Moneyball: The Art of Winning an Unfair Game
I'm not sure I agree with you about the media coverage of Moneyball. I've actually been impressed by how few reviewers have misread the book. Most reviewers seem to have a decent grasp on the fundamentals of Oakland's success and have done a good job of distinguishing between the method (sabermetrics, roughly speaking) and the man (Beane). (This is a testament to Lewis' ability to make complicated ideas accessible.) The people who have misread the book, ironically, are people in the baseball community. Of course, as you explained yesterday, they've got a psychological investment in doing so.
But you're right that, even though statistics are obviously important to the A's, the real key to the team's success is the way "they think not only about statistics, but nearly everything else that comes with running a baseball team." In a column I wrote about Beane for The New Yorker last year, I talked about a similar idea: What made the A's distinctive was the way the organization was built around a core set of principles (talent evaluation, player development, and game tactics), so that from top to bottom everyone was working, in a sense, from the same page. Among those principles would be, I think, things like:
On-base percentage is the most important offensive category.
Speed is overvalued in the marketplace.
The most valuable pitchers get lots of ground-ball outs, give up few walks, and give up few home runs.
When drafting pitchers, college pitchers are better investments than high-school pitchers, whose futures are almost impossible to project. (This is also true of college hitters, but to a lesser extent.)
With the right conditioning program and careful monitoring of pitch counts, you can significantly improve your chances of keeping pitchers healthy.
I don't think that any of these ideas, on its own, is exactly revolutionary (though there are far too many baseball executives who would still dismiss them). But when you put them all together and make the entire organization live by them, then you have a kind of strategic coherence that's hard to match.
Having said all this, there are aspects of Beane's decision-making that I confess I still don't understand. For instance, during an interview on CNBC yesterday, he said something he's said before (to each of us, actually): namely, that when he makes a trade he never worries about the player he's giving up, only about the player he's getting. As long as the A's get the players they want, Beane thinks it's a good trade.
Now, I've thought about this explanation a lot, and frankly it still makes no sense to me. Roughly speaking, it seems to be the equivalent of an investor saying that as long as he invests in a good company, it doesn't matter what price he pays for the stock. Brian Cashman—the GM of the Yankees—may be able to take this approach, but the whole point of Moneyball is that the A's always have to husband their resources—they succeed by finding undervalued talent and underpaying for it. So, what can it mean for Beane to say he doesn't care who he's giving up?
Take last year, when Beane traded Eric Hinske, a seemingly excellent third-base prospect, to the Toronto Blue Jays for relief pitcher Billy Koch. By Beane's logic, this was a good deal because the A's needed a closer, Koch was a strong one, and he was affordable. Hinske went on to be Rookie of the Year for Toronto. According to Beane's logic, that didn't matter, because the A's had already got what they wanted out of the deal.
But trading Hinske for Koch meant not being able to trade Hinske for someone else down the road. Lewis explains the deal unconvincingly, by saying that the A's didn't need Hinske because they already had a great third baseman in Eric Chavez. But, as he also notes, "Finding pitchers who could become successful closers wasn't all that difficult." So, why give away a potentially great hitter for an eminently replaceable player?
Beane wasn't necessarily wrong to do so. But there are aspects to what he's doing that remain mysterious to me, even after reading Moneyball. Yesterday I said that Beane lets the data make his decisions for him. But sometimes it's hard to see what the data are. Here's another example: Last summer he decided to trade away Carlos Pena, whom everyone had expected to be the A's starting first baseman for a long time to come. Pena seemed like a classic Oakland player: He had good plate discipline and excellent power, he was young, and the A's had basically stolen him from the Texas Rangers. Pena had a pretty good April, then went into a deep slump, got sent to the minors, and then was traded away.
Now, Lewis says Beane dumped Pena because "in a short two months … Carlos Pena had transformed himself from a player Billy Beane coveted more than any other minor leaguer into a player everyone valued more highly than Billy did." But one of the key ideas in sabermetrics—as in all statistics—is that sample size matters. In other words, whether a player does or doesn't hit over a stretch of 150 at-bats isn't proof of anything. It's only in the long run that you can be sure that the numbers are telling the truth. Two months isn't a big enough sample size to decide that the guy you thought was the best hitter in the minor leagues is actually overvalued. So, what did Beane see in those two months that made him so convinced getting rid of Pena (who is, by the way, currently struggling with Detroit) was the right thing to do?
What I really wonder about these deals is: Are they examples of Beane allowing emotion to get the better of him (giving away Hinske, a great long-term prospect, for Koch because he really wanted an established closer; dumping Pena because he was frustrated with his underperformance), or are they examples of Beane having evidence that we just don't know about? We think, especially after having read Moneyball,that we understand what's inside the Oakland black box. But maybe it's still full of secrets.
James Surowiecki writes the financial column at The New Yorker.