The Myth of Clutch

A-Rod

If I took 100 pennies and I threw them up in the air, about half of them would land heads and the other half tails, right? Now, if I looked around closely, I’d probably find some heads grouped together in a cluster. What does that mean? Does that mean anything?—A Civil Action

Statistics are great. They help us find the answers to important questions. Need to know if smoking causes lung cancer? Look at the data. Wonder if height is correlated with material success? There’s probably a study you can find. Think Albert Pujols is a better hitter than Mickey Mantle? Look it up. Statistics aren’t the final answer to any of these questions, but they certainly help.

The problem with statistics is that, like most great things—the automobile, plutonium, superpowers—they can be very dangerous in the wrong hands. One need only to look at the myth of baseball’s “clutch players” to see how statistics can be misinterpreted.

One week ago, Alex Rodriguez and Vladimir Guerrero were first-ballot Hall of Famers. CC Sabathia was one of the best pitchers in the game. But all three had reputations as guys who couldn’t come through in the playoffs. They were not “clutch players.” Clutch players are guys like Derek Jeter, David Ortiz, and Cole Hamels.

Except here’s the thing: Rodriguez and Guerrero each had clutch ninth-inning hits in their division series. Sabathia gave up one run in seven innings vs. the Twins. Meanwhile, Ortiz went 1-for-12 with no walks and three strikeouts, and Hamels gave up four runs in five innings at home (Derek Jeter had a great series, but that’s because Derek Jeter is fucking awesome).

So what happened? Did A-Rod, Vlad and CC all suddenly learn how to be clutch players? Did Ortiz and Hamels just forget? Neither. The truth is this: THERE IS NO SUCH THING AS A CLUTCH PLAYER.

First, we ought to establish what is meant by “clutch.” Clutch, when applied to situations, certainly exists. We all know that calling an instance clutch means it’s important, critical, or crucial. Calling a player “clutch,” however, implies that this player is particularly prone to coming through in these types of situations.

In actual fact, though, no such players exist. Many studies have been done (most at Baseball Prospectus, which you need a paid subscription to see, sorry), and they pretty much all confirm this fact. As James Click says, “Clutch hits exist, clutch hitters do not.”

But what about common sense? What about looking at the simple fact that Alex Rodriguez hadn’t had a good playoff series since before the 2004 ALCS?

Well, this is why statistics are dangerous. We look at clusters of pennies and think they have meaning.

In Game 1 of the NYY-MIN series, Rodriguez had an RBI single in the fifth inning. The estimable Chip Caray announced that it was his “first postseason hit with a runner in scoring position since 2004.” When someone says something like that, the reaction is “Oh my God! That was five years ago! Barack Obama wasn’t even a U.S. Senator then. The O.C. was still popular. Kanye West had just released his first solo record. Nobody knew who the hell Miley Cyrus was.” And it seems like it was a long time ago, because it was.

But in all that time Rodriguez had a grand total of 18 at-bats with runners in scoring position. During the regular season, he gets that many in a good week. Major League hitters are evaluated over the course of a season, usually with over 500 at bats. If you throw 500 pennies in the air, about half will land heads and half will land tails. You might look closely and find clusters of 18 heads in a row, but what does that mean? Does that mean anything?*

*Actually, it probably means the coin is fixed, since this would be incredibly unlikely for a simple coin toss. But the odds on a coin toss are 50-50, and no hitter ever hits .500 in the majors.

A sample of 18 is incredibly small when it comes to professional hitters. In 2009, for example, Mark Teixeira had a terrible April, in which he hit .200 with and OPS of .738. That was over the course 70 at-bats, or almost four times what Alex Rodriguez’s “unclutch” reputation was based off of, and yet Teixeira still finished the year as one of the AL’s leading MVP candidates.

Every player has stretches of 18 bad at-bats, even in the playoffs,* even if they are supposedly “clutch.” David Ortiz, for example, the clutchiest of clutch hitters, hit .095 in 21 at-bats during his first playoff series in Boston. Even Jeter (even Jeter!) went 3 for 17 in the 2007 ALDS, 4 for 27 in the 2001 World Series, and 5 for 25 in the 1998 ALCS. 

