Clutch vs. choke - do these numbers mean anything?

[notin]

Hard to trust that guy since he was an escaped convict who made friends with a talking duck named Howard and ended up dying in Vietnam after being given an experimental drug by the Army.

 

But he was a Viking!!

[moonslav59]

I'm just glad you guys spelled Choke correctly.

[Kimmi]

Sorry, what anecdotal evidence did I provide?

 

The data that you posted is anecdotal because there is no statistical significance to it.

[Kimmi]

We will never resolve these things. The analytic people will deny the existence of clutch, hot/cold , momentum, etc. all day. But it seems that the players, managers, coaches, etc. all believe in these things. Who is right? I don't know.

 

For so many years, baseball people have accepted their beliefs as fact, mostly because that's the way it's always been. To be fair to the analytic people, they are not out to prove or disprove anything. They are out to find answers. If the players, managers, and coaches are right, the geeks would be more than happy to confirm those beliefs.

[Kimmi]

For sure , confidence and poise are important for success in anything. If doubt starts to creep in, you are headed for trouble in baseball or most anything else.

 

I 100% agree with this statement.

[Kimmi]

The Analytical Folks pour through tons and tons of data. But the players/coaches/managers have lived through it all, which should be more valuable.

 

When asked what the most useless stat was, Torii Hunter once said “whatever one tells you Derek Jeter is a bad defender.”

 

Needless to say, not all players are on board with analytics.

 

But then, some of their beliefs are also pretty out there. Wade Boggs thought eating chicken before every game made him a better hitter. Justin Moreau believed to be at his best, he had to leave his car at exactly 13 minutes after the hour. Nuke LaLoosh thought wearing women’s underwear made him a better pitcher.

 

So yes, I can see why some players’ wisdom might not be so sagely. And in the end, in a game dominated by math, I find myself trusting the math.

 

Except when it comes to Pat Tabler. That dude was a Clutch Freak Show…

 

Thinking that Jeter was a good defender is exactly why you have to trust the math.

 

When the math very strongly suggests that something is or isn't, I find it very hard to deny what the math is saying.

[Kimmi]

Also, the notion of "line up protection" is mostly false.

[moonslav59]

Also, the notion of "line up protection" is mostly false.

 

..and player comfort slots in line-ups.

[Bellhorn04]

The data that you posted is anecdotal because there is no statistical significance to it.

 

I didn't know data could be anecdotal. And I didn't draw any conclusions about the significance. You seem to just want to shut down the whole discussion.

[Bellhorn04]

In the clutch-choke debate, postseason numbers get rejected because they are not large enough sample sizes.

 

At least with the leverage numbers you get some higher sample sizes.

[Kimmi]

..and player comfort slots in line-ups.

 

Well, I don't agree with that statement.

[Kimmi]

I didn't know data could be anecdotal. And I didn't draw any conclusions about the significance. You seem to just want to shut down the whole discussion.

 

To me, data that has no statistical significance is anecdotal. I'm not at all trying to shut down the discussion. This is the type of discussion that I love.

 

My point is that people will always be able to show cases of where players had very clutch seasons. It is not a repeatable skill. Even with the players who had 'clutch careers', I'm guessing that their clutchness varied from season to season.

[Bellhorn04]

My point is that people will always be able to show cases of where players had very clutch seasons. It is not a repeatable skill. Even with the players who had 'clutch careers', I'm guessing that their clutchness varied from season to season.

 

Of course there's variance from season to season. Like there is with everything in baseball. I don't see how that's an argument.

 

So now you're saying that even if you have a big enough sample to show a 'clutch career', it means zilch.

Edited August 13, 2021 by Bellhorn04

[moonslav59]

Well, I don't agree with that statement.

 

I knew you wouldn't.

[moonslav59]

To me, data that has no statistical significance is anecdotal. I'm not at all trying to shut down the discussion. This is the type of discussion that I love.

 

My point is that people will always be able to show cases of where players had very clutch seasons. It is not a repeatable skill. Even with the players who had 'clutch careers', I'm guessing that their clutchness varied from season to season.

 

They did have varying seasons, and chokes like Bonds became clutch, and clutch like Beckett became chokes.

 

To me, it's the same as saying player X is great on Tuesdays but sucks on Fridays. It's a skill.

[Bellhorn04]

They did have varying seasons, and chokes like Bonds became clutch, and clutch like Beckett became chokes.

 

No, Beckett did not become a choke. He had a fine postseason career. In the 2008 postseason he was pretty obviously pitching hurt, but came through with a gutty performance in Game 6 of the ALCS.

