Going with a three-man rotation during the final two playoff rounds last season was a less complicated call for the Yankees. For one thing, they did not have a better alternative for Game 4. Chad Gaudin was an unappealing candidate who was never truly considered, and in that light, Sabathia was the only possible choice.
This year, the Yankees have Burnett, whose 5.26 E.R.A. was the highest in franchise history among pitchers who threw at least 180 innings. Girardi was noncommittal when asked whether Burnett would start in the A.L.C.S., but giving him a start in a seven-game series would be less risky, especially because the postseason schedule is not nearly as favorable.
Last year’s A.L.C.S. had an extra day off between Games 4 and 5 that allowed Sabathia, in a dominant Game 4, to be the only one to pitch on short rest. Burnett worked Game 5 on normal four days’ rest and Pettitte, after a rainout, started Game 6 on five days’ rest. Barring rainouts, no such flexibility exists this year. Major League Baseball, bowing to complaints that last year’s series took too long (6 games in 10 days), eliminated that superfluous off day. It means that over a potential seven games for the Yankees — if they committed to using three starters — their final four games would be pitched on short rest.
We should all have such problems. But at first consideration, it strikes me that the Yankees — and for all I know this flies in the face of game theory — should reserve a decision on what to do in Game 4 until they know what happens in the first three games.
If they emerge from those contests up 2-1 or 3-0, then they might as well roll the dice with Burnett in Game 4, knowing that at very worst they’ll be looking at a best-of-3 series with their three best pitchers on regular rest.
If they’re down 2-1 or with their backs to the wall, then they probably need to go with the big fella, as there’s almost no chance they should risk burying themselves with Burnett on the mound.
The problem with that, I imagine, is that Sabathia, Hughes and Pettitte would have to prepare differently if they were planning to start on three days’ rest than on regular rest. So, you know, nevermind.
Tedquarters 
Ted, I would say that you are becoming something of a game theory expert.
The nature of the game will change depending on the outcome of the first 3 games. Therefore, the payoffs for game 4 will be different, as you explained. The ideal starter for game 4 depends on the circumstances.
One of the things Ted Burke and I concluded from many extremely nerdy late-night conversations in college was that economics is really good at assigning fancy names to stuff you should figure out on your own if you thought about ’em hard enough.
So true. But if we all just simply referred to “Rock, Paper, Scissors” instead of “Game theory” the acceptance speech for the Nobel Prize in economics would not sound nearly as smart.
They could also start Sabathia on short rest in Game 4, and then start Burnett in game 5, Hughes/Pettite on full rest in game 6, and Sabathia on short rest in game 7.
That way you get three starts from Sabathia without having to pitch Hughes and Pettite on short rest. Downside is you still are starting Burnett for one game.
Not arguing that Yankess should do this, but could be considered, depending on situation.