How many pitchers do you take? I'm thinking 8 and maybe 9 is the right number. That feels high to me but I want to be ready for two things;

1. Matchups - Teams should be able to work toward a platoon advantage as much as possible late in games.

2. Long Games/short starts - I don't want to be the guy who runs out of pitchers first in a 15 inning game. It's an extreme example but the Red Sox and Royals both used 8 pitchers in a 12 inning game earlier this year. I was at that game and at no time did I feel either team was being particularly ridiculous with their pitching changes.

If I take 9 that leaves me with 16 position players. The first 12 are fairly obvious;

9 starters
Infielder
Outfielder
Catcher

I then want two pinch hitters, ideally slotted as LHB/RHB depending on the makeup of my utility players above and a pinch runner. That leaves a 16th man. I think a lot at that point depends on what I have. If I don't have a guy who is more useful I'd take a third catcher but I'd prefer something a bit more useful than that.

I think it depends on your pitchers, to some extent. If the Yankees end up in the one-game playoff (oh, Christ, I hope not) they might be disinclined to put someone like Pettitte on the roster--unless he's starting, of course--because you don't know how he reacts to being used out of the pen. On the other hand, unless Sabathia pitches the day before, he is a dