[CubinNY]

Notes:

1. I don't have them, but I think numbers throughout the league differ in "clutch" situations, or late-situations, etc., from normal. Pitchers work more carefully, nibble mroe, walk more (OBP rises), but allow fewer hits. It's common to consider rotation pitchers superior to relievers, but year after year relievers allow lower batting averages but higher WHIPs. For the hitter, the situation is *not* the same as normal, because pitchers don't pitch the same as normal.

 

2. My understanding is that in baseball, studies indicate little evidence for such thing as a super-clutch hitter, guys who can significantly or consistently **outperform** their norm in clutch situations. Notable stuff is typically a result of small sample size, and tends to flatten out. (As suggested by reggie Jackson in WS versus playoffs...)

 

3. However, it's my understanding that many players are able to perform about as well in "clutch" situations as they do under normal circmstances. That's my idea of a clutch guy. He doesn't choke when the pressure is on. Doesn't overthrow, doesn't overswing, doesn't expand his strike zone, etc..

 

4. My hypothesis is that while few if any players can consistently "overachieve", that there may be some who "underachieve" in the clutch. Guys who are "anti-clutch". When the pressure is really on, the pitcher starts to overthrow, or his delivery starts to erode a little. When the pressure is really on, a hitter maybe swings too hard, or begins to expands his strike zone.

 

Vance, I think there have been saber stuff on supposedly "clutch" performers. Have there been saber studies on "anti-clutch" performers?

 

If my hypothesis was true, then a "clutch" guy is simply somebody who performs as well as normal in crisis, whereas an "anti-clutch" guy is somebody who doesn't.

 

The other thing is, if a guy is responding to stress or pressure or crisis or whatever, what constitutes a pressure situation for one guy might be quite different for somebody else.

 

Craig,

 

In order for data to support your hypothesis there would have to be enough observations. In order to do that one would have to define clutch. Using the most liberal terms one could say: two outs with a runner in scoring postion. It wouldn't matter what inning or which game. Then the performance would have to be significantly worse or better then the player's performance in non clutch situations (probably at least two standard deviations from the mean for that player).

 

Given these assumptions I would be willing to wager that data to support the existence of anti-clutch or clutch would be difficlut if not impossible to find

[treebird]

A really good point Tim makes everytime this argument comes up is that, in order to make it all the way to the major leagues, you have to perform under extreme amounts of a lot. You have to perform in front of scouts in high school and college, and in front of more scouts in the minor leagues. That's an enormous amount of pressure, in some cases you could probably argue that there's MORE pressure on the kid in high school.

 

It seems silly to say that you'd be able to make it to the major leagues without being able to perform under pressure.

[RynoRules]

A really good point Tim makes everytime this argument comes up is that, in order to make it all the way to the major leagues, you have to perform under extreme amounts of a lot. You have to perform in front of scouts in high school and college, and in front of more scouts in the minor leagues. That's an enormous amount of pressure, in some cases you could probably argue that there's MORE pressure on the kid in high school.

 

It seems silly to say that you'd be able to make it to the major leagues without being able to perform under pressure.

 

That's a good point. I also like craig's notion that a guy who is truly "clutch" may be one who simply performs at normal levels in abnormal situations. I might buy that.

 

That said, I think its fair to say that the Sox had an abnormal amount of clutch hits and pitching performances during their playoff run.

[BigbadB]

That said, I think its fair to say that the Sox had an abnormal amount of clutch hits and pitching performances during their playoff run.

 

Was that "clutch" hitting, or good timing?

 

I think "good timing" is a better terminology than "clutch". That "clutch" hit at a very opportune time can come from the worst player in baseball, therefore I buy into it being "good timing" rather than "clutch".

 

To hit .050 percentage points higher than the normal batting average with runners in scoring position isn't necessarily clutch. It's performing better under pressure situations.

 

To truly be clutch, I'd expect to see a guy almost always coming through in run producing situations.

