Jump to content

2022-23 Detroit Tigers Offseason Thread


chasfh

Recommended Posts

59 minutes ago, gehringer_2 said:

I would say the issue here is that to simply have some detectable level of effect is not a very high bar (this is also the bugaboo about drug trials). You could have enough additional success for you statistics to still show some effect, but how much would it be worth in terms of making enough difference matter given all the other drivers in a given game? I tend to think that in general, those things which have high the highest correlation are likely to be causative and the lower correlated effects are likely simply secondary correlates of those primary drivers. I think this logic holds in particular when you can at least posit the direct action of factor (e.g. a faster faster fastball required a batter with faster reflexes), as compared to more diffuse or generalized outcome measurements,

OK, so bottom line this for me: does it mean anything? Is it worthless? Can anything or nothing be taken away from this? Is there any point to any of this? Is the universe simply a series of random events? What is the meaning of life?

  • Haha 1
Link to comment
Share on other sites

1 hour ago, mtutiger said:

Another former Harris employee hits the streets.

I remember him from his time with the Cubs. His walk-up song was December 1963 (Oh What a Night). It must have been dad's favorite or something.

Tommy was good in his role, but he's never gotten much more than 300 plate appearances in a single season, and it would seem unwise to give him his first-ever full-time starting job at age 34.

I also remembering citing him as Exhibit A for just how crazy the juiced ball was in 2019, as in, if Tommy Fucking La Stella can jack 30 bombs in a year, something is seriously off. (He ended up with 16 that year because he got hurt.) I know I said so on the old site—I'd be linking that post here right were it available.

Tommy is owed $11.5 million by the Giants in 2023, as the tail-end of a 3/18.75. Talk about irrational exuberance!

Link to comment
Share on other sites

6 hours ago, chasfh said:

Finally got around to doing this, and the result is more surprising than I imagined it could be. Not because I guessed wrong correlations on the wrong attributes, but that correlations were stronger for those attributes I probably thought would be pretty weak.

I looked at the correlations of each of the 14 attributes on the Statcast cards of those pitchers who had values reported for both 2021 and 2022. As a reminder, Statcast card values are reported as a percentile versus the league—if you have the greatest strikeout percentage in the league, your percentile will be 100; if you have the worst, it will be 1; if you're dead balls average, you're a 50. Rinse and repeat thirteen more times.

Here are the results:

xwoba/xera       0.47
xba       0.57
xslg       0.52
xiso       0.49
xobp       0.42
brl_percent       0.38
exit_velocity       0.48
hard_hit_percent       0.37
k_percent       0.70
bb_percent       0.63
whiff_percent       0.73
fb_velocity       0.90
fb_spin       0.90
curve_spin       0.93

We could have guessed that the correlations are highest for attributes completely physically controlled by the pitcher: velocity and spin. We could also guess that outcomes more closely related to those physical attributes would also be pretty high: K%, BB%, Whiff %.

What I didn't expect to see were correlations so high for the attributes related to outcomes that are influenced by luck, which is basically everything else. Particularly xba and xiso, which you would think is a strong residue of BABIP. I didn't think the correlation would be zero or anything, but I imagined they would come out more in the 0.10 to 0.20 range than they would in the 0.37 to 0.57 range.

I guess I underestimated the effect that a pitcher's stuff would have on outcomes, specifically on overcoming oppositional force, and the luck of that force on as regular a basis as it does.

Good work.  What I'd be most interested in is how the statcast results correlate with performance stats such as FIP in the same year and future years.  Are they adding anything that is not already explained by BB%, K% and HR%?  

Link to comment
Share on other sites

21 minutes ago, Tiger337 said:

Good work.  What I'd be most interested in is how the statcast results correlate with performance stats such as FIP in the same year and future years.  Are they adding anything that is not already explained by BB%, K% and HR%?  

Just did a quick correlation for FIP, 2022 vs 2021, minimum 50 innings each season, n=218. Correlation is 0.42, significantly less than K% or BB% alone. HR/9 correlation is 0.35. Take from this what you will, but in terms of whether we need the three different measures when FIP contemplates all of them together, personally, I like knowing how components of an all-in-one number break down. I think it helps give me better understanding of the player involved.

I think of OPS the same way. Player A—let's call him "Jesse"—has a .344 on base and .344 slugging; Player B—let's call him "Marcell"—has a .274 on base and a .413 slugging. OPS would tell you they are practically the exact same type of hitter, but if you break it down to the components, you can see they arrived at their OPS in very different ways, and because of that, they are very different types of hitters.

Link to comment
Share on other sites

2 hours ago, chasfh said:

I remember him from his time with the Cubs. His walk-up song was December 1963 (Oh What a Night). It must have been dad's favorite or something.

Tommy was good in his role, but he's never gotten much more than 300 plate appearances in a single season, and it would seem unwise to give him his first-ever full-time starting job at age 34.

