Who Has Been Clutch?

by James Kannengieser on Aug 25, 2009 6:00 AM EDT

The are seemingly infinite ways to measure "clutchness" -- performance w/RISP, late and close situations, September, RISP and 2 outs, etc. Beyond the Boxscore suggested one other method, which I will use to look at the Mets' performance in the clutch. The method entails comparing Batting Runs (BR, the offense component of Fangraphs WAR) to WPA. By subtracting BR from WPA*10, we can see how clutch each Met has been. "Clutch" here is a measurement of how much better or worse a player has fared when factoring leverage, compared to overall context-neutral performance. Here are the top 5 and bottom 5 Mets this season, minimum 50 plate appearances, through Sunday's games. Remember that 10 runs = 1 win and click here for the full list of Met hitters (pitchers not included):

Rank	Player	BR	WPA*10	Clutch
1	David Wright	24	31.3	7.3
2	Brian Schneider	-9.5	-4.3	5.2
3	Cory Sullivan	0.8	4.3	3.5
4	Jeremy Reed	-7.2	-4.7	2.5
5	Alex Cora	-10.6	-8.5	2.1

Rank	Player	BR	WPA*10	Clutch
14	Jose Reyes	1.7	-1.4	-3.1
15	Luis Castillo	10.6	7.1	-3.5
16	Angel Pagan	4.9	-3	-7.9
17	Jeff Francoeur	1.8	-7.1	-8.9
18	Fernando Tatis	-3.7	-18.1	-14.4

This isn't a predictor of future performance, as we know that the "clutch hitter" is almost certainly a myth. Just look at last year's results, which pegged Tatis as one of the best clutch hitters (also note the name at the top of the list). Keep in mind that luck often plays a role in WPA, so, for example, Francoeur's unfortunate line drive turned triple play on Sunday really hurts him here. This technique for assessing past performance is not infallible, but is nonetheless worth looking at.

Tweet Comment 23 comments | 0 recs |

Do you like this story?

Comments

The fact that schneider is the second clutch player

and castillo is 15th is absurd

"We're investigating the investigative procedure of the investigation of Tony Bernazard"---Omar Minaya (he really didn't say it but he would"

by firejerrynow on Aug 25, 2009 7:10 AM EDT reply actions

Interesting post

I remember blogging about Tatis and his clutchness last year. You could hardly rely on any one stat to determine future performance. Who could have predicted Wright would simply shave 70 points off his SLG% this season?

Are those Francoeur’s full season or his Mets split?

by Joe D. on Aug 25, 2009 8:38 AM EDT reply actions

Doesnt make sense

I checked Francoeur and that is his Mets split.

Now, considering he has twice as many RBIs than the next highest player during his tenure, how can he be that unclutch?

Something smells rotten or these particular stats are seriously flawed.

by Joe D. on Aug 25, 2009 8:49 AM EDT up reply actions

One more thing...

Credit Keith Hernandez for this, but on Friday he said that what he liked about Francoeur was his productive outs. Are productive outs considered when measuring clutch? Shouldn’t they be?

by Joe D. on Aug 25, 2009 8:55 AM EDT up reply actions

no

outs are only really productive if you are tied or behind by one in the 9th, and you move a runner from first to second. Other then that, they really aren’t.

"We're investigating the investigative procedure of the investigation of Tony Bernazard"---Omar Minaya (he really didn't say it but he would"

by firejerrynow on Aug 25, 2009 9:30 AM EDT up reply actions

RBI's don't measure clutch

"We're investigating the investigative procedure of the investigation of Tony Bernazard"---Omar Minaya (he really didn't say it but he would"

by firejerrynow on Aug 25, 2009 9:30 AM EDT up reply actions

Joe, I suggest reading up on WPA to better understand this method

WPA factors game situation, so an RBI when the Mets are up 10 runs is nowhere near as valuable as 1 when the game is tied in the 8th or 9th inning. And, as fjn noted below, RsBI, in a vacuum, don’t really tell us much about a ballplayer.

by James Kannengieser on Aug 25, 2009 10:42 AM EDT up reply actions

thanks

I agree that some RBI’s are more valuable than others, and that would explain a few things and answer my question as well. I appreciate the explanation..

by Joe D. on Aug 25, 2009 10:02 PM EDT up reply actions

I love that 3 of the 5 clutchest Mets

have sucked so bad, that sucking slightly less in high leverage has led to their clutchness.

by dtro on Aug 25, 2009 8:45 AM EDT reply actions

I was going to say the same thing

I think the list of Mets with positive WPA for the whole season is probably very short.

by JoshNY on Aug 25, 2009 9:36 AM EDT up reply actions

i really hate WPA to evaluate individual players

interesting to look at in the team context, especially game-to-game with fangraphs’ game charts, but i really don’t think the stat tells us much about an individual’s performance. batters, for example, too often get penalized in WPA for the baserunning mistakes of their teammates.

side note: david eckstein is the fifth most “clutch” hitter in MLB this season. maybe we’ve been wrong all along.

by englishgrey on Aug 25, 2009 10:29 AM EDT reply actions

I thought it was a myth

that there is no such thing as clutch. It is not very important, but there is evidence of a slight effect. But wtfdik?

http://www.hardballtimes.com/main/article/the-color-of-clutch/

by wobatus on Aug 25, 2009 10:49 AM EDT reply actions

Not a total Myth IMO

I don’t believe a player can actually perform better in a “clutch” spot, but they surely can come up short.

