Hey lads. Can’t believe the coke Girardio’s been taking from the msm boys. Well yes I can. I’m feeling very confident about game 6. Brainnnnnnnnnnnz
I’m watching Diary of the Dead. I kind of like it. Waiting for a crazed guy wearing # 55 to jump out.
Hey B.

As soon as you start talking about “AJ-like guys” you’ve left the realm of objective analysis and entered the realm of speculation.

maybe you (someone) could see if certain pitchers consistently have higher variance in some metric, say tRA or FIP or xFIP.

or I think baseball reference or somebody has a “game score” metric for pitchers.  That would be useful since it is meant to be used on a per game basis.

Well, I’m not saying that your “someone” is going to be me, but for the sake of argument, how should someone define “higher variance”?  And what’s the right sort of metric to use for a game by game analysis?  I wouldn’t think FIP, because it might tend to be too easy on a guy giving up lots of line drive hits in a given game.

I mentioned tRA because it counts more events than FIP.  By higher variance I mean look at the standard deviation per pitcher.  Maybe separate the pitchers into quartiles and see if what quartile they are in for month A is independent from month B.  I’m not sure how rigorous that method is though.

So so Curb, mez a mez Dexter. Middling night.
Pin- T the K is on the way.
Lucky Owlsfoot and all. Crow-bar in hand. New gaskets and an Alpine head unit in the squirrel sack.
VCP will be dialed in like a John Force funny car.

SG,

Regarding the 5% hit on short rest numbers; Did you come up with that through your own analysis or was that taken from Sean Forman’s recent study on rest performance in the post season since 1969?

Today I read chart here that shows 3 day rest pitchers take a major hit since the expansion to three rounds in 1995.

I know normally we would go with the larger sample size but weren’t the 70’s a major outlier in ML history in regards to pitchers going on 3 days rest? Would that be enough to skew Sean’s study?

Intuitively, I’m thinking pitcher usage today has changed dramatically from ‘69 until let’s say the mid 80’s, and that the second study might have more relevance to usage patterns today. Then again, I could be wrong.

Thanks for any thoughts on this.

Doomsday looms for the Phillies.

“As soon as you start talking about “AJ-like guys” you’ve left the realm of objective analysis”

This translates to, “If I don’t like the argument, it’s stupid”.

“rampant in some quarters, that there was no chance of bad AJ showing up on five days rest”

Well, I can argue against the Santa Claus hypothesis too, but there’s a conversation going on here about stuff we’re actually interested in.

“Also, if a guy has a 20% clunker propensity and short rest increases that by 20%”

Do I need to write down the algebra for you?  You have to come back with, The clunker rate can’t be worse than x for this, so his expected ERA can’t be worse than y.  Well, if you don’t want to just beg the question.

Today I read chart here that shows 3 day rest pitchers take a major hit since the expansion to three rounds in 1995.

I’m not disputing the chart, but holy crap, the article the chart was attached to was SOOO bad.

“Lee was good enough to beat the Yankees”

What does that even MEAN? He gave up FIVE runs!!

Yeah, but he got the Win, which Joe Morgan tells me is all that matters.

[101] If I ever make a zombie flik, Matsui will be in it. He won’t understand why, but he’ll be in it.

I’m not disputing the chart, but holy crap, the article the chart was attached to was SOOO bad.

Yeah, disregard the commentary. The chart did give me something to think about though. There is also no comparison numbers on pitchers who’ve had exactly four days rest and those who’ve had 5+, but I just assumed they are better than what’s posted because those are pretty bad.

Regarding the 5% hit on short rest numbers; Did you come up with that through your own analysis or was that taken from Sean Forman’s recent study on rest performance in the post season since 1969?

Nope, neither.  I took it from Chapter 7 of The Book (Playing the Percentages in Baseball).  I don’t have it in front of me, but I believe they looked at pitchers from 1999-2002 and compared all the pitchers who pitched on three days rest to how they did on four and five.  In terms of wOBA allowed, they went from something like .342 on four or five days rest to .359 or so(going off memory, so those may not be exact, but I know there was a five percent difference).

I didn’t see that Forman study until just now, but it seems to back up the data from <u>The Book.</u>

Re: Cano, I think he has excellent strike-zone judgement.  It’s also that he believes there are several balls that are out of the zone that he can hit with authority.  And for the most part, he’s right.  What Cano *doesn’t* do very often is take strikes.  E.g. the pitcher throws an 0-0 pitch on the corner at the knees, Cano is likely to swing.  When he’s going well, he’ll lace a LD into the gap.  When he isn’t going well, they’ll be popouts or weak GB.

He often gets compared to Soriano, but I don’t think that is correct.  Soriano would literally swing at ANYTHING.  Cano won’t.  He doesn’t often swing at pitches way out of the zone.  At least no more often than Teix or Swish (patient hitters) do.

And yes bebop, I understand you saying “never” was a bit of hyperbole.  I was calling you on it.  Perhaps with a bit too much snark.

[109] Wow.  Your responses here are all completely out of touch with anything I was trying to convey.  I must have really expressed myself exceptionally poorly on this.  Sorry.

WRT “AJ-like guys” my point is that there is a definitional aspect to this sort of thing that is always going to be subjective.  Trying to quantify the qualitative, as it were.  Do you disagree?

As for the “AJ could have blown up on extra rest, too” point; what’s your actual disagreement here?  It seems to me that it goes without saying, as we have all seen.  It’s got nothing to do with Santa Claus, and it has been ignored in much of the discussion, both by first- and second-guessers.  I was not trying to imply that you personally had made that argument.  If that’s what annoyed you, then my apologies again.

