As long as Joe Torre has been managing in the Bronx, there has been one, and only one, opinion on which every observer, from the jeering Mets fans to the Paul O’Neill worshipper in the right field bleachers, have all been able to agree: Torre works his bullpen too hard. Every year there’s someone like Scott Proctor or Tom Gordon or Paul Quantrill who pitches impossibly well, gets called in from the bullpen more or less every day in the first half, suffers a period of dead arm, and ends up useless when the team needs him most.

This year, in addition to his many other problems, Torre is managing to botch the bullpen yet again. The Yankees are third in the league in relief innings pitched, and while most of that can be blamed on the team’s various injuries and the presence of a number of Class AA pitchers in the rotation, Torre’s inimitable work is on display. Lefty specialist Mike Myers, who last threw 50 innings in 1998, has already racked up 30. Kyle Farnsworth, famously unable to pitch well without long periods of recovery time between appearances, has been run out again and again on one day of rest, with predictable results. Andy Pettitte and Roger Clemens have made relief appearances; meanwhile, potentially useful pitchers like Sean Henn are plowing away in the bush leagues. It’s all a bit of a mess.

What’s interesting, though, is that while we’ve seen the unwise use of relievers, we haven’t seen anyone designated as a dray horse. Brian Bruney, seemingly just the sort of pitcher to whom you’d expect Torre to do horrible things, pitched five times in one six-game stretch at the beginning of June, but he’s otherwise been given plenty of rest, rather than being flogged until falling over. It isn’t just the lack of one designated workhorse that’s notable, either; as a whole, the team’s better setup men haven’t really been worked that hard compared to those who play for other American League teams.

The standard for “better” here is murky; judging reliever performance is difficult. Imagine two pitchers, each appearing in five games. One escapes four basesloaded jams and allo ws two solo home runs in the fifth game, the other allows base-clearing doubles in four different games and allows no other runs. Going by ERA, the second will look better, but he has been terrible, and the first pitcher superb. A good reliever’s ERA can also be inflated by one nightmarish inning. For these reasons, I used WXRL, a Baseball Prospectus statistic that gives pitchers credit for stranding inherited runners and the like. Context is everything in relief, and by this measure, Henn, Kyle Farnsworth, and Luis Vizcaino have been the Yankees’ three most effective setup men.

Those three accounted for 88.3 innings through the start of play yesterday, meaning that Torre was able to juice 1.15 innings a game out of the three (one of whom, granted, has pitched one game in the last two months). That’s an in-between figure. Minnesota, with a deep corps of extremely effective relievers, has had its three best setup men throw 1.48 innings a game. On the other end of the spectrum, The Chicago White Sox’ three best have accounted for only .52 innings a game, a horrific number, because most of its relievers have been so awful that one Dewon Day, who’s pitched in seven games, has been the team’s second-best setup man once you account for game situations. The league average is 1.06, so, really, Torre’s hand hasn’t been all that heavy relative to those of other managers, especially considering the team circumstances.

This is all imprecise and meant only to point out broad tendencies — I wouldn’t argue that Henn has been the Yankees’ third-best setup man, no matter how his performance looks in context according to WXRL or any other statistic, or that my sketch here captures usage patterns anywhere near precisely. It is interesting, though, not least because one of the things that I and others have criticized Torre for is inflexibility, which is really at the core of the complaints about his handling of the bullpen. The problem isn’t so much that he’ll overwork someone like Proctor if he has a chance as that he sees managing as fitting players into slots. If at all possible, for instance, he’ll have a setup man he can work to death and a first baseman who can’t hit and has an inflated defensive reputation. Given the chance to platoon two players, Torre will always alternate between playing them two weeks at a time.

If in this of all seasons, though, Torre isn’t working the pen to death nearly as much as one might think (Myers excepted), doesn’t that show something the critics might take into account? After all, while the general presumption is that Torre will be gone after this season, things are always mercurial in Yankeeland, and it would be a good idea to at least consider that he may remain the best man for the job even if the Yanks miss the playoffs. Especially if, after more than a decade, he can show that he’s a bit more flexible than he sometimes gets credit for.

tmarchman@nysun.com