It’s hard to be humble ...by Shane Tourtellotte
February 06, 2013
It's an exciting time for baseball writers, and readers. Pitchers and catchers report to spring training just days from now, bringing a climax to hot-stove speculations. It's also awards season, something a little new for us. SABR has inaugurated its Analytic Research Awards, honoring the best in commentary and analysis for 2012. You've seen the links to the award information all over our site and others, and if you've been smart or at least fortunate, you've followed them, read some or all of the finalists, and maybe voted already.
Speaking of being smart or at least fortunate ... I have personal reasons for highlighting this. I am one of the nominees in the Historical Analysis/Commentary category, honored for "And That Ain't All, He Stole Home!," the first sabermetric piece I had published at The Hardball Times or anywhere else. (If nothing else, this demonstrates that there is such a thing as beginner's luck.)
This brought me quite a gush of pride, both for myself and for THT: No other publication had more nominations. Pride, though, is a dangerous thing. It's like shrubbery, needing vigorous pruning to keep it in its proper dimensions. I've had this feeling a couple times before, early in my science-fiction writing career when I got a couple of award nominations. Back then, I learned the best method for forestalling a swelled head, and I used it again here. I read the other nominees.
It's no surprise to me, and I hope not to you, that there is a lot of good sabermetrically inclined baseball writing being done today. We as a whole can feel justifiable pride in the strength and growth of the field. Yet one can get a little too fond of the numbers we produce, the meanings we attribute to them, and the presumed permanence of what we discover.
Fortunately, some of us have this in mind already. Two of the nominated pieces this year flow from an awareness of the limitations of sabermetrics, a humility—in one case a surprisingly personal one—about what the numbers can and cannot tell us. They were just the needed corrective for someone briefly feeling awfully, awfully good about himself.
The first was Dave Cameron's "Why I'm Not a Fan of Losing on Purpose," dealing with Jeffrey Loria's fire sale of the Miami Marlins' big player acquisitions after one poor season. A plausible case can be made that blowing up the roster and starting over is a smarter move than staying committed to a mediocre lineup that never really contends; 80 wins a year doesn't get you the playoffs, and it doesn't get you the big boost in team revenues for marginal wins that kicks in when you start seriously contending for a playoff spot.
Cameron counters (and I am seriously compressing, so go read what he says yourself) that projections of the wins a team will have are uncertain enough, their error bars wide enough, that presumed mediocrity has more value than it appears. The systems may miss something, or you'll catch a few breaks, and suddenly you're not just contending, you're excelling. Cameron notes the 2012 Oakland A's as an example of an apparently not-so-good team confounding the forecasts to its great benefit. He could have cited the Baltimore Orioles to similar effect.
It's not a surprising point when you take time to think about it. Good sabermetricians always look for a confidence level higher than 51 percent before declaring something to be gospel truth. But it is refreshing to see someone as respected as Cameron not only be modest about what sabermetrics can do, but to show how that modesty itself can intelligently inform the actions of a major league team—even if it didn't this time.
The second piece was by Russell Carleton, returning to Baseball Prospectus after two years working with the Cleveland Indians. You could forgive someone who got to work in the bigs on the strength of his number-crunching for thinking sabermetrics the be-all and end-all. But Carleton doesn't need that forgiveness. His thoughts are quite different, in an essay he provocatively titles "Hire Joe Morgan."
Morgan, the one-time color commentator for ESPN baseball, is a stand-in for the whole "old school": the ex-players who cast jaundiced eyes on the number geeks, the coaches who trust what they see to the exclusion of what the computer printouts say, the Oakland A's scouts Brad Pitt mocks in Moneyball. You know, morons. They're everything sabermetrics is supposed to be superseding. They do it wrong, so Carleton and his companions have to be right. Right?
Except that he found that he wasn't. He discovered he had his own blind spots in creating statistical models. He learned that things he had taken for granted, some of them cornerstones of sabermetric doctrine, were not true. He even began to suspect that the old "eyeball test" might reveal things to a baseball lifer that Carleton didn't know existed. The conclusion of the piece (which again I'm compressing and which again you should give your own eyeball test) is a sincere apology to Joe Morgan, for responding to Morgan's narrow certainties with a narrow certainty of his own.
Mind you, he isn't saying Morgan and company are right: He isn't being that humble. Not having all the answers yourself does not mean some antagonist has all the rest. Even where they do have a point or three, they might never be methodical enough to comb out the right stuff from all the other stuff they think and say. Anyway, Carleton is speaking more to his own attitudes than to theirs.
His bit of introspection brings to my mind a piece of rarefied mathematics-cum-philosophy, Godel's Incompleteness Theorem. I'll skip the complex underpinnings (which I can barely follow myself) and cut to a boiled-down core. Mathematician Kurt Godel determined that there is no mathematical system that can prove all mathematical truths: No single theory can cover everything.
If you'll excuse a paragraph of high-flown speculation, it's my belief that Godel's principle covers just about every complex system, including the universe itself. For roughly the last century, we've had two theories, general relativity and quantum mechanics, that explain the operations of the universe on different scales but just don't mesh with each other. Scientists from Einstein on have spent generations trying to combine the two into a Theory of Everything, and have not managed it. I suspect they never will, not because of their stupidity but because of the underlying nature of reality.