*Sometimes particularly in the playoffs, since the pitchers are typically better. Guerrero’s poor postseason reputation, for example, is largely based on going 1 for 20 in the 2005 ALCS, but that was a series in which the Chicago White Sox had FOUR CONSECUTIVE complete games. Guerrero wasn’t the only one who couldn’t hit those pitchers.

Now, you may be looking at those last numbers and thinking, “Wow, Derek Jeter sucks. How on Earth did he get his great reputation?” (or, if you’re Tim, you were thinking that already). Well, Derek Jeter has played in an absurd number of playoff games. The 2009 ALCS will be his 27th postseason series in 13 seasons. He has 505 playoff at-bats, or about as many as a typical season. How do his numbers compare, you ask? Well, his postseason stats are .311/.380/.477 and his regular season numbers are .317/.388/.459. Wow, those numbers are shockingly similar. Almost suspiciously similar. It almost leads one to believe that a player, given a lot of opportunities, will perform more or less the same in the postseason as he does in the regular season.

This statement is borne out again and again. Ortiz in the playoffs: .283/.388/.520. Ortiz in the regular season: .282/.377/.545. Few other players have enough playoff appearances to effectively compare, but you can also look at regular season “clutch stats.” A-Rod’s numbers in these situations (close and late, two outs and runners in scoring position, etc.) are more or less the same as his career numbers. These facts are also borne out by BP’s research. The aforementioned Click says that “Good hitters are good clutch hitters; bad hitters are bad clutch hitters.”

But people don’t want to look at these numbers; fans look at the clusters and convince themselves that they MUST mean something.

“Clutch situations can’t be defined statistically,” people say. “Not every time up with two outs and runners in scoring position is equal. Not every playoff at-bat is equal. These stats don’t tell the whole story.”

There is a grain of truth to this objection: There is no textbook or objective definition of a “clutch” situation. Statistics can only reveal some of the truth. A clutch situation is one that is crucial or especially important, but sometimes this importance is hard to quantify. A home run to stop a losing streak is clutch. A big hit that makes up for a teammate’s error can be clutch. A walk-off hit against a major rival can be clutch. It’s also fair to say that some clutch situations are more clutch than others (“clutchness” differs from “uniqueness” in that respect). But evaluating the relevant scales and magnitudes is not clear-cut and differs for every fan and participant.

And this is a crucial point: It’s hard to always know what situations the players themselves consider clutch. Clutchness is often explained by highlighting players’ ability to succeed under pressure, but athletes feel pressure for all sorts of ambiguous and personal reasons, and everyone reacts differently. Some supposedly “clutch” players “raise their game” under pressure (Derek Jeter, David Ortiz) and others appear to “keep their cool” (Manny Ramirez, Mariano Rivera). Some guys feel pressure approaching a personal statistical milestone, others feel it in big games, others against intimidating opponents.

It’s not that pressure doesn’t exist, it’s that it is impossible to quantify all the reasons someone might feel added pressure (like, say, if you promised a kid in the hospital two home runs). Whenever I rail against the subject of “clutchness,” people accuse me of ignoring the “human element” of sports. In fact, I think I’m recognizing just how complicated it is. Clutch situations occur all the time in sports, that’s what makes them so much fun. But it’s not as simple as some guys have clutch and others don’t; different players respond to pressure in different ways and in different situations. It’s next to impossible to tell what’s going on in a player’s mind, and how he will react to it. So with the game on the line, I’ll take the player with the better batting average over the one with the better reputation.

Advertisements

14 responses to this post.

  1. Posted by Tim on October 13, 2009 at 9:30 PM

    While I generally agree with your point that there are more clutch hits than clutch hitters and that “clutchness”–your word, not mine–is to a large extent exaggerated by the media (and to a greater one by Chip Caray), I still hold that there are some “clutch” players who tend to come through in these situations more than others.

    Mariano Rivera has appeared in 79 postseason games–more than he has in any regular season–and has an ERA of 0.74, or roughly half of his best single-season ERA. This is despite the fact that he is both facing better competition and pitching longer in most of his appearances. Mariano Rivera is clutch.

    Pitchers such as Dave Stewart and Orel Hershiser have career postseason ERAs almost a run lower than their regular-season ones. And Barry Bonds, despite his 8-for-17 with 4 HR World Series in 2002, hits 53 points worse in the postseason. Stewart and Hershiser are clutch; Bonds is not.