[vance_the_cubs_fan]

I have a question for those of you who don't believe in clutchness... or clutchitivity... or clutchosity... anyways... when you're at a baseball game with a buddy and the game is tight and a guy steps up to the plate, your buddy says "I have a feeling he'll drive them in here, this guy is great in the clutch" what do you say to him? I'm just curious.

 

I just want the best hitter possible at the plate. I'm not concerned with what his numbers in the "clutch" are.

 

I'm not saying that in reality some hitters may or may not be better in certain situations. I'm sure some are and some aren't. But, it is impossible not only to measure but to accurately label most of these players. For the most part, truly great "clutch" hitters are already great hitters. They make pitchers pay for mistakes and they have the ability to hit great anyway. So, when placed in a big situation, they are more than likely to rise to the occasion. On the other hand, mediocre hitters are more likely to be poor to mediocre in such situations.

[Ding Dong Johnson]

Perceptions of clutch are quite similar to the weird feeling you get when you are thinking of somebody, the phone rings, and it's the person you were just thinking of. That event stands out as something special, but the times you were thinking of Jane and Dick calls (which is mostly the case) go completely unnoticed.

 

Derek Jeter's post-season reputation is a perfect example. He makes a few really good plays, and he is considered a clutch performer. His post-season line is 307/379/463/842. His regular season numbers are slightly better though: 314/386/461/847.

Edited January 5, 2006 by Ding Dong Johnson

[Careless]

are you implying that manny doesn't produce in "clutch" or, as i like to say "unicorn" situations?

If you don't believe there's such a thing as a clutch player, you must not watch much sports, and you clearly didn't play many either.

 

I watch a lot of sports and have played plenty in my time and I don't believe there is a "clutch player."

 

 

The numbers don't bear that out. If being "clutch" is an ability, then "clutch players" would have good numbers in "clutch" situations from year to year or at least numbers that are consistent with their career numbers. They do not.

So does that mean someone has addressed the matter of clutch pitching and how it affects hitters? Because that's(highlighted) simply not a statement you can make unless that's been done.

[noisesquared]

Every time I read a clutch hitting thread, my mind always goes back to this Bill James article someone posted a link to a while ago:

 

http://www.sabr.org/cmsfiles/underestimating.pdf

 

The pertinent part from Bill James:

 

We ran astray because we have been assuming that random data is proof of nothingness, when in reality random data proves nothing. In essence, starting with Dick Cramer’s article, Cramer argued, “I did an analysis which should have identified clutch hitters, if clutch hitting exists. I got random data; therefore, clutch hitters don’t exist.”

Cramer was using random data as proof of nothingness—and I did the same, many times, and many other people also have done the same. But I’m saying now that’s not right; random data proves nothing—and it cannot be used as proof of nothingness.

Why? Because whenever you do a study, if your study completely fails, you will get random data. Therefore, when you get random data, all you may conclude is that your study has failed. Cramer’s study may have failed to identify clutch hitters because clutch hitters don’t exist—as he concluded—or it may have failed to identify clutch hitters because the method doesn’t work—as I now believe. We don’t know. All we can say is that the study has failed.

 

As I understand it, James is saying that due to the small sample size of 'clutch' at-bats per year and the high amount of luck involved in every at-bat, it is impossible to compare yearly 'clutch' numbers - there is too much luck involved.

[Tim]

In every field, it is nearly impossible to prove that something does not exist. I don't believe in ghosts, but I cannot prove they are not real. However, the burden of proof should lie with the people who are positing that something does exist. I have yet to see any compelling evidence that there are players who are clutch.

 

What cracks me up about this argument is when people say it has a huge impact....but can't be measured. That is what I'd really like to have explained sometime.

[T2E]

The issue of Tejada's clutchiness led me to remember an article by noted Sabermetrician tangotiger who argued that clutch hitting was a slight yet detectable skill. One pertinent quote from the article:

 

TangoTiger[/url]"]So, who are the clutch hitters? From 1999-2002, Jason Giambi and Miguel Tejada have shown to have the true talent clutch ability to add 2 runs per year. That's it. That's the effect of clutch ability.