I think this also applies to the wonderfully complex system that is baseball. Godel's theorem leads me to believe that there is no one system of thinking about baseball that will be able to answer all the questions we have about it. The numbers do not, and cannot, cover everything, just as the old-school methods do not and cannot. That doesn't imply that every faction has an equal portion of the truth, just that nobody's got it all. And maybe we should admit that the other guy may have part of it, and try to figure out what part it is.
There's another reason for us sabermetricians (if I may count myself in that group) to show some restrain in our belief in ourselves. Ironically, it's something thrown into sharp relief by the building flood of information available to us, from old play-by-plays to the various F/X systems producing reams of data on every movement of the ball and swing of the bat. It's that for all we do know, there is so much else that we don't—or worse, think we know that may get overturned in a couple years.
Carleton already encountered the latter, in those proprietary things he learned with the Indians that he isn't at liberty to reveal to us. When we're being honest and attentive, and a little modest, other writers will catch this stuff, too. At other times, we'll get bound up by what we think are certainties, and pay a price.
I'm drawn to the work Baseball Prospectus has done on the most controversial baseball matter of our era: steroids. In its book Baseball Between the Numbers, published in 2006, BP concluded that the effect of PEDs on player performance is small, even statistically insignificant for pitchers. The staff of BP seems to have taken this to heart in judging recent PED users: an in-house survey had more than 90 percent of them supporting Barry Bonds and Roger Clemens for the Hall of Fame. Why should they be penalized for cheating when that cheating didn't really matter?
In the very same book, BP found there was no evidence that certain catchers could produce more strike calls through how they framed their pitches.
They've certainly changed their minds on that subject. It's been BP writers like Dan Turkenkopf, Mike Fast, and Max Marchi* who have uncovered that evidence and determined the sometimes massive effect that pitch framing can have. This is so far the sabermetric discovery of the decade, comparable to Voros McCracken's work on pitchers' influence over balls in play. We are still trying to assimilate the framing knowledge: I don't think we've gotten a handle yet on how this should affect our perception of pitchers' effectiveness, especially when changing catchers.
* All three have also written at The Hardball Times, but a lot of their framing work has been at Baseball Prospectus. Besides, I'm making a side point on the PED madness—which you're free to ignore if it's too distracting.
New information from PITCHf/x produced that sudden about-face, a technological breakthrough triggering an analytic breakthrough. So what other doctrines we hold could be swung 180 degrees with the addition of the right set of data? Even more disquieting, what truths about baseball might simply be impossible to unearth through the methods we're committed to using?
This is not a denunciation of sabermetrics. I think sabermetrics is a beneficent revolution in baseball history, adding to the excitement and pleasure of the game for any fan—or player, manager, executive—willing to embrace it. If someone's circulating a petition to get Bill James inducted into the Hall of Fame, hand me a pen.
But sabermetrics is not the perfect machine, and I believe it never can be. We need to accept the limitations of our wonderful toolbox, if for no other reason than to keep us from becoming as insufferable as those know-it-alls who refuse to know things about baseball by the same method that we know them. And by "we," I mean, first and foremost, "me." This started off about keeping my head from ballooning up, and I will not forget that.
Of course, there is a countervailing pitfall in such modesty, especially for a writer, and very especially for a writer trying to reach conclusions and convey them to readers. Hedging and qualifying and modifying can be intellectually honest, but it is lousy, flabby writing. It's passive where you need to be active; it limply waves to readers to follow you rather than grabbing them and carrying them forward. It's Winston Churchill saying "This, on the whole, seems to have been their finest hour."*
* That wet-noodle quotation was actually spoken by British politician Lord Halifax, who had been expected to be made Prime Minister over Churchill, and who was expected to be "reasonable" in dealing with Hitler. Thank God for unreasonable people sometimes.
So I need to be clear and confident in what I write, yet, since I'm not locking horns with Hitler, something less than immovable moral certainty is wise. That should provide a roomy middle ground, yet now and again I can sense that I've weaved over the line trying to hold that course. The writer in me wants that driving active voice; the sabermetrician, sensing the Godelian limits of the system and not having the strongest grasp of deep statistical methodology yet anyway, wants to temporize.
This is where the baseball writer could use a little help. And this is where you can provide it.
You can do so by being thinking, critical readers. Be aware that one article is not enough to etch its conclusions into stone suitable for Mount Sinai. Be aware that a writer may have backspaced his moderating language into pixel purgatory to keep his readers' eyes from glazing over. Always have a grain of salt handy. Maybe more.
I would say that this won't be a problem, that ours is a readership self-selected for smarts and critical thinking. But my whole point has been about keeping our pride from running wild, so I'd better not. I'll say instead that this is the goal we must strive to achieve.
And if you need some practice applying your smarts and critical thinking, I can think of no better place than with the SABR Analytic Research Awards finalists. I said before, there's lots of great analytic writing being done. Go make your choices for the best.
P.S. In case you're worried that I've undermined the home team, I will note that Cameron and Carleton are nominated in a category without any finalists from The Hardball Times. I didn't choose their articles as my examples for that purpose; it was the luck of the draw.
Shane Tourtellotte is a long-time, occasionally-nominated science fiction writer, currently living in Asheville, North Carolina. He will tell you all about the baseball novel he’s shopping if you give him an inch.