All these moves are the result of a search for edges that don't exist. This isn't the normal annoying tinkering we've all seen in 100 boring playoff games; it's a compulsive effort to control randomness. The difference between Aceves and Robertson is about 12 runs per 500 batters faced, meaning that the lesser of the two gives up about .02 more runs per batter than the other. Neither player is more likely to get any particular hitter out—whatever differences there are in the speed or break of their pitches are utterly irrelevant next to the role of sheer random chance. An "intelligent" manager like La Russa or Girardi consults his color-coded charts as he thinks deeply about whether Howie Kendrick's swing will work better against Robertson's curve or Aceves' cutter. A wise manager understands that there's really no difference and there's nothing he can do but call on a decent pitcher and cross his fingers.
In this World Series, Girardi will be matched against a very wise manager, Philadelphia's Charlie Manuel. He speaks with a deep West Virginia drawl and doesn't boast degrees of any kind, having turned down a scholarship to Penn to help support his family. While Manuel isn't often tagged as a deep thinker, he does understand probability. Going on the recent evidence, that's a hell of a lot more than you can say for his opponent.
Perhaps the most useful sabermetric insight is that the most recent at-bat, game, or season isn't necessarily a good predictor of the next at-bat, game, or season. Barring injury, a pitcher with 10 years in the majors and a 4.00 career ERA will likely give up four runs in his next nine innings and 40 in his next 90. There are hot streaks and there are cold streaks, but they behave so much like random variation that the best plan is nearly always to ignore them.
While Manuel isn't about to hire Bill James as his bench coach, he clearly understands this principle. He pinch hit for a regular exactly once in the five-game NLCS and made just two defensive substitutions, clear evidence not only that he has a solid lineup but that he knows the soundest strategy is usually to roll out your best guys and hope they play well.
Perhaps even more impressive was Manuel's handling of Brad Lidge. The Phillies closer had one of the all-time rotten seasons this year, and it got worse as it went along. He gave up nearly a run per inning in the last month of the season, and received wisdom was that Lidge wasn't going to see a meaningful inning in October. Then the strangest thing happened: Manuel gave him the ball, and he pitched brilliantly, as he has most of his career. In five playoff games, he hasn't given up a run. The whole time, Manuel has allowed that he's mostly been pitching Lidge for want of other options, but he's also said again and again that Lidge has been a terrific pitcher over his career and that nothing has happened recently to make him think he isn't. It's worked well enough to win him a second straight pennant.
Manuel's general understanding that baseball is about players throwing and hitting and catching a ball is in perfect accord with the most sophisticated study of the sport. So is his simple insistence that players should be judged by more information rather than less. The Yankees have the better team, and would probably win the World Series if they were managed by a bag of sunflower seeds. The bag of seeds, at least, wouldn't pinch run for Alex Rodriguez, bench Jorge Posada, or swap relievers because of some minute difference that doesn't really matter. Girardi isn't wholly useless—he's refreshingly willing to use his closer outside of ninth-inning save chances, and he's dismissed calls to tinker with his batting order to aid slumping Nick Swisher. Still, if the World Series comes down to who has the smarter manager, the Phillies will win a second straight championship.