 

While the effect is 2 clutch runs for Giambi and Tejada, these clutch runs come at the most opportune time. The effect of the 2 runs, which normally contributes 0.2 wins, is 0.6 wins when considering the timing. Essentially 2 runs in a clutch situation is equivalent to 6 random runs. Therefore, it is best to consider these 2 runs as 6 leveraged-runs. 2 runs in a clutch situation, 6 random runs, or 0.6 wins. They are all equivalent.

 

Another sabermetric support of clutch hitting came from Andy Dolphin.

 

I personally agree with tangotiger, that clutch hitting does, in fact, exist, although its effects are generally small and most claims of clutchness are generally exaggerated.

[pete the dog]

In every field, it is nearly impossible to prove that something does not exist. I don't believe in ghosts, but I cannot prove they are not real. However, the burden of proof should lie with the people who are positing that something does exist. I have yet to see any compelling evidence that there are players who are clutch.

 

What cracks me up about this argument is when people say it has a huge impact....but can't be measured. That is what I'd really like to have explained sometime.

 

 

Exactly.

[Transmogrified Tiger]

Every time I read a clutch hitting thread, my mind always goes back to this Bill James article someone posted a link to a while ago:

 

http://www.sabr.org/cmsfiles/underestimating.pdf

 

The pertinent part from Bill James:

 

We ran astray because we have been assuming that random data is proof of nothingness, when in reality random data proves nothing. In essence, starting with Dick Cramer’s article, Cramer argued, “I did an analysis which should have identified clutch hitters, if clutch hitting exists. I got random data; therefore, clutch hitters don’t exist.”

Cramer was using random data as proof of nothingness—and I did the same, many times, and many other people also have done the same. But I’m saying now that’s not right; random data proves nothing—and it cannot be used as proof of nothingness.

Why? Because whenever you do a study, if your study completely fails, you will get random data. Therefore, when you get random data, all you may conclude is that your study has failed. Cramer’s study may have failed to identify clutch hitters because clutch hitters don’t exist—as he concluded—or it may have failed to identify clutch hitters because the method doesn’t work—as I now believe. We don’t know. All we can say is that the study has failed.

 

As I understand it, James is saying that due to the small sample size of 'clutch' at-bats per year and the high amount of luck involved in every at-bat, it is impossible to compare yearly 'clutch' numbers - there is too much luck involved.

 

Exactly, there's way too much variation for performance in 'clutch' situations to be predictive or more than an afterthought in player evaluation.

[MPrior]

Generally speaking, I'm with the "clutchiness-is-inconsistent-and-of-little-relative-significance-or-at-the-very-least-is-extremely-difficult-to-define-and-measure" camp, but the point Craig (I think) brought up - that is, that "clutch" hitting is more about being able to perform at normal levels under pressure, and that the true difference comes with players that can't perform at normal levels - is something that I had thought of that makes more sense to me. A good example of this (yes, I know we were mostly talking about hitting, but whatever) is LaTroy Hawkins. He's a great reliever, and if you look at his numbers, 2004 was a pretty good year for him too. He just could not save games to save (heh heh) his life. Particularly, if I remember correctly, 1-run saves were difficult for him - and not just because a 1-run save is harder. I seem to recall that he was quite effective when the team was up by 2 or 3 runs, but seemed incapable of pitching well with only a 1-run lead.

 

By the way, all that stuff about LaTroy Hawkins is based on my personal recollections - and I'm the first to admit that my personal recollections are both limited and subject to fault. So - I could be wrong.

[Transmogrified Tiger]

Generally speaking, I'm with the "clutchiness-is-inconsistent-and-of-little-relative-significance-or-at-the-very-least-is-extremely-difficult-to-define-and-measure" camp, but the point Craig (I think) brought up - that is, that "clutch" hitting is more about being able to perform at normal levels under pressure, and that the true difference comes with players that can't perform at normal levels - is something that I had thought of that makes more sense to me. A good example of this (yes, I know we were mostly talking about hitting, but whatever) is LaTroy Hawkins. He's a great reliever, and if you look at his numbers, 2004 was a pretty good year for him too. He just could not save games to save (heh heh) his life. Particularly, if I remember correctly, 1-run saves were difficult for him - and not just because a 1-run save is harder. I seem to recall that he was quite effective when the team was up by 2 or 3 runs, but seemed incapable of pitching well with only a 1-run lead.