[Kimmi]

Of course there's variance from season to season. Like there is with everything in baseball. I don't how that's an argument.

 

So now you're saying that even if you have a big enough sample to show a 'clutch career', it means zilch.

 

The variance is large enough that there is no correlation in the stat from season to season. If there is no correlation, then clutch or choke is not a repeatable skill. It's a random factor. A player might have had a clutch career, but not because he is a clutch hitter.

 

I looked at Fangraphs' "Clutch" stat on a season by season basis for each of the players that you mentioned in your opening post, except for Montgomery because I didn't know which Montgomery you were referring to. Also, I used the "Clutch" stat rather than BR's High/Medium/Low leverage because I could see all seasons on the same page (my laziness.)

 

Gallo was pretty consistently bad in clutch situations, except for 2020 in which he had a positive clutch stat.

 

For the other players, Judge, JD, and Ortiz, there is no rhyme or reason to their season by season data. They were unclutch just as often as they were clutch.

[notin]

In the clutch-choke debate, postseason numbers get rejected because they are not large enough sample sizes.

 

At least with the leverage numbers you get some higher sample sizes.

 

It’s not just the sample size with postseason numbers; it’s also that they are often spread out over a range of years.

 

Also they assume every at bat is some sort of important key at bat. When the Yankees are pummeling the Sox in game 3 of the 2003 ALCS, did you watch the hitter driving in the 19th run of that game and think “Nice clutch hit!!”?

[Kimmi]

They did have varying seasons, and chokes like Bonds became clutch, and clutch like Beckett became chokes.

 

To me, it's the same as saying player X is great on Tuesdays but sucks on Fridays. It's a skill.

 

Exactly. There are clutch moments. There are clutch games. There are clutch seasons, and maybe even clutch careers.

 

But 'clutch' is not a skill. It's pretty much random.

[Bellhorn04]

It’s not just the sample size with postseason numbers; it’s also that they are often spread out over a range of years.

 

That argument irks me too. To me the fact that David Ortiz was great in the World Series in 2004, and again in 2007, and again in 2013, should be a point in favor of the clutch premise...

[Kimmi]

That argument irks me too. To me the fact that David Ortiz was great in the World Series in 2004, and again in 2007, and again in 2013, should be a point in favor of the clutch premise...

 

The fact that David Ortiz is a great hitter, period, has a lot to do with his performance in the postseason.

[Bellhorn04]

The variance is large enough that there is no correlation in the stat from season to season. If there is no correlation, then clutch or choke is not a repeatable skill. It's a random factor. A player might have had a clutch career, but not because he is a clutch hitter.

 

And that's where I throw up my hands. Pretty hard to argue against that kind of logic, I guess.

[Bellhorn04]

The fact that David Ortiz is a great hitter, period, has a lot to do with his performance in the postseason.

 

Of course. But other great or very good hitters have not had good postseason numbers.

[moonslav59]

Of course. But other great or very good hitters have not had good postseason numbers.

 

That's totally expected.

 

Other great and very good hitters hit poorly from July 15th to August 10th.

[Bellhorn04]

That's totally expected.

 

Other great and very good hitters hit poorly from July 15th to August 10th.

 

Not a relevant argument.

[moonslav59]

Exactly. There are clutch moments. There are clutch games. There are clutch seasons, and maybe even clutch careers.

 

But 'clutch' is not a skill. It's pretty much random.

 

Yes, if you randomly fed a career .931 OPS guys numbers into a random generator, you have a few outputs like what Papi did, a few very bad and many around .930. The random samples would mirror reality, but with names attached not blank data.

[moonslav59]

Not a relevant argument.

 

Sure it is.

 

You are choosing October 3 to October 30th as your sample size and calling it a skill.

[Bellhorn04]

Sure it is.

 

You are choosing October 3 to October 30th as your sample size and calling it a skill.

 

No. The argument is that postseason is higher leverage.

[Kimmi]

And that's where I throw up my hands. Pretty hard to argue against that kind of logic, I guess.

 

How do you explain a player with good overall career clutch numbers (a supposedly clutch player) being below average in clutch in as many seasons as he was above average in clutch?

 

If there is no way to correlate or predict what a player is going to do from one season to the next in clutch moments, how can it be classified as a skill? It's just like BABIP.

[mvp 78]

How do you explain a player with good overall career clutch numbers (a supposedly clutch player) being below average in clutch in as many seasons as he was above average in clutch?

 

If there is no way to correlate or predict what a player is going to do from one season to the next in clutch moments, how can it be classified as a skill? It's just like BABIP.

 

Clutch is about how a fan feels as a player. Similar to how pitchers feel about throwing to certain catchers.