The Great Clutch Project

by tangotiger on Mar 24, 2008 12:16 PM PDT Comment 25 comments

The Great Clutch Project

Time to put up or shut up:
http://www.tangotiger.net/clutch/

5 recs | Comment 25 comments

Story-email Email Printer Print

Comments

Major recommendation.

Everybody should do this.

stat-addled alien overlord

by salb918 on Mar 24, 2008 12:29 PM PDT reply actions 0 recs

So many guys on the A's roster who I haven't watched play. Even if I was someone who strongly believed in clutch performances, my only knowledge about Emil Brown, CGon, Ryan Sweeney, etc, comes from looking at their stats.

So maybe I'll just click over to the Blue Jays roster and mark down Scoot as a CLUTCH GOD!!!

by colin on Mar 24, 2008 1:09 PM PDT reply actions 0 recs

I may be underestimating his true platoon split

but Hannahan's good contact skills and great strike zone judgment make him a pretty good "clutch candidate" for me.

You could probably say the same for Barton and Buck, but they're already good hitters in all situations. I suppose they could still "turn it up" in the ninth inning though.

I think the one thing that A's fans will actually agree on is that Chavez is not clutch... although I think that's partly an artifact of his awful platoon split.

Your 2008 Athletics: It's Nothing Personal.

by PaulThomas on Mar 24, 2008 1:14 PM PDT reply actions 0 recs

You pegged my responses.

Barton, Buck and Hannahan all got above average ranks from me. Nearly everybody else got average, except for Chavez and Crosby.

The key to this for me was not that they would outperform themselves in high pressure situations, but the other team members. I think Buck, Barton and Hannahan will be like Buck, Barton, and Hannahan (respectively) in "clutch" situations.

Chiba Lotte lost simultaneously in the fighter plane of the Japanese ham.

by JediLeroy on Mar 24, 2008 5:03 PM PDT up reply actions 0 recs

JL, are you and PT therefore advocates of the Devo Conundrum?

Namely, that "clutch" per se doesn't exist, but unclutch (or, as PT more precisely puts it, "artifacts of awful platoon splits") does?

way to speed humanities demise douche bags! @('.')@

by monkeyball on Mar 24, 2008 7:15 PM PDT up reply actions 0 recs

I have my own conundrum? nice!

"It's for your own good. Big strong Devo knows whats best for Poppy" -- Mossback

by devo on Mar 24, 2008 7:25 PM PDT up reply actions 0 recs

well, here's the "conundrum" part

... apologies if I don't recall your response to this from some previous thread, but it seems to me that if the "soft unclutch" argument ("clutch" doesn't exist; a small handful certain batters due to situationality are better able to maintain their mean perf levels in clutch situations; and an identifiable set of players consistently underperform their mean perf levels in clutch situations) holds true, then we'd see a small but distinct overall drop in mean performance level in clutch situations -- which I don't think we do.

way to speed humanities demise douche bags! @('.')@

by monkeyball on Mar 25, 2008 8:00 AM PDT up reply actions 0 recs

It isn't that simple

because there are some subtle selection bias problems to deal with. The league as a whole hits slightly better with men on than with the bases empty, largely because bad pitchers (or good pitchers having a bad day) are more likely to put runners on in the first place than good ones are. Countering that, managers will be more inclined to use better relievers in close and late situations. The net effect is going to depend on exactly which situations you consider clutch.

There are (at least) 3 plausible explanations, beyond it simply being an innate talent, that might explain clutch/unclutch differences:

1. PaulThomas' theory, aka the Chavvy effect, that unclutch players are largely those with extreme platoon splits or some other exploitable weakness. There is clearly some truth in this in particular cases, but in general the effect will be small, because the range of true platoon splits is very small to begin with, and hard to detect, because you'd be trying to correlate two things that each have low signal-to-noise.

2. devo's theory, aka the Scutaro effect, that clutch players are good fastball hitters. That's a pretty clever idea that certainly fits some of the anecdotal data, and I don't think it's one I've heard elsewhere. But again, it's going to be hard to test systematically, for essentially the same reasons as #1.

3. My (unoriginal) theory, aka the Moneyball effect, that older players are better in clutch situations than younger ones. I don't remember the details, but I think there was some study that did find a small but significant effect of players getting better in clutch situations as they grew older. That still isn't explanatory in and of itself, and ultimately could be related to psychology (veterans get more comfortable in clutch situations as they gain experience), a form of PaulThomas' theory (veterans are less likely to have exploitable weaknesses than younger players), or another form of selection bias (the survivor effect, that better players tend to have longer careers).

In any case, as I understand it, Tango's experiment here isn't going to say much about whether clutch ability exists, it's just going to show that fans aren't very good at identifying that ability, which I think most of us already accept.

"Tomorrow it may rain." - Leo Durocher

by andeux on Mar 25, 2008 10:27 AM PDT up reply actions 0 recs

all men should

I'm here to talk about the past.

by 67MARQUEZ on Mar 26, 2008 12:37 PM PDT up reply actions 0 recs

I don't know that our positions are identical

My postulate, I guess you could call it (although I'm not convinced it's strictly true) is that most of what appears to be "clutch hitting" or "choke hitting" is non-psychological, and an artifact of game-situational concerns like usage of situational relievers.

To reduce it to absurdity, if every team used relievers for the first 3 innings and a "starter" (obviously a misnomer, but bear with me) the rest of the way, a lot of "unclutch" hitters would appear "clutch," and vice versa. I believe Curtis Granderson would become "clutch" in that scenario, for instance.

Devo may or may not agree with this, but I don't think that's what he means when he says "clutch hitters don't exist but unclutch hitters do."

