh dozens of pressure-filled moments that do (or don't) reinforce our conceptions.
Some define "clutch" as "playoff performance." All that does, though, is build a story off small sample sizes. Players with only a few playoff appearances get tagged as "good" or "bad" playoff quarterbacks—but the more playoff games they play in, the more it irons out to their career norms.
New England Patriots quarterback Tom Brady, an unheralded sixth-round draft pick, was thrust into the starting role due to an injury. He went on to win his first nine playoff games, securing three Super Bowl championships and a reputation as one of the most clutch players ever.
He's gone 8-7 since, going one-and-done twice and losing two Super Bowls in the process. Did he suddenly become un-clutch?
Not at all: He's just playing to his ability. In 24 playoff games, Brady's postseason numbers (listed first) are very close to his regular-season averages—in fact, they're slightly worse:
Brady neither has any playoff magic nor drops off from his regular-season performance. His lower playoff touchdown-to-interception ratio is likely due to playing only playoff-caliber defenses.
Lies, Damned Lies and Statistics
For years, baseball statisticians have been trying to prove (or disprove) the idea of clutch hitting, and basketball statisticians have been doing the same with clutch shooting.
They always come up with the same answer: There is such a thing as measurable performance in big moments, but there aren't players who consistently play better when the stakes are high, year after year.
Does it exist in football? We have a way to find out.
Brian Burke of Advanced NFL Stats generates a statistic called "Win Probability Added," which measures the change in probability a team will win from one play to the next. When a player makes a play that puts his team in a much better position to win, he adds lots of win probability.
A five-yard run on 1st-and-10 from a team's own 20-yard line in the first quarter doesn't add much WP. A five-yard run on 4th-and-4 in the red zone in the fourth quarter when the team's down by six adds a whole heck of a lot. Over the course of a season, WPA tells us who's doing the most to help his team win.
Burke has another stat, Expected Points Added, which is similar but doesn't take into account time or score. So that five yard run on 4th-and-4 in the red zone generates more Expected Points than the five-yard run on 1st-and-10, but EP doesn't care how much time is left or who's winning.
So if EPA measures a player's production and WPA measures how much a player tips the scales between winning and losing, we can figure out how "clutch" someone is. If the production comes when it counts, the WPA will be higher. If the production comes in the first quarter and in garbage time, the EPA will be.
Burke did exactly that in 2010, generating "expected WPA" given all quarterbacks' EPA production, then comparing their expected WPA to their actual WPA:
The red line there represents the difference between how much WP Matt Ryan's production should have generated and the WP he actually generated: over two full wins. Ryan, then, was the "clutchest" quarterback of 2010.
But was Ryan the clutchest quarterback of 2011 or 2012?
This January, Burke revisited the idea, comparing the seasons of all NFL quarterbacks with more than 200 attempts per season since 2000. He measured correlation for year-to-year EPA, WPA and that "clutchy" differential factor.
He found that quarterbacks' EPAs stayed the most consistent from season to season, with WPA varying a little more. That "clutch" difference between expected WPA and actual WPA, though, was all over the place—there's no year-to-year persistency; it's essentially random.
So statistically, football "clutchness" is the same as that of every other sport: There are good players and great players and not-so-good players, but none of them are consistently better or worse when the game is on the line.
So how do we define clutch? As our brains do, with the biggest moments imprinting the boldest memories. Just because no player consistently plays better in the playoffs than any other time or steps his game way up in the fourth quarter doesn't make those moments any less electric or any more forgettable.
Clutch plays are real, and great players often make them.