See also Gil Meche POTD and Gil Meche's Deception.

…………………….. 

On this webpage there is an interesting debate:  will Gil Meche sustain his success?  After you scan this, you'll see why the debate centers around the idea of HR/FB ratio.

There are two different ways to use pitching "luck" indicators like H%, HR/FB ratio, 1-LOB%, and all the rest:

1) Dr. D:  Use "luck indicators" as "metal detectors" to find candidates for your list of "pitchers who may have been pitching in good luck."  Then get to work on your list to identify which pitchers actually were lucky, and which were not.

2) USSM:  assume that any deflection from league average on "luck indicators" — especially on HR/FB ratio — is due to nothing other than luck, and that any deflection must quickly "regress to the mean." 

USSM's "baloney call" on Gil Meche, is interesting, and they may be correct. They advise flushing him, quick, if any GM is loopy enough to offer a good prospect. The soundness of their analysis stands or falls with the idea that Meche's low HR totals are due to nothing that he has done. 

The premise is that once the ball is up in the air, it will go over the fence at the same rates for all pitchers.  If you're lower than HR/FB = 11%, then it cannot be because you're keeping hitters off balance.  It is because luck is running your way. 

……………………….

First of all, this is a difficult debate.  We found this idea fascinating and asked guys who are the best in the sabr biz, inside ML orgs and out, and their responses were basically, "we don't know much about HR/FB ratio."  So we're not trying to bust anybody's chops for not having mastered this issue.  Baseball hasn't mastered this issue.

I personally think that the BABIP craze was far too dogmatic at the start.  I also believe that analysts have gone way overboard in their search for "corollaries" to BABIP that can be used with confidence to assess MLB pitchers without even watching them pitch.  (USSM believes in tools scouting, but I didn't see anything in that particular argument that couldn't have been deduced from a monitor only.)

Ever since AVG = (AB-HR-K)*0.3 was discovered as a general guideline, it has been all the rage to assume that a pitcher controls nothing other than whether the ball is in the strike zone.

Does that sound a little fishy to you?  To me too.  All thigh-high fastballs are not created equal, LOL.

………………………… 

Well, for sure Meche has been lucky — his ERA is 1+ his last six starts.  Obviously he can't sustain what he has done the last 6-7 starts.  Nobody could.

And I sympathize with folks who are so ANGRY at Gil Meche that it would take a thorazine drip to get them to calm down and consider the idea that Gil Meche Finally Arrived.  That's 95% of the blog-o-sphere.

…………………………. 

But you don't have to break down Meche's starts as "first 11 vs last 7" … you should actually review his "first 3 vs next 15." 

Since his 3rd start:

7-4, 3.41 ERA (would be #11 in the AL, ahead of Bonderman, Schilling, and Buehrle) 

95 IP

80 H

80 K

38 BB

10 HR

7.6 K, 3.7 BB, 0.9 HR

He has been fanning 1 batter for every 1 who got a hit — all year, except his first 3 starts. 

I'm not saying Meche is a hero.  But there is no question that he is a different pitcher than he has been in the past — and that has been the case all year long.  See the June POTD, where in mid-June we opined that Meche had probably turned the corner.  Since then, most people have come to believe that he has turned some kind of corner.

.

=== Rubber Meets The Road Dept. ===

Is HR/FB = 0.11 a guideline?  Or an absolute?

. 

Pedro Martinez for his career has a HR/FB rate of 8.3% — that is a full 25% below the league average, despite pitching most of his life in Fenway Park.

In some years in his prime, Pedro allowed HR's on fewer than 6% of his fly balls — despite Fenway.

Pedro is a great pitcher.  But remember the premise here — that once a batted ball in the major leagues goes up in the air, it is NOTHING other than luck (and park) that determines whether it will be an HR.

Pedro apparently refutes this idea, at least in its absolute version.  Pedro Martinez radically affected the percentage of fly balls that went for HR's.

……………………

Hitters routinely have seasons in which they have 25% or 30% of their fly balls go for home runs — Mark Teixeira is an example of this.

Granted, hitters are not pitchers.  But the implications are clear.  If HR/FB = 0.11 were an absolute, hitters would show much more "regression to the mean" than they actually do.

Some hitters hit THREE TIMES the league average on HR/FB.  This should suggest to us, intuitively, that HR/FB = 0.11 is a guideline, not an absolute. 

Further, we know that certain styles of pitchers are more effective against power hitters.  Merely excelling against the Teixeiras of baseball would necessarily affect your HR rate.

………………….. 

Dontrelle Willis, the 2nd guy I checked after Pedro, has a career HR/FB rate of 7.6% — which is 31% lower than the league's HR/FB rate — and his HR rate is about the same road and home.

……………………

Barry Zito, the 3rd guy I checked, has a career HR/FB rate of 7.9% — despite generally giving up more HR at home than on the road.

Certainly Zito is an example of a flyball pitcher, with unremarkable K rates, who maintains low HR rates.  His "change speed" factor — the deception on his curve — keeps hitters off balance.

……………………

That's not a study, but it's suggestive.  Pedro hasn't deflected from HR/FB=0.11 by accident.  I'm guessing you could find many pitchers who deflect from 0.11 signficantly, over the courses of their careers — and that you could find pitchers who deflect hugely from 0.11 over a period of 3-to-5 prime years.

.

=== Dr's Prognosis Dept. ===

The idea of "HR/FB = 0.11 as an absolute" deserves further study.  I wasn't able to find any BP or THT articles that study its persistence from year to year.  (Certainly the Pedro, Willis, and Zito examples are very suggestive.)

On an intuitive level, I find it incredible to think that there is no such thing as a pitcher who keeps hitters off balance.  I also find it incredible to think that there is no such thing as a pitcher who makes lots of mistakes and has gopheritis.

I also find it strange to analyze the 2006 Gil Meche in terms of the randomness of his batted balls in play, as opposed to attending to the FACT that hitters are now having trouble reading his pitches.

But I could be wrong.  HR/FB ratio is a metric that still is not well understood, that I can tell.  I'm not saying we've proven anything here.  We are only saying that our intuition tells us that HR/FB ratio should be treated as a (LOOSE) guideline rather than an absolute.

And we're saying that pitching analysis should not start with and pivot around "luck indicators" on batted balls.  Such a method seems shaky to me, but that's just my opinon.

……………………..

In the meantime, you know where we're at.  detectovision.com advises BUY on Gil Meche.

Cheers,

Dr D