Bill James on "clutch hitting"

[RoseBowlMtg]

Last night on Homer's show, Homer asked James about clutch hitting. He said Ted Simmons over a 21 yr career hit .272 with no one on and .299 with runners on with an OPS 100+ points higher. Is that clutch? James went on to say he would have to look at it closer and he is reevaluating his take on clutch hitting. He went onto say that in the 70's he read an article by someone that got him into believing clutch hitting didn't exist. He now believes much more studying needs to be done on the subject.

 

Interesting coming from the guru on stats. Anyone else hear the interview?

[TooLiveBrew]

I didn't get to hear that - thanks for sharing. I'm kind of surprised the concept of "clutch" isn't yet more widely regarded as simply straying as little as possible from normal production. Sure, there are guys like Simmons who for whatever reason actually performed better (depending on how you calculate what "clutch" is defined by), but for the most part "clutch" simply means you don't suffer from the added stress of higher-leverage situations.

[logan82]

It seems like clutch would be almost impossible to measure until somebody's career was over. You don't get enough at bats in clutch situations, whatever those may be, to tell wether or not it is just random or a skill. At least that would be my assumption.

[TerraceReserved]

And that's the problem with James. He has no idea whats going on in sabermetrics. He had an article in the THT annual, and in it he stated the debate had ended and no studies have been done, despite extensive work on the subject. In The Book, they found clutch hitting exists, albeit in a small way. Tons of others studies have been done. And then of course he's written junk in the past claiming clutch is a joke and has said "we're supposed to believe (supposed clutch hitters) are better people." James was a pioneer in the field of statistics, but he's pretty irrevelant now.

[brewjihad]

The arguement frequently used on Fire Joe Morgan is probably the best way to 'dispute' clutch hitting. Basically, a hitter is trying as hard as possible to get a hit in any particular situation. It's not like Jeter tries harder to get a hit in the late innings than A-Rod. The best "clutch hitters" are also the "best hitters". I remember an arguement in the debate between Counsell and Cirillo- who would have been better as a utility/pinch hitter last year. Someone said that Cirillo was clutch and Counsell wasn't.

 

Cirillo 06: .319/ .369/ .414

 

Counsell 07: .220/ .323 /.309

[sbrylski]

Ted Simmons is one of the most extreme cases in difference between his situational stats over the course of his career. But there is a very, very small number of other players like that.

[TooLiveBrew]

Hey, just a tip-off/reminder to anyone interested, a piece on Bill James is the 'cover story' for 60 Minutes tonight.

[logan82]

Hey, just a tip-off/reminder to anyone interested, a piece on Bill James is the 'cover story' for 60 Minutes tonight.

Thanks for nothing dude. Nah, just givving you a hard time.http://forum.brewerfan.net/images/smilies/wink.gif It looks like you posted at 7:51, but 60 Minutes is on at 6:00 here.

[TooLiveBrew]

Whoops - well, ... "Hey west-coast Brewers fans, just a heads-up..."

 

http://forum.brewerfan.net/images/smilies/wink.gif

[rluzinski]

Yes, James is getting a bad habit of calling for more research to be done on a topic that is being continuously researched. People have never stopped looking for statistical evidence of cluth hitting and some have found evidence of a small effect. It's not going to make Jetter a better option than Arod when a clutch hit is needed, however.

[Playing Catch]

Does anyone have the statistics on "anti-clutch"? In the pressure situations I was in during High School, my knees were a-knockin'. But maybe athletes (especially world-class athletes) generally have the "fight" response rather than the "flight" response when the adrenaline is pumping.

[rluzinski]

I'm sure some major leaguers are anti-clutch to some degree. It's just tough to find statistical evidence of it when the anti-clutch isn't dramatic. I suppose that, in an AB in VERY important situation (in a clutch situation in the playoffs, for instance), a player could be terribly anti-clutch. You could just never prove it, statistically.

[Playing Catch]

Thanks rluz. I guess being able to play up to one's "normal" numbers has always been my definition of clutch.

[Leinie Lodge]

A basic "anti-clutch" formula might be:

(#K's with Runners On + #Popups with Runners On + #GIDP's) / total AB's with Runners On

 

Could be interesting ...

[owbc]

I've always had a hard time dealing with clutch.

 

I agree that there is no convincing statistical evidence for clutch hitting, and there probably never will be.

 

I'm not sure if I necessarily agree with the definition of at-bats that raise/lower a team's % chance of winning the game the most. If you're down by 2 in the bottom 9th with 2 outs and nobody on, you don't have a chance to win/lose the game, but the AB is extremely important for the win expectancy, but a HR isn't going to change the outcome of the game. The next batter comes up and hits another solo HR to tie the game. The third batter wins the game with a 3rd solo HR

 

The first batter raises the win expectancy by 3%.

The second batter raises it by 51%.

The third batter raises it by 45%.

 

I'm not a math expert, but I don't see how you could write a formula that weighs each of these HR in terms of "clutch" without creating a large amount of disagreement.

 

Personally, I like looking at numbers with 2 outs and runners in scoring position. It's a simple stat that is readily available and the majority of the time, it is an important AB during the game.