Using numbers to hide

I'm a little more than halfway through Sheldon and Alan Hirsch's "The Beauty of Short Hops: How Chance and Circumstance Confound the Moneyball Approach to Baseball."

In spite of the second half of the title, it's not a strict takedown of Michael Lewis' "Moneyball," although the authors do a pretty good job picking apart what they see as the flaws in the book, including the same criticism that Tony LaRussa had of the movie (which hewed pretty closely to the book, as I recall) ... namely that in the desire to paint Billy Beane as a genius, it ignores the main reasons the A's were successful, in particular Tim Hudson, Mark Mulder and Barry Zito.

But not only are the Hirsches not LaRussa acolytes -- they're no fans of his approach to closers -- their book isn't strictly an anti-sabermetrics screed. Their view is considerably more nuanced, as this passage on page 62 states:

"The problem that animated Bill James's revolution was colossal ignorance pervading the baseball world. He helped cure it. The problem he inadvertently ushered in was excessive faith in a particular path to knowledge and insufficient appreciation of how much can never be quantified."

As does this on pages 110 and 111, after the authors have given credit for sabermetric advances in the understanding of on-base percentage and closers, imperfect though those advances may have been:

"It is one thing to recognize a deficiency in baseball tactics, quite another to correct it, particularly without unleashing unanticipated negative consequences. Focusing on numbers, while overlooking the nuances and subtle forces that can't be quantified, sabermetricians fail to appreciate the complexity of the game they seek to transform."

It points out much more eloquently than I have in arguments with friends the main problem I have about those who worship at the altar of numbers -- they have a tendency toward absolutism.

In their world, batting average and RBI are virtually useless statistics, as are pitching wins. There's also no such thing as "intangibles," "clutch" or "chemistry," and Derek Jeter can't field. (One thing we agree on ... David Eckstein is highly overrated.)

But before I came to the first quote noted above, the Hirsches spent some time on Game 7 of the 2003 American League Championship Series. The numbers said Red Sox manager Grady Little should have taken out Pedro Martinez after seven innings, given his tendency to tire after throwing about 105 pitches. Little's gut said otherwise, so he left him in. You know how it ended up ... the Red Sox blew the lead, Aaron Boone happened, Little got fired. (I enjoyed all of this tremendously.)

From page 62:

"You follow your gut, or your best guess, and then you watch the game unfold in its unpredictable glory. Unless you're a devout sabermetrician, in which case you think the decision is made for you by the numbers. If you take the latter approach, you may sleep easier (win or lose), but you misunderstand messy reality and shortchange the beauty of the game."

Would Grady Little have been fired if he took Pedro out and the Red Sox lost, anyway? Well, consider this nugget from the story about Little getting fired:

"The Red Sox have insisted that Little's future will not be determined by the one decision, but according to the (Boston) Herald, the team was concerned about Little's lack of reliance on stats."

If Little's situation is any indication, relying on the stats means never having to say you made a mistake (or getting fired for it).