    Furthermore, it’s disingenuous of you to write off Derek Jeter’s postseason numbers as “suspiciously similar” to his regular-season ones, and then say it’s not weird for other players to struggle in the postseason because the pitching is better. I’d say that Jeter’s ability to hit as well as he does against that better pitching makes him more clutch than, say, Alex Rodriguez.

    It’s also disingenuous of you to compare A-Rod’s 18 ABs to Teixeira’s 70. Rodriguez’s struggles were non-consecutive; that is, it can’t be attributed to a single slump. He went 3-for-29 in back-to-bak postseason years in 2005-2006. In Sept. of 2005, he hit .330 with eight home runs. In Sept. of 2006, he hit .358 with eight home runs. In the playoffs, then, he went 3-for-29.

    Let me put it to you hypothetically: Let’s say A-Rod is a Muslim, and after hitting .330 with eight home runs in a month, he decides to fast for Ramadan. During Ramadan, he goes 2-for-15. The next year, A-Rod hits .358 with eight home runs for a month before his fast. During his fast, he then goes 1-for-14. Would you think that his fasting had something to do with his struggles? Or would you simply say, “Hmm…coincidence”?

    On an entirely different note, you do only consider “clutchness” in the context of baseball. I would argue that it plays a much bigger role in individual sports, such as golf (it’s HUGE in golf) and tennis (particularly women’s tennis). It’s less visible in less individualized team sports, like basketball and football (Eli Manning the one exception).

    Reply

  2. Posted by John S on October 14, 2009 at 12:49 AM

    Yeah, I stuck within in the confines of baseball, and even within that sport mainly within hitting, because it’s the most quantifiable sport and the most quantifiable aspect of that sport. Even so, I think, my points generally hold across sports: It’s not simply a matter of some guys being clutch and others not, it’s mainly that different people react differently at different times. Look at guys like Peyton Manning, Phil Mickelson, and Kobe Bryant: These guys are all considered incapable of coming through in the clutch until the moment they do. It’s the same phenomenon at work. That is, fans extrapolate from a small sample (even though no golfer wins every tournament, no QB wins every game, no basketball player wins every title) until that sample gets big enough and evens out.

    As for your potentially offensive Ramadan example: You’re being a little disingenuous. Obviously fasting during daylight hours is going to affect performance in a much more direct and predictable way that something as psychologically nebulous and elusive as “clutch” situations. It’s true that it isn’t one single slump in which A-Rod is performing, but it doesn’t change the fact that 18 is an unfairly small sample size (The truth is that A-Rod had one really bad series in 2006 against Detroit. In 2005, his BA was low, but he walked six times and was hit twice; he wasn’t seeing many good pitches. In 2007, he went 4 for 15 with a HR–not the stuff of legends, but not totally unforgivable either. But since he never got a decent number of at-bats, he never really had a groove until this year). What I meant by the Teixeira comparison wasn’t that it’s merely one slump; I was pointing out that judging a player by one small cluster of at-bats is unfair and misleading.

    I also think you’re misinterpreting my “superior competition” argument. The playoffs are conducive to facing more dominant pitchers, particularly in a single series with a deep staff or a great ace. The Guerrerro example is the best: He was facing four pitchers basically at the height of their games. But what kind of drop-off should we expect over the course of a few hundred playoff at-bats? Most guys I looked at with 200+ at-bats have numbers that are pretty similar. I think over a lot of playoff games, you’re going to face your fair share of bad pitchers, particularly in the Wild Card era, so this makes sense. In a single series, though, one pitcher or staff can more or less shut even the best hitters down. This is, granted, just a hypothesis, but I don’t think it’s a contradiction or “disingenuous,” as you imply, for me to make both arguments.

    As for your Hershiser/Stewart examples, they seem to indicate clutchness, until you look closer. Their playoff numbers are, understandably, taken from seasons in which they were at their most dominant. Almost a third of Hershiser’s postseason innings, for example, come from his exceptional 1988 season. Similarly, almost all of Stewart’s numbers come from the 1988-90 stretch when was great, but those seasons only account for less than a third of his career innings. It’s not that those guys pitched better in big spots; it’s more that they tended to have their big spots when they were better pitchers.