Finally, 0.2 x 0.2 = 0.04.  Is that what you’re looking for?  It’s arithmetic, not algebra.  I didn’t think it was worth writing that down because I don’t see much actual value in multiplying a number pulled out of your ass by a number pulled out of my ass.  The point is simply that when you multiply two fractions you get a smaller fraction.  I didn’t think I needed to do a demonstration.

WRT “AJ-like guys” my point is that there is a definitional aspect to this sort of thing that is always going to be subjective.  Trying to quantify the qualitative, as it were.

My little test wouldn’t be able to tell you who the AJ-guys are, but it might say who they probably are, and if they could exist.

[116] Ok, sorry to have overreacted, but several things about your post irked me.  You’re arguing a strawman - the question around here has been, what sort of arguments can one sensibly re AJ’s start, and you’ve been pushing something irrelevant.  Of course AJ could have a bad start on any amount of rest - but that’s in fact true of _any_ pitcher.  It’s true of Andy tonight and CC if need be.  What’s interesting is whether the likelihood of a blowup moves the estimated Phillies’ win % from SG‘s 10.2 w/AJ below his 10.7 w/Gaudin, or revised estimates of the latter which depend on stuff some might call unobjective.

The equation is Yankees’ budget = y = total now - guys leaving + CC + AJ + x, where x = max available for Teix.  Clearly depending on how much y is, x can be large.

Re AJ-like, of course, but everything done here is susceptible to this stance.  “Like” is just shorthand for some metric, which you might find useful or not.  Saying the metric isn’t “objective” before you’ve even seen it strikes me as, hmm, irksome.

Saying the metric isn’t “objective” before you’ve even seen it strikes me as, hmm, irksome.

Well, if that’s the case then you still aren’t understanding me.  I did not intend to argue that a hypothetical metric which had yet to be proffered was not objective.  I intended to argue that the hypothetical effort to use some yet-to-be-proffered metric to define “AJ-like” was so likely to be subject to selection bias and confirmation bias as to lack sufficient objectivity.  Of course, I may very well be wrong in making that argument as well, but I’d at least prefer to be shown to be wrong about what I’m actually arguing.

And I don’t see how the point about Burnett’s propensity for blowing up on regular or extended rest can be dismissed as irrelevant now that we’re in the post hoc phase and the argument has been put forth that his poor performance on Monday was directly attributable to short rest.  He pitched poorly in game 5 of the ALCS on regular rest.  How is that a straw-man?  Perhaps the problem is that you are taking my general responses to arguments in the MSM and elsewhere in the blogosphere as specific responses to posts made here.  So I’ll try to be clearer about that sort of thing in the future.

“And I don’t see how the point about Burnett’s propensity for blowing up on regular or extended rest can be dismissed as irrelevant now that we’re in the post hoc phase and the argument has been put forth that his poor performance on Monday was directly attributable to short rest.”

The point is that for all we know short rest may make Burnett much more likely (e.g., twice) to blow up.  I thought before that that was a risk - probably not one outweighed by the risk that Gaudin would be rusty etc., but maybe enough to argue for Marte et al - also the consequent ripples in the rotation might be beneficial.  That’s all.  It would be interesting to see if one can reasonably estimate or anyway limit the blow-up risk - for next year’s playoffs, for example.  Maybe Girardi should have considered the umpire on Monday has a tight zone, maybe AJ was a total slam-dunk decision.  Maybe that 5% estimate suffers from an obvious selection bias that makes it inapplicable.  I don’t care either way, I just don’t think it’s reasonable to dismiss the discussion a priori.  Not looking when one thinks one knows the answer is also biasing.

What about believing that the answer is not knowable?  Is that also biasing?  Or is the effort to find a statistical justification for not using a Burnett-type pitcher on short rest tantamount for looking for your keys where the light is better instead of where you dropped them?

I thought the Burnett blowup was a risk too.  I just didn’t think it was markedly increased by the rest difference.  The ripple effect on Pettitte is potentially real as well, but in any case the baseline for comparison is Gaudin, and the risk there is much more than rust.  His track record shows that even in mid-season regular work form, he is highly likely to struggle mightily against a good lefty-heavy lineup.  As I’ve said before, I was not in love with the short-rest strategy.  I just preferred it to the actually available alternatives.  I’m pretty sure that I was the first one here to bring up the bullpen game approach, but we all know that was never going to happen.

“What about believing that the answer is not knowable?”

That’s perfectly sensible, but of course then one can’t claim to put AJ in or not for rational reasons.  I suspect that you instead have a model which estimates the risk of A as < n% at some confidence level and the risk of B as > m% and the weighting factor as k +/- l% so you say let’s put in AJ.  If you won’t examine n, m, etc. though you end up in the other case.  This is the same thing as signing (oops, “buying”) Teix.  Either one doesn’t know, and takes that seriously, or one tries to know as best one can.

I don’t think there’s much basis for comparison between what should go into pitcher deployment decisions and our wild ass guesses about what the Yankees payroll parameters are.

I also disagree that the belief that certain things are unknowable is incompatible with being rational.  One can rationally consider all the things that one believes to know and make a decision on that basis while conceding that he may very well be wrong because of the things he doesn’t know.

But before the season your stance wasn’t (as far as I could tell), “we don’t know if the Yankees can afford Teix” (my stance), it was, “they can’t afford him, full stop, no point in looking at the parameters”.  And I had the impression that your criticism of the AJ-blew-up-entirely-due-to-resters, if there are such people, wasn’t that we don’t know how much the marginal odds were it would happen or that the non-negligible marginal odds didn’t outweigh the Gaudin risks.  Maybe I’m just not reading you right, but I don’t hear what seems like the full skeptical stance towards your own preferred position implied by what appears to be our shared view of knowledge.