Kosuke the Clutch
Down three runs in the ninth inning on opening day, Kosuke Fukudome belted a three-run homer off of Eric Gagne to tie the game. The Cubs would ultimately lose, but Kosuke had earned himself the clutch label, and labels can be very hard to break. In certain cases these labels are undeserved, like calling David Ortiz clutch based on past results; the stats here at Fangraphs show he was one of the least clutch players last season and he has a -0.32 score this season. Likewise, players with negative labels will find themselves hard pressed to shake off the opinions pointed in their direction.
Looking at Kosuke’s numbers, however, it seems he was properly labeled, at least so far. In case you cannot recite all of his numbers off the top of your head, here they are:
Fukudome 2008
46 GP, 51-167, 11 2B, 2 3B, 2 HR, 17 RBI, 30 BB, 27 K
.305/.409/.439, .355 BABIP
1.11 WPA, 0.58 WPA/LI, 0.52 Clutch
More facts about him? Well, his home/road splits have been pretty drastic:
Home: .393/.505/.562
Road: .205/.292/.282
And his splits with runners on base are drastic as well:
Nobody On: 104 PA, .261/.346/.337, 12 BB, 18 K
Runners On: 95 PA, .360/.479/.547, 18 BB, 9 K
The clutch score is what stuck with me, though, as he ranks 10th in the NL. Looking at a few other areas that could determine clutchiness, the results seemed to match:
High Leverage: 40 PA, .412/.462/.647
Within 1 Run: 49 PA, .361/.463/.482
Tie Game: 100 PA, .419/.490/.581
The clutch stat here does not measure performance in clutch situations but rather how a player performs in the high leverage situations as compared to context-neutral situations. Based on that definition, Kosuke is definitely stepping up in the situations in which his team needs him to. He might not be lighting the world on fire with extra base hits but these numbers clearly suggest Kosuke still deserves the label assigned on opening day.
23
The opening day HR accounted for .454 WPA in itself. I don’t know how much of that is “Clutch” and how much you get for an average HR, but doesn’t that mean that the majority of his clutchiness came on that one swing? If so, it seems like he hasn’t really been all that clutch since being given the label.
Good idea. The clutch formula is:
Clutch = WPA/pLI – WPA/LI
pLI refers to the player’s average LI in the game and WPA/LI is the context-neutral wins. In the opener, Kosuke’s home run had a +.454 WPA, with an LI of 3.52. He had four at-bats which, when averaged together, had a pLI of 1.47. His WPA/LI for the play would be .454/3.52, or 0.13.
Kosuke HR = .454/(1.47 – 0.13)
Kosuke HR = .454/1.34 = 0.34
So, please correct me if my calculations are wrong, in the pLI sense, but if not it looks like his opening day HR accounted for 0.34 of his Clutch Score, or 65.3%; if this is the case, while he hasn’t been unclutch since the opening day HR, it did account for 2/3 of his clutch.
You’ll get about .13 wins on average for a home run. So if you do take that away, he’s still been a little bit clutch over the course of the season. He might also have one particularly “unclutch” moment too, so unless you’re looking at his full “negative Clutch” and “positive Clutch” (which we don’t calculate and haven’t, though maybe it’s something to do), I think it’s misleading to take one hit and say without that he wouldn’t have been clutch, because you might very well be able to say the opposite in that without a particular “unclutch” moment, he would have been clutch.
Of course… he’s had all of 200 plate appearances, and who knows if he’ll continue to be clutch. Who knows if anyone will continue to be clutch. That’s what the The Great Clutch Project is all about.
Yeah, exactly. Just like with WPA, Kosuke might have something like +1.02 positive Clutch and -0.50 negative clutch, so the sum thus far has given him a high clutch score, even if he hasn’t been clutch in every single plate appearance or game.
If the calcs I did above are correct, then that HR would account for +0.34 positive clutch, but that’s not to say it wasn’t erased or lessened by other unclutchy plate appearances. It would be +0.34 of his positive clutch, not overall clutch.
“I think it’s misleading to take one hit and say without that he wouldn’t have been clutch, because you might very well be able to say the opposite in that without a particular “unclutch†moment, he would have been clutch.”
I understand this; I was only making my original observation because of when we (whoever “we” is) started thinking of him as clutch.
So BCL (Before Clutch Label) he was +0.34 clutch in one game (thus causing the clutch lablel), and ACL he’s been +0.18 clutch in 45 games. That’s all I was getting at.
Vegas Watch: Yeah, he definitely has been as “clutch” as he was in his first game.
When I think of him, I think of that first game and his big “clutch” hit, but that’s only because I haven’t really been following him closely. What will be interesting to see is how long he’ll be able to continue with this “clutch” label in people’s minds.
How much good will with the fans does that first clutch hit give him? I know with someone like A-Rod, clutch hits don’t seem to stay in memory for that long. ;)