    The Rivera thing….well, I can’t explain that. I will concede that Mariano Rivera is the one and only clutch player who has ever lived.

    Reply

  3. Posted by Carl on October 14, 2009 at 1:51 AM

    wow, you all have a lot of time on your hands… :::yawn:::

    Reply

  4. Posted by janechong on October 14, 2009 at 8:51 AM

    No need to subscribe for baseball-specific studies. Dan Ariely’s done a good deal of work dispelling the mystery (dunno about dispelling the myth) of clutch. Pretty straightforward, as you suggest. Boils down to same percentage, more tries. http://marketplace.publicradio.org/display/web/2009/07/06/pm_clutch_players_q/?refid=0

    Reply

  5. Posted by Douglas on October 14, 2009 at 12:13 PM

    Screw you, Carl!

    Reply

  6. […] for everything. There is some truth to that—I’ve already discussed how people like to look at disparate, inconclusive evidence and conclude that there is some larger meaning—but this is not the only element at work here. We want random, unconnected events to have […]

    Reply

  7. Posted by Brady on November 2, 2009 at 7:08 AM

    It’s funny you chose Mariano Rivera as a “clutch performer.” Closers have a huge advantage in the postseason because they are often playing teams that dont normally see them often. But when they do play a team that see’s them a lot, they usually get worked. Case in point, when the Red Sox caused Rivera to blow back to back save opportunities back in 2004 in games 4 and 5, thus making the greatest comeback in MLB history. Rivera also blew a 9th inning game 7 lead against the Diamondbacks. If you look at Jonathan Pappelbon, he just gave up a postseason run for the first time in his career, having had pitched 27 scoreless innings in a row in the postseason, a record for a closer!
    The Yankees have only missed the playoffs once in Mariano’s career, thus he has had far more opportunities to rack up innings pitched. Give it a few years, and I dont think you’ll see much difference in Pappelbon’s and Rivera’s numbers, except that Mariano will have more innings pitched in the Postseason, because the yanks buy every big name free agent.

    Reply

    • Posted by John S on November 2, 2009 at 8:56 PM

      Sorry Brady, but this reply is going to be mean, since you seem to be a Red Sox fan (or at least one of those Yankee haters who thinks signing free agents is either against the rules or unique to the Yankees) and you forced me to remember the 04 ALCS, which I have speant 5 years trying to repress. First of all, Papelbon’s 27 scoreless innings are not a record: Rivera threw 34 1/3 scoreless postseason innings from 1998 through 2001. Second of all, I think you’re overstating to say the Red Sox “owned” Rivera in the 2004 ALCS, as his ERA for that series was 1.29.

      It’s also true that closers may benefit from facing hitters who are unfamiliar with them in the World Series, but MOST of Rivera’s 100+ IP in the playoffs come against AL teams who face him every year. Also, the fact that Rivera’s career ERA in the postseason is .25 lower than Papelbon’s despite the fact that Rivera has thrown over 100 more playoff innings than Papelbon, indicates that they are hardly similar.

      Reply

  8. […] Yeah, those examples seem like fans seizing on particular failures and missing the general trend. I’ve touched on this before. Meanwhile, the Colts just took the lead back. Josh, you may have picked out your Super Bowl song […]

    Reply

  9. […] off the like of Joe Nathan and Brian Fuentes in the postseason? That A-Rod finally disproved the myth of clutch? That CC Sabathia, A.J. Burnett, and Mark Teixeira proved that they were worth every penny? Because […]

    Reply

  10. […] Rodriguez is “clutch” now, so we have to come up with new crimes for him to commit, like violating the age-old rule about not […]

    Reply

  11. […] Speaking of athletes, the NBA Finals have people questioning whether or not Kobe Bryant is actually a clutch player. When will people learn that clutch players are a myth? […]

    Reply

  12. […] Backe-Colby Lewis comparison, I don’t accept the premise. As you know, I don’t buy the idea that any players perform markedly better in the postseason than they do in the regular season. Even Backe only had two great starts in the postseason–we […]

    Reply

  13. […] it turns out Derek Jeter is actually clutch. Even though clutch players don’t exist. In other baseball news, an interview with Ken Singleton and Jonah Keri on the inflated value of […]

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: