Here’s the clearest way to look at it. Take the two simplest cases. Bad Scenario: we’re down 3-0; Good Scenario: we’re up 3-0.

In the Bad Scenario, we have to win all four remaining games. In the Good Scenario, we have to win one of the remaining four games.

I’m assuming (I think everyone is) that an AJ start gives the Yanks a considerably lower chance of winning, but also that starting one of the other pitchers on short rest reduces his chance of winning. So the relevant question is whether the reduction in strength of the better pitchers outweighs the advantage of leaving AJ out of the sequence.

I think it takes a fairly odd collection of probabilities to make it *better* to leave AJ out in the Bad Scenario but better to give him the start in the Good Scenario.

For a little practice model, I assumed that CC has a 60% chance of winning on full rest, Phil and Andy have 55%, and each of them loses 5% on short rest. I gave AJ only a 30% chance of winning.

In this model, leaving AJ out gives the Yankees better chances in either scenario.

Do those numbers seem very unrealistic? (I set up a little spreadsheet so I can re-run it with different numbers very easily.)

]]>If Phil had looked bad the other day, I’d feel differently. ]]>

That way you get CC, Andy, Phil, on regular rest.

If it were just CC, then sure, go short rest. But if CC were to be slotted for Games 1, 4, 7, you’d *still* have to have AJ in there, or else the other two studs on short rest, and if the former then I think having each of the three good pitchers on regular rest is better than CC twice on short rest, and the latter seems like a bad idea.

So, sure, maybe AJ spits the bit. But (a) he might be good, and (b) even if he’s bad the Yankees could still win that game (against the opposing #4 pitcher, after all), and (c) if the Yanks lose at least they’re set up really well for the last three game.

]]>