 

By the way, all that stuff about LaTroy Hawkins is based on my personal recollections - and I'm the first to admit that my personal recollections are both limited and subject to fault. So - I could be wrong.

 

As I recall, Hawkins was pretty bad with 1 run leads, and he dealt with an abnormally large amount of them too. Bad combo.

[badger1679666666]

I'm sure Fred could help you two out...

[pete the dog]

Generally speaking, I'm with the "clutchiness-is-inconsistent-and-of-little-relative-significance-or-at-the-very-least-is-extremely-difficult-to-define-and-measure" camp, but the point Craig (I think) brought up - that is, that "clutch" hitting is more about being able to perform at normal levels under pressure, and that the true difference comes with players that can't perform at normal levels - is something that I had thought of that makes more sense to me.

 

 

Maybe, but I think the point is that there is no good evidence that any players generally perform any better or worse in "clutch" situations from year to year. I'm certainly no expert on the topic, but that is my understanding of the available data.

 

In any science, the null (assumed) hypothesis is that there is no real effect in your study. You seek evidence that forces you to abandon the null hypothesis, and accept that something real is happening. A failure to disconfirm the null hypothesis does not mean that there is no effect: It simply means that your study found no evidence that such an effect exists. James' description of this as a "failed" study is not quite right. There is no basis to call the study "failed" in any way. The study may accurately show that no real effect exists. Or not. Such an outcome simply means that if there is an effect to be found, you'd better look somewhere else (e.g., at a different set of statistics). As Tim rightly points out, the onus is on the proponents of an effect to statistically demonstrate its existence. The suggestion that "it's real but can't be measured" is a cop-out to the extreme, and is as good as admitting it's not real. Try publishing that in a scientific journal! :shock:

[MPrior]

Generally speaking, I'm with the "clutchiness-is-inconsistent-and-of-little-relative-significance-or-at-the-very-least-is-extremely-difficult-to-define-and-measure" camp, but the point Craig (I think) brought up - that is, that "clutch" hitting is more about being able to perform at normal levels under pressure, and that the true difference comes with players that can't perform at normal levels - is something that I had thought of that makes more sense to me.

 

 

Maybe, but I think the point is that there is no good evidence that any players generally perform any better or worse in "clutch" situations from year to year. I'm certainly no expert on the topic, but that is my understanding of the available data.

 

In any science, the null (assumed) hypothesis is that there is no real effect in your study. You seek evidence that forces you to abandon the null hypothesis, and accept that something real is happening. A failure to disconfirm the null hypothesis does not mean that there is no effect: It simply means that your study found no evidence that such an effect exists. James' description of this as a "failed" study is not quite right. There is no basis to call the study "failed" in any way. The study may accurately show that no real effect exists. Or not. Such an outcome simply means that if there is an effect to be found, you'd better look somewhere else (e.g., at a different set of statistics). As Tim rightly points out, the onus is on the proponents of an effect to statistically demonstrate its existence. The suggestion that "it's real but can't be measured" is a cop-out to the extreme, and is as good as admitting it's not real. Try publishing that in a scientific journal! :shock:

 

Oh I'm with you - and I'm pretty familiar with the null hypothesis - I even had a minor argument with someone on a related note about it. I think we were loosely discussing clubhouse chemistry.

[Magnetic Curses]

are you implying that manny doesn't produce in "clutch" or, as i like to say "unicorn" situations?

If you don't believe there's such a thing as a clutch player, you must not watch much sports, and you clearly didn't play many either.

 

um, okay, i don't really know you, but feel free to insult me personally if you wish. you must not watch many friends, and you clearly don't have any either.