Your 2008 Athletics: It's Nothing Personal.

by PaulThomas on Mar 24, 2008 9:53 PM PDT up reply actions 0 recs

Nope ... not what I mean ...

though I agree that situationality plays a big role. What do Marco Scutaro and David Ortiz have in common? They both produce a disproportionate portion of their production off of fastballs. What do late inning relievers tend to rely on more than starters?

I'm working on a way to test this hypothesis ....

"It's for your own good. Big strong Devo knows whats best for Poppy" -- Mossback

by devo on Mar 24, 2008 11:01 PM PDT up reply actions 0 recs

During a quiet moment,

you might mention that to Keith Foulke.

I like Cindi. A. She never pretends to know more than she does. B. She has unbridled enthusiasm for her "Hotties," and isn't afraid to show it. -IM4Oakgal

by Nico on Mar 26, 2008 3:27 PM PDT up reply actions 0 recs

I didn't say it was necessarily a good fastball ...

but Foulke does throw it 65% of the time ... (I'm not sure if that's a lot or not ... thought next Staturday you might ...)

"It's for your own good. Big strong Devo knows whats best for Poppy" -- Mossback

by devo on Mar 26, 2008 3:55 PM PDT up reply actions 0 recs

It's not a "good" fastball,

but it's an effective fastball thanks to location - as ManRam can attest to not once but twice!

I like Cindi. A. She never pretends to know more than she does. B. She has unbridled enthusiasm for her "Hotties," and isn't afraid to show it. -IM4Oakgal

by Nico on Mar 26, 2008 4:09 PM PDT up reply actions 0 recs

I'm not quite sure what I believe

But that has always seemed more plausible to me.

Chiba Lotte lost simultaneously in the fighter plane of the Japanese ham.

by JediLeroy on Mar 25, 2008 6:19 AM PDT up reply actions 0 recs

Variance

I've been wanting to test a theory for a while, which would piggyback nicely on Tango's experiment. I've always felt that clutch hitting had an awful lot to do with variance. For example, a guy who goes 1-4 every day is a .250 hitter; a guy who goes 4-4 every 4 days is also a .250 hitter. What we perceive when we think "clutch" is a player's consistency when that clutch opportunity arises. In other words, Robot Scutaro who hits 1/4 every day is much more clutch than Feast-or-Famine Cust. Here's why it seems so:

Give a certain random probability that a player gets a clutch opportunity - it'd be some function of how often that opportunity comes up for the team, and the likelihood the batter in question gets to bat. Say, as an extremely off the cuff example, that the clutch opportunity happens once every 4 games, and that opportunity encompasses 3 batters (3 outs). So a player may have a 1/12 chance of being in a clutch situation. Now here's where the perception kicks in:

If you've got someone who goes 0-4 on three days, and 4-4 on the 4th day, then you basically only have 1 day where he gets a hit in that clutch situation. If the batter goes 1-4 every day, then he's got all 4 days where he can be "clutch."

Testing this isn't all that hard, but I don't know where to get datasets en masse. That, and when I pick out just a few players randomly, the variance of, er, their variance is too much. Also, I have never had a definitive "clutch" list to compare against.

Any help in obtaining those two things would be much appreciated.

by dscel on Mar 25, 2008 2:25 PM PDT reply actions 0 recs

That's not mathematically correct ...

Say, for instance, that on average both Marco Oneforfourero and Jack All or Custing both will come up in a clutch situation one out of every ten games (the actual number doesn't matter).

Marco Oneforfourero will definitely get a hit at some point during that game -- but he only has a one in four chance of getting a hit when it matters. 1/1 * 1/4 = 1/4

Jack All or Custing, on the other hand only has a one in four chance of getting a hit during that game -- but if he gets a hit it will definitely come at the right time. 1/4 * 1/1 = 1/4

Both would then come through four times over the course of the season (1/4 * 1/10 * 162)

See? Same thing.

"It's for your own good. Big strong Devo knows whats best for Poppy" -- Mossback

by devo on Mar 25, 2008 3:00 PM PDT up reply actions 0 recs

I'll get back to this

When I'm more sober tomorrow, maybe during the game, I will get back to this. Biggest thing is that, while I'm very unsureof my hypothesis, you're still missing the variance point. Jack doesn't have a one-in-four chance of a hit during the game, because averages are not representative of the actual distribution.

I think you have to look at the day-by-day distribution. My idea is that you only have a chance once per game to be clutch. If you are a player that hits 4/4 every 4 days, then those first three at-bats don't mean anything toward a perception of clutch. On the other hand, in retrospect, a player who hits 1/4 every day has a chance every day to be clutch. Like I said, I think it only builds in a perception of a player being clutch, and not in an actual "clutch ability."

by dscel on Mar 25, 2008 10:17 PM PDT reply actions 0 recs

Clutch vs. "Unclutch"

I think it's very plausible that "clutch" doesn't exist but "unclutch" does. It's easier to perform worse (try too hard, press) than it is to perform better from a starting place of "best possible performance as the norm" - which you have to be to begin with in order to be a major league hitter.

I like Cindi. A. She never pretends to know more than she does. B. She has unbridled enthusiasm for her "Hotties," and isn't afraid to show it. -IM4Oakgal

by Nico on Mar 26, 2008 3:32 PM PDT reply actions 0 recs

which brings me back to Hendu

When asked what he was doing differently that allowed him to shine at critical times, he said (and I am paraphrasing here) "I am not doing anything different. They are."

Now granted, it seems silly to think that every time Hendu came through, it was because of the pticher, because wouldn't every hitter capitalize on a nervous or pressing pitcher?

I'm here to talk about the past.

by 67MARQUEZ on Mar 26, 2008 5:47 PM PDT up reply actions 0 recs