I would define being clutch as not performing worse than that player typically does when there is a higher leverage situation.

by Balagast on Aug 25, 2009 11:06 AM EDT up reply actions

I generally subscribe to this

by James Kannengieser on Aug 25, 2009 11:11 AM EDT up reply actions

yup

I tend to agree. I think anonynmous and i discussed this in another thread a while back. Whatever your take on it generally, it would not make sense for someone to perform better than they are generally capable of (although there was always the “Ollie gets up for big games” haha) under pressure, but human nature would suggest someone could come up slightly worse.

In any event, the stats evidence is there is some small effect evidently.

Now, for a contra example:

In the book With the Old Breed, by Eugene Sledge, who was marine infantry at Peleliu and Okinawa, he wrote about how some americans had some japanese troops pinned down. One by one the japanese would run a guy out every 30 seconds or so. He would be killed easily. But eventually the american riflemen became lazy. It was a turkey shoot, and they would not aim properly, or shoot from the hip, and one or 2 would get out this way, maybe hit but not killed.

Not really relevant, but there must be times in a blowout in a long season you just go up hacking and say I got me some dinner reservations, let’s get the f outta here.

by wobatus on Aug 25, 2009 11:53 AM EDT up reply actions

I cannot remember one clutch hit

that Brian Schneider has had all year. He hit 2 3-run homers, and a 3-run double, that’s been his season.

by David G on Aug 25, 2009 11:21 AM EDT reply actions

Schneider

Here is his play log. Those 2 3-run homers were hit in decently big spots, leading to his high spot on this list. Conversely, it looks like he hasn’t really been “un-clutch”, as in he hasn’t had many large -WPA plate appearances.

by James Kannengieser on Aug 25, 2009 11:25 AM EDT up reply actions

One of those homers

was very early in game. Would like to have seen Church’s rating here. He was bad with risp, but had a couple big hits before being traded, including one vs. Pitt in extra innings if I remember correctly.

by David G on Aug 25, 2009 1:45 PM EDT up reply actions

Clutchness

Given it’s proven that human beings perform better, worse, or at the very least, different under high stress, I find it odd that some people claim that it does not exist in baseball.

Granted baseball has a larger random chance factor than other sports, but still, I think everyone can immediately think of a bunch of performances in other sports where players clearly choke. Golf putts, football field goals, basketball free throws, etc. You watch it, and you can clearly recognize a human being psychologically imploding.

I think in baseball, the random factor obscures the data, such as Francoeur’s line drive triple play. Or conversely, a check-swing game winning infield hit.

I think that’s where actually watching the games provides a better subjective feel than trying to crunch numbers on extremely small sample sizes. Crushing a line drive with RISP in the 9th results in my brain concluding that was a very good thing, whereas by the box score number, it’s a triple play, and the worst possible outcome.

In any case, the question I would ask is, “Why is baseball the only human activity in which people do not exhibit any difference in performance when under pressure?”

I believe that some players, being human, perform better under pressure than others, but there isn’t an adequate way to measure it, mostly because trying to quantify “pressure” is a losing battle. It varies wildly by person and situation.

Going back to Francoeur’s triple play, even if you agree that hitting a hard liner was a good thing, I don’t know if I would call that a “pressure” situation. There were no outs at the time, and it came in a game that is basically meaningless for the Mets.

So there was more pressure than if the Mets were up 10-0 in the 1st inning, but less pressure than if there were 2 outs, or if it was a crucial game.

It gets nebulous fast, and is so subjective that I’d rather just watch the game, and form my own opinion of who I want up at the plate in critical spots.

by Mex_17 on Aug 25, 2009 1:48 PM EDT reply actions

Well

No one really denies that clutch hitting exists, just that it is not predictive of future clutch hitting. See: David Wright RISP 2008 v. David Wright RISP 2009.

by Jeffrey Paternostro on Aug 25, 2009 2:15 PM EDT up reply actions

you know what? there's really no point in rehashing this every time someone uses the word "clutch"

But the point of the anti-clutch argument is that this is a naive oversimplification of what “high-stress” means — it basically is based on wishfully pretending that the fans’ stress moments are the same as the players’. Every MLB player ever to play the game has already been selected for their performance over thousands and thousands of incredibly stressful at-bats, on which their career effectively depended. Anyone who makes it to MLB has already outperformed 99-point-whatever percent of humanity at high-stress baseball performance.

by anonymous on Aug 25, 2009 3:55 PM EDT up reply actions

true

But his point seemed to include, hey, so had jean Vandervelde when he skidded an easy chip into a brook at the British Open, or Watson when he yipped a short putt this year. Then again, I am sure that’s just selective bias too. Everyone misses ’easy" putts now and then.

And somehow, Eddie Murray just never seemed stressed to me, spring training bases empty, risp with 2 outs late in a pennant race. Like he could give a shit.

I do think you are onto something though. I don’t think the game is as stressful to the players as say driving on 95 Labor Day weekend traffic. The streesful stuff is not in “clutch” situations it seems but when normal performance routine starts failing them: Ankiel or Blass throwing even close to the plate, Sasser returning the ball, sax and Knoblauch soft tosses to first.

Ankiel did melt down in the playoffs. But Blass was great in the series. he failed the NEXT year, early.

by wobatus on Aug 25, 2009 4:47 PM EDT up reply actions