 

the clutch stat in baseball is the unicorn stat, it does not exist. players who are better hitters hit proportionately better in clutch situations. numbers vary from year to year, but will likely be consistent with their career avg, obp, slg., etc.

[vance_the_cubs_fan]

Shouldn't this discussion be moved to "Baseball Discussions"?

[nick23]

I think there are certain players that rise up to the occassion who obviously are always good players but seem to be even better in pressure situations - intangibles if you will.

 

MJ is a shining example of one of them.

 

I don't think it makes sense to compare across sports.

 

 

Why not. People were debating whether clutch was real, and I gave an example of an athlete that was by far the most clutch player in sports history.

 

 

I'm generally for objective analysis but I don't think statistics tell you all there is to know about every ball player. You could have two equal players based on basic metrics, but one might just respond better to pressure situations.

 

I don't think that concept is so foreign that it should be mocked (not saying you are mocking it goony - but treebeard sort of did). I wouldn't make 90% of my decisions based on it, but as former assistant coach/private coach I had an idea of who I would want under pressure, and I certainly didn't study numbers.

 

It's simple. You want your best player up in crucial situations. Not the guy who seems like he always comes through.

Edit: I have no idea why my font is so large. It's on normal.

[Magnetic Curses]

I think there are certain players that rise up to the occassion who obviously are always good players but seem to be even better in pressure situations - intangibles if you will.

 

MJ is a shining example of one of them.

 

I don't think it makes sense to compare across sports.

 

 

Why not. People were debating whether clutch was real, and I gave an example of an athlete that was by far the most clutch player in sports history.

 

 

I'm generally for objective analysis but I don't think statistics tell you all there is to know about every ball player. You could have two equal players based on basic metrics, but one might just respond better to pressure situations.

 

I don't think that concept is so foreign that it should be mocked (not saying you are mocking it goony - but treebeard sort of did). I wouldn't make 90% of my decisions based on it, but as former assistant coach/private coach I had an idea of who I would want under pressure, and I certainly didn't study numbers.

 

It's simple. You want your best player up in crucial situations. Not the guy who seems like he always comes through.

Edit: I have no idea why my font is so large. It's on normal.

 

despite your large font, you've hit the nail on the head.

[fromthestretch]

I have no idea why my font is so large. It's on normal.

 

Be careful. Some posters might get font envy.

[nick23]

One question for you guys believing in clutch hitting. Let's pretend like Jauque Jones outpreforms Derrek Lee this year in "clutch" situations by a decent size margin. Now lets pretend that we're playing Houston in a one game playoff for the wild card. There's 2 outs in the bottom of the 9th inning and Brad Lidge is on the mound. Runners are on 1st and second. Very, very crucial situation. Now remember, Jones has clearly outpreformed Lee this year in terms of clutch hitting. Who would you rather have up Derrek Lee or Jauque Jones?

[BigbadB]

One question for you guys believing in clutch hitting. Let's pretend like Jauque Jones outpreforms Derrek Lee this year in "clutch" situations by a decent size margin. Now lets pretend that we're playing Houston in a one game playoff for the wild card. There's 2 outs in the bottom of the 9th inning and Brad Lidge is on the mound. Runners are on 1st and second. Very, very crucial situation. Now remember, Jones has clearly outpreformed Lee this year in terms of clutch hitting. Who would you rather have up Derrek Lee or Jauque Jones?

 

Alex Gonzalez! :D

[nick23]

One question for you guys believing in clutch hitting. Let's pretend like Jauque Jones outpreforms Derrek Lee this year in "clutch" situations by a decent size margin. Now lets pretend that we're playing Houston in a one game playoff for the wild card. There's 2 outs in the bottom of the 9th inning and Brad Lidge is on the mound. Runners are on 1st and second. Very, very crucial situation. Now remember, Jones has clearly outpreformed Lee this year in terms of clutch hitting. Who would you rather have up Derrek Lee or Jauque Jones?

 

Alex Gonzalez! :D

 

Well that goes without saying. No player was better in the clutch than Kgonz. Unless we're talking about defense. :(