I also remembering citing him as Exhibit A for just how crazy the juiced ball was in 2019, as in, if Tommy Fucking La Stella can jack 30 bombs in a year, something is seriously off. (He ended up with 16 that year because he got hurt.) I know I said so on the old site—I'd be linking that post here right were it available.

Tommy is owed $11.5 million by the Giants in 2023, as the tail-end of a 3/18.75. Talk about irrational exuberance!

I’m in favor of letting him clear waivers and picking him up.  Checks a couple boxes, if healthy.

Link to comment
Share on other sites

2 hours ago, chasfh said:

Just did a quick correlation for FIP, 2022 vs 2021, minimum 50 innings each season, n=218. Correlation is 0.42, significantly less than K% or BB% alone. HR/9 correlation is 0.35. Take from this what you will, but in terms of whether we need the three different measures when FIP contemplates all of them together, personally, I like knowing how components of an all-in-one number break down. I think it helps give me better understanding of the player involved.

I think of OPS the same way. Player A—let's call him "Jesse"—has a .344 on base and .344 slugging; Player B—let's call him "Marcell"—has a .274 on base and a .413 slugging. OPS would tell you they are practically the exact same type of hitter, but if you break it down to the components, you can see they arrived at their OPS in very different ways, and because of that, they are very different types of hitters.

It's always good to look at the components.  However, my interest is in knowing whether the statcast numbers tell us more about run prevention than just looking at BB%, HR% and K%.  Maybe we could do a regression where runs allowed = all the individual stats and see if there is a combination that tells us significantly more about run prevention than BB%, HR% and K%.  

Link to comment
Share on other sites

2 hours ago, Edman85 said:

R^2 depends on the context... In the context we use in my line of work, .9-.95 is the general threshold for the specific application.

In physical science your are usually looking for direct causality, and that should have a very high correlation at least at across some range and linearization - otherwise you are probably just looking at the correlate of some more fundamental factor you don't know - or your data is just too noisy. In social science or human survey data it's almost the opposite, you may be looking for things that may only be weakly causative or that only a fraction of the population responds strongly to. Sports metrics are probably somewhere in the middle of that continuum - some things very directly causative (FB velo), some things maybe tied to other factors that are partially hidden - for instance maybe curveball spin matters more at particular spin axes. And then some things probably get into more psychologically dominated effects where some players may react and some don't at all so you get the average results of sub populations with high and low correlation respectively.

Link to comment
Share on other sites

14 hours ago, gehringer_2 said:

In physical science your are usually looking for direct causality, and that should have a very high correlation at least at across some range and linearization - otherwise you are probably just looking at the correlate of some more fundamental factor you don't know - or your data is just too noisy. In social science or human survey data it's almost the opposite, you may be looking for things that may only be weakly causative or that only a fraction of the population responds strongly to. Sports metrics are probably somewhere in the middle of that continuum - some things very directly causative (FB velo), some things maybe tied to other factors that are partially hidden - for instance maybe curveball spin matters more at particular spin axes. And then some things probably get into more psychologically dominated effects where some players may react and some don't at all so you get the average results of sub populations with high and low correlation respectively.

 

Yeah, in my case it is a short cut to see if we can use load spectra for adjacent finite elements for fatigue life predictions. Without a crazy high R^2, it's no good.

Link to comment
Share on other sites

1 hour ago, Edman85 said:

 

Yeah, in my case it is a short cut to see if we can use load spectra for adjacent finite elements for fatigue life predictions. Without a crazy high R^2, it's no good.

can you guys take this nonsense to the Math thread? 

You're getting in the way of our complaining about management's inactivity.

Link to comment
Share on other sites

57 minutes ago, Tenacious D said:

can you guys take this nonsense to the Math thread? 

You're getting in the way of our complaining about management's inactivity.

"Math Suks" -- Jimmy Buffett

If necessity is the mother of invention
Then I'd like to kill the guy who invented this
The numbers come together in some kind of a third dimension
A regular algebraic bliss.

Let's start with something simple, like one and one ain't three
Any two plus two will never get you five.
There are fractions in my subtraction and x don't equal y
But my homework is bound to multiply.

Math suks math suks
I'd like to burn this textbook, I hate this stuff so much.
Math suks math suks
Sometimes I think that I don't know that much
But math suks.

I got so bored with my homework, I turned on the TV.
The beauty contest winners were all smiling through their teeth.
Then they asked the new Miss America
Hey babe can you add up all those bucks?
She looked puzzled, then just said
"Math Suks".

Math suks math suks
You don't even have to spell it,
All you have to do is yell it...
Math suks math suks
Sometime times I think that I don't know that much
But math suks.

Geometry, trigonometry and if that don't tax your brain
There are numbers to big to be named
Numerical precision is a science with a mission
And I think it's gonna drive me insane.

Parents fighting with their children, and the Congress can't agree
Teachers and their students are all jousting constantly.
Management and labor keep rattling old sabers
Quacking like those Peabody ducks.

Math suks math suks
You don't even have to spell it,
All you have to do is yell it...
Math suks math suks
Sometime times I think that I don't know that much
But math suks.

Edited by casimir
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...