Page rendered in 0.8300 seconds
41 querie(s) executed
— Where BTF's Members Investigate the Grand Old Game
Tuesday, June 22, 2004
Painting a Fake Tunnel on a Blind Alley
Pitch counts and abuse of the research process.
I don’t have enough evidence to say that David Leonhardt is an idiot, but his article about pitch counts that appeared in the June 13th edition of the New York Times is another sad chapter in the disinformation campaign about pitcher abuse and pitch counts—-a situation that, over the past few years, can be summed up by a familiar visual metaphor: imagine, if you will, a dog who won’t let go of a bone no matter what.
Leonhardt’s article is full of so much contradictory information that it’s hard to take it seriously—-except for the fact that it has the unmistakable aroma of a counterattack on this controversial topic, one that has ignited a great deal of heat and very little light ever since it first appeared in 1998.
And the group of neo-sabermetricians that Leonhardt quotes exclusively on this subject—-those cheery fellows at the Baseball Prospectus—-really need a counterattack right now. Their research effort on pitch counts has recently been eviscerated by the key figure of sabermetrics, Bill James, in his definitive essay appearing in a new book, The Neyer/James Guide to Pitchers.
Leonhardt is conspicuously silent on this fact, despite his apparent rapt interest in the subject matter. The fundamental weakness of the research conducted by the two principals, Rany Jazayerli and Keith Woolner, is that it doesn’t predict injury or ineffectiveness better than anything else, up to and including the use of a divining rod. As James points out in his own studies that appear within his essay, the pitchers with greater overall workloads pitched far better in subsequent seasons than those with lesser workloads.
James concludes that the research is so flawed that it is virtually useless. He states this a bit more gently:
“We have driven a long way here down a blind alley, and we’re not going to fix it by backing up a few feet and trying again.”
In their rebuttal to James that appears in the book, the Prospectus folks circled the wagons (something that, more than anything, they are truly gifted at), stating that they had merely wished to “spur more research.”
Now what we see, with efforts such as Leonhardt’s, that rather than own up to the fact that the “pitcher abuse movement” was ill-conceived and ill-advised, a second front of “damage control” and spin is going to be aggressively put into place.
And that front actually began a few days earlier with Prospectus proxy Will Carroll, whose June 4th column at mlb.com studiously ignores James and tries to salvage some legitimacy for this pitch count hocus-pocus. Carroll focuses on other aspects of science to suggest a connection between biomechanics and pitch counts, something that, objectively speaking, is about as well-founded as the connection between Saddam Hussein and Al Qaeda. Despite the fact that James’ findings indicate that there is no connection between higher workloads and injury/performance, Carroll clings valiantly to the wreckage of an idea that was grounded more in fervor than in objective science.
Careerism Trumps Research
Nothing could be less surprising to those who have watched the sabermetric movement get hijacked by careerism over the past six years. It should be abundantly clear now to anyone with an ability to read wall scrawl that it isn’t research and objective knowledge that the Baseball Prospectus is interested in, its dollars. Their MBA-driven approach has proven successful in a world where rhetoric and carefully manipulated data are now becoming as much a part of the insider culture as has been the case in other industries.
Such an occurrence is, as we all know, by no means an unalloyed gain for the industries in question—-and the impact on baseball will be similarly problematic.
As with most things in the world of consulting, appearance is far more important than reality. And in that environment, how you approach the numbers is often designed to convey how you want reality to appear, as opposed to providing the most faithful representation of reality. With their work on pitch counts, the Prospectus has demonstrated that they have firmly grasped and applied the invidious lessons of more-or-less-professional consultation.
A Brief History of the “Pitch Count Crusade”
Of course, the first thing you do is use your friends. When Pitcher Abuse Points first appeared as an idea in 1998, ESPNet columnist and author Rob Neyer was lukewarm in his assessment. Somewhere over the course of the next nine months, though, Neyer was converted into the fold; when the 1999 edition of the Baseball Prospectus appeared, he suddenly trumpeted the original PAP (their choice of acronyms, mind you, not mine) as a turning point in the battle against AIDS, world hunger, and the cross-wire brassiere. Neyer then embarked on what amounted to a two-year witch hunt on managers who dared to give pitchers more than 120 pitches in a single start, despite the fact that there was no clear evidence that occasionally hitting this number was actually dangerous.
A great deal of opposition to this simplistic approach was raised, including my own in 1999, which was burned to a crisp by the “prairie fire” of the emerging neo-sabe hegemony. In it, I dared to question the motives of the creator of PAP, suggesting that he had created it more to create publicity than to foster research. This was met with the indignant howls of a movement that had blood in its nostrils, and the “pitch count police” became even more aggressively entrenched as a result. Neyer, who is actually a very good researcher, reached back into history to discover that Orioles manager Paul Richards had been obsessively conservative with his young pitchers during the late 1950s, which was an interesting but utterly inconclusive finding (the kind that occasion much of what passes for real research in our current “neo-sabe” daze).
Leonhardt, of course, fails to make this connection between Richards’ handling of young pitchers and the current experiment of limiting pitch counts in the lower levels of the Cincinnati Reds’ farm system. He would apparently prefer you to think that these are new and original ideas. They’re not.
But Bill James has a far more considered view of the issue that is also pertinent here. He suggests that too much coddling of arms may produce unanticipated problems later on:
“Backing away from the pitcher’s limits too far doesn’t make a pitcher less vulnerable; it makes him more vulnerable. And pushing the envelope, while it may lead to a catastrophic event, is more likely to enhance the pitcher’s durability than to destroy it.”
In the face of such statements, and in ignorance of other strains of thought regarding this complex issue, the Prospectus cadre and Leonhardt spin the facts to show that what is quite possibly a needless exercise in caution is “cutting edge research being put into practice for a better, safer tomorrow.”
That’s what I’d call the “PR approach to baseball research,” which is a sure sign that you are dealing with a neo-sabe. Despite the backlash engendered against my 1999 critique of PAP, it was necessary for the Prospectus cadre to revise the method in 2001. However, in doing so, the original impetus was derailed: instead of focusing on young pitchers, an attempt to focus on overall workload limits took precedence. The study, as conducted by Keith Woolner, took on the patina of complexity but ultimately became muddled. In the coy fashion that is the apparent signature of Woolnerian research, the conclusion circled back to the idea that 120 pitches (actually, 122: go figure) was dangerous at any time and any place, and a new measure (called “stress”) was connected to this assumption.
“Spurring Research” The PR Way: Ignore It
In the meantime, other research (the work that Jazayerli claimed they had hoped to spur) was simply ignored. One of the key fallacies in the original PAP was that high single-game pitch counts were inherently catastrophic; this fallacy is repeated almost verbatim in Leonhardt’s article when he states that a pitcher throwing 130 pitches in one game and 70 in the next has worked harder than one who threw 100 pitches in each game.
The notion of workload clustering—-an alternative hypothesis that identified several problematic practices with greater potential to produce injury and ineffectiveness—-was overlooked by the Prospectus team of “PR researchers” (though not by Neyer, who was by mid-2001 slowly distancing himself from the issue).
Other evidence of the non-catastrophic effect of occasional high single-game pitch counts was provided by the Allan Roth Dodger dataset, a compilation of detailed game logs from 1947-64 that the great statistician had kept, which included pitch count information. The treatment of the Roth data at the hands of Jazayerli and Woolner was almost comical in its obtuseness; they simply missed an entire clutch of potential insights in their zeal to crunch the pitch count numbers.
What was most striking about the data was that it showed how pitchers who pitched well had an inherent “economy of scale” in their pitch counts, while those who struggled did not. (This tendency has become more pronounced in recent times, as offense has both improved and changed its shape.) Simply put, it appears that the number of pitches a starter can throw on any given day without ill effects is more strongly related to the quality of their performance than has been previously considered.
The data also showed a primitive ecology for pitchers that often worked as a natural barrier to the most dangerous type of overuse—a long consecutive string of 100+-pitch outings, which recent research (noted by Neyer, but bypassed by Prospectus) has shown to produce injury in close to 50% of its incidences. That type of usage pattern is relatively rare today (and was so in 1998, when pitch counts first became a cause celebre), but it was even rarer in the 1950s, because pitchers who struggled in the early innings of a ballgame were simply removed. Despite the fact that pitching staffs were smaller, managers used their long men earlier and more often; they spread out their innings (and pitches) more in this fashion, and the occasional low-count, short-inning appearance prevented starting pitchers from accumulating long strings of moderate-to-high pitch count games.
What’s important to realize is that none of these ideas can possibly explain the dynamics of workload and injury (or, as James puts it in the title of his essay, “Abuse and Durability”). They are merely signposts for a complex process that is being obscured by oversimplified data and emotionally-charged “crusaderism.” By continuing to focus on the overly cautious approach, the Prospectus cadre and Leonhardt give us only one side of the story, much in the same way that the current administration and the media give us a slanted view of the “twin wars.”
As is often the case, Bill James sums up this unfair and unbalanced approach quite well:
“What I believe has happened is that this balancing of risks has just gone completely haywire. Two things have happened to cause it to go haywire:
1) We have introduced new information—-pitch counts—-into a previously organized way of thinking about the issue, and 2) We have focused undue attention on the risk of chronic overuse, which in reality is merely one of many potential problems for a pitcher.
There is a natural balancing of risks, between avoiding a catastrophic event, and developing a tolerance for more work. In order to develop a real understanding of this issue, we are going to have to take account of both ends of the injury spectrum.”
One notes that James is careful not to suggest that the Prospectus cadre had less-than-pure research motives at heart, which was one of the things that got me into so much trouble (and will doubtless do so again). I read this as a kindly mentor’s attempt to extricate a favorite pupil from an embarrassing episode where feeling and fervor clearly trumped good sense. The fact that Leonhardt is now trying to spin this disaster of misguided research into “the wave of the future” should indicate that the Prospectus plans to keep its wagons as tightly circled as ever. Just like George Bush Junior, they can’t bring themselves to admit that they were wrong, or that their motives were impure. And, as noted, it’s not just Leonhardt; it’s the Prospectus family itself, with Will Carroll claiming hazard duty pay as Spin Doctor.
So let’s not be hoodwinked into letting these guys paint a fake tunnel on a blind alley. That kind of escape attempt only works in cartoons. But, then again, much of neo-sabermetrics is cartoon-like, so maybe we shouldn’t be surprised by such a tactic. And with spin from a source as high-toned as the New York Times, perhaps our pesky “pitch count paranoids” will be able to emulate kindred spirit Wile E. Coyote and fall from the 500-ft. mesa to the valley floor below with only a puff of smoke to mark the spot. Teflon, anyone?
You must be logged in to view your Bookmarks.
Loser Scores 2015
(12 - 2:28pm, Nov 17)
Loser Scores 2014
(8 - 2:36pm, Nov 15)
Winning Pitcher: Bumgarner....er, Affeldt
(43 - 8:29am, Nov 05)
Last: ERROR---Jolly Old St. Nick
What do you do with Deacon White?
(17 - 12:12pm, Dec 23)
Last: Alex King
(15 - 12:05am, Oct 18)
Nine (Year) Men Out: Free El Duque!
(67 - 10:46am, May 09)
Who is Shyam Das?
(4 - 7:52pm, Feb 23)
Last: RoyalsRetro (AG#1F)
Greg Spira, RIP
(45 - 9:22pm, Jan 09)
Last: Jonathan Spira
Northern California Symposium on Statistics and Operations Research in Sports, October 16, 2010
(5 - 12:50am, Sep 18)
Mike Morgan, the Nexus of the Baseball Universe?
(37 - 12:33pm, Jun 23)
Last: The Keith Law Blog Blah Blah (battlekow)
Sabermetrics, Scouting, and the Science of Baseball – May 21 and 22, 2011
(2 - 8:03pm, May 16)
Last: Diamond Research
Retrosheet Semi-Annual Site Update!
(4 - 3:07pm, Nov 18)
What Might Work in the World Series, 2010 Edition
(5 - 2:27pm, Nov 12)
Last: fra paolo
Predicting the 2010 Playoffs
(11 - 5:21pm, Oct 20)
SABR 40: Impressions of a First-Time Attendee
(5 - 11:12pm, Aug 19)
Last: Joe Bivens, Floundering Pumpkin