Dejan Kovacevic doesn't like it very much when I pick on him, and I know he thinks it's unfair, but look -- if any other well-known Pittsburgh media personality who I'm supposed to take seriously posted something this provocative and ridiculous, I'd be talking about that, too.
Has the Moneyball crowd yet trademarked the term "Meh?" How about "ad hominem?" Or "straw man?" Look, I love advanced stats ... But the amount of groupthink that occurs within this group flies completely in the face of two other favorite terms, "objective" and "intellectual." True objectivity -- and I'm not claiming purity here -- comes from, duh, an open mind. Maybe there's a REASON closers pitch the ninth.
The last sentence is hilarious. I'm sure there are reasons why closers pitch (only) the ninth. (There's no controversy about closers pitching the ninth sometimes, or even often, obviously; the issue is how much they should be used in non-save situations.) But you'll notice that Kovacevic isn't going to tell you what they are. You'll just have to trust him, and all the managers who continue the practice despite it flying in the face of all common sense and evidence. If you don't, you're a sheep. "Duh."
So, what are the reasons closers pitch the ninth? Well, a sportswriter named Jerome Holtzman invented a statistic called the "save" sometime around 1960 and became an official stat about a decade later. Eventually, managers began to tailor their reliever usage to the save statistic, which typically limited the number of innings their best relievers pitched, and required their relievers to pitch in relatively-low-leverage situations, such as in the ninth inning with a three-run lead and none on, instead of, for example, in the seventh inning with a one-run lead and two on. If there is a reason for managers to use closers that way, a sportswriter named Jerome Holtzman seems to have invented it.
As for why managers continue to do it, it's probably mostly because that's what their relievers want, and because, for managers, doing something different would be a high-risk, low-reward proposition. Yes, using their best relievers in higher-leverage situations would probably make their bullpens better, but trying something even remotely unconventional opens them to all kinds of second-guessing from the more reactionary elements of the media, and it runs the risk of making their "closer" unhappy, since he may perceive that he'll be worth less money on the free market if he isn't allowed to accumulate saves. (Or, less cynically, he may have an easier time preparing to pitch if he knows he's going to enter in the ninth. But he should be able to get over it, since no other relievers besides closers know exactly what inning they're going to pitch in.)
The 2003 Red Sox tried going with a "closer by committee," and the media predictably freaked out. The "closer by committee" didn't work (probably because the Red Sox simply didn't have a very good bullpen, not because the strategy itself was flawed), and Boston acquired Byung-Hyun Kim. Sadly, that's probably the most lasting memory most fans have of recent teams who tried something different. But it doesn't prove much of anything, and there are counterexamples of teams that used the back ends of their bullpens more flexibly, and quite successfully. Check out the 1990 Reds, for example, who used Randy Myers as closer most of the time but frequently allowed Rob Dibble to collect saves in multi-inning relief appearances, usually to great effect. Or the 1986 Mets, which used righty Roger McDowell and lefty Jesse Orosco in a sort of closer platoon, with both frequently pitching multi-inning appearances. Or the 1985 Cardinals.
That using your closer in (only) the ninth doesn't expose your closer to the highest-leverage innings isn't really even arguable. The main reason teams don't do anything about it seems to be inertia, frankly -- conservative elements within the game and outside it who don't want things to change. But that doesn't mean that, on a systemic level, current closer usage makes any sense at all.
The funny thing about this that the "Moneyball crowd" is really the group that's trying the hardest to think outside the box and look at what the evidence says on this issue. I can't remember the last time I even saw an earnest defense of current closer usage. Correct me if I'm wrong, but as far as I know, no one is even trying to make serious arguments that it makes any sense, at least not beyond saying that some relievers tend to be more comfortable if they know their roles. And yet somehow it's the "Moneyball crowd" who are practicing "groupthink" on this issue.
To be clear, I think the problem here is a nuanced one -- in this age of specialization, teams are probably using their bullpens better than they did 40 years ago. And the number of players in today's game who could handle a Goose Gossage-type role is pretty small. (The debate on closer usage doesn't really have much bearing on the 2012 Pirates, either -- Joel Hanrahan has been used fairly rigidly, so his outings have been only slightly higher-leverage than Jason Grilli's, which is actually good, as far as it goes, given that Grilli has been the better pitcher.) But that doesn't mean that the rigid approach some teams take with the back ends of their bullpens doesn't border on absurdity.
My guess is that, within a generation or so, teams will think about their closers differently than they do now. But it's going to take a long time. There will have to be teams who are willing to experiment, and the current barriers to experimentation are pretty high. The first teams to experiment will probably be teams that don't have much to lose, and those kinds of teams generally don't have good bullpens. It will take awhile for things to change. But teams are currently leaving plenty of wins on the table.