What is Clutch?

[BrewCityUnit]

Or should I say, who is clutch?

 

Derek Jeter. They just showed a stat that said he had the highest batting average with 2 outs and runners in scoring position (.436) and he just jacked a 3 run bomb with 2 outs in the 8th inning to give the Yankees the lead over the Red Sox.

[1992casey]

I think "Who is clutch?" is nearly impossible to define.

 

"What is clutch?" has a chance of being defined, but tends to be defined poorly. As an example, people love to use RISP stats. A game is tied in the bottom of the ninth. The first batter triples. One of the following batters singles him in. My gut says the triple is at least as "clutch" as the single--and probably more "clutch", but it doesn't count in the RISP stat. (I bet Russ can tell us for sure. http://forum.brewerfan.net/images/smilies/smile.gif)

 

Here's an interesting thread with some background:

 

Article On Clutch Hitting

[mothershipconnection]

"Late and close" is a much better indicator than RISP, if you want to talk "clutch". RISP is crappy because it depends on another hitter having reached ahead of you. Thus, it can also reflect how well you hit when a pitcher is beginning to unravel (in that he's already allowed 1 hitter to reach scoring position). A homer with a man on 1st is more impressive than a single with a man on 2nd.

[1992casey]

"Late and close" (I think it's officially "close and late") would seem to be on the right track. But it has horrible sample size issues. There are few players where you can come up with the 1,500 or so ABs you need to make the sample meaningful. And if you're only dealing with a small number of players, how reliable a calculation can you actually get?

 

From what I understand, the 'next step' toward attempting to identify clutch hitters would include every AB, properly assigning a leverage factor to each of them.

[TooLiveBrew]

David Ortiz was officially clutch before popping out to end the game. Now he's officially not clutch. If he can get a walk-off hit again, then he'll officially be clutch again. See how it works?

 

http://static.yuku.com//domainskins/bypass/img/smileys/wink.gif

[DougJones43]

I always like to look at Win Probability Added. It takes game situations into account very well, better than RISP or "close and late" in my opinion. For what it's worth, Ortiz is #6 in the ML in WPA, and Jeter is #37 (behind our own Corey Hart) although today's heroics aren't reflected as of yet. Prince is #4. By this metric, we have about 5 more wins than we would have without Fielder in the lineup. This stat is only for hitting, defense is not taken into account.

[Principessa21]

I don't believe in clutch, esp. since last August when Mr. Clutchity Clutch Clutch (aka D. Ortiz) himself had a panic attack during the most important series of the year - vs. the Yankees. http://forum.brewerfan.net/images/smilies/wink.gif

 

ETA:

http://nomaas.org/images/jeterspanksschilling.jpg

 

also from nomaas.org:

http://nomaas.org/images/Ortizdancing1.jpg

[jason21nl]

I know most on this site really dislike Jeter but when dude needs to come up with a hit he seems like he always gets it. If I needed a guy to come up with a 2 out hit to win a world series, Jeter would be the guy Id want up. If I needed a 2 run shot, Id probably want up Ortiz.

[sbrylski]

There is a statistical "clutch" monitor called Leverage Index (LI). It measures the statistical importance of the plate appearance to the outcome of the game. WPA, as DougJones mentioned, is the stat that weights the outcome of a PA based on its importance to the game. So, if you look at a players WPA in high LI situations, I think that's a pretty good clutch indicator in retrospect. As far as predicting which players will be clutch in the future, good luck!

 

Fangraphs has a solid clutch stat, IMO. Its defined as [WPA - (OPS Wins * LI)] and states that Albert Pujols as been the most clutch player this season.

[endaround]

And again Jenkins is one of the most clutch players in the league. Which pretty much means people can't even tell clutch when they see it.

[craigharmann]

Joe Dillon was clutch yesterday.

[Baldkin]

When its 9:30 and your fridge looks like this:

http://www1.istockphoto.com/file_thumbview_approve/2629099/2/istockphoto_2629099_empty_fridge.jpg

 

This guy, is the most clutch.

http://freewheelbike.com/merchant/589/images/site/Sam%20w%20Keg.jpg

[Fatter than Joey]

Wow -- Ben Sheets is not clutch.

[sgtcluels]

Clutch is a band from Maryland!

 

For thanksgiving we had taters, succotash and Studebakers

[brettac1]

That HR was most definitely clutch. Where you run into issues regarding the idea of clutch is whether or not a player has an inherent ability to repeatedly raise his performance (or keep it the same as in other situations) during those types of situations. Everything I have seen seems to say that players performance in these situations will come to very closely resemble their normal stats after enough AB's.

 

In short, the hit is clutch, the hitter isn't.

[1992casey]

It's pretty easy to identify an individual game situation as clutch. Prime examples would be Dillon yesterday and Cordero's eighth inning performances the past two games. Adding the situations up and trying to get meaning out of them is where problems arise.

[TuesdaysWithRillo]

There is no such thing as clutch. Even ask Sabremetrics people

[Ennder]

Clutch is the bunching of events over a small sample. Its variation just like anything else.

[TooLiveBrew]

In short, the hit is clutch, the hitter isn't.

 

That's probably the best way to sum it up.

[edod]

Joe Dillon was clutch yesterday.

 

Could we also say Yost was clutch in this scenario?

[mothershipconnection]

Adding the situations up and trying to get meaning out of them is where problems arise.

 

That's the problem - people are biased toward the event most salient to them. A player could K 4 times in game six, where a big hit could win it, then hit a homer in the 9th in game 7 to win it and he's clutch. What about the times he choked in situations that were very big, but just not quite as big? Then there's the guy who hits the homer in the 8th to put his team within one (not clutch) compared to the guy who hits a 2 run shot in the 9th to win ("Clutch-Nasty"). Or the guy who has the huge hit to tie the game in the top of the ninth, but then his team blows it in the bottom, so he is post-hoc not clutch.

 

Anyone who'd rather have Mr. Clutch Craig Counsell hitting in a pivotal situation instead of unproven Prince Fielder or Anti-Clutch Arod simply doesn't pay enough attention.

[logan82]

I think in general guys who are clutch usually are good hitters to begin with. Usually with good plate dicipline(low strikeouts) and the ability to just put a ball in play instead of trying to hit a homer. Like mother said, clutch is more of a matter of perception and faulty memory than something that cam actually be measured. If it is measured clutch depends on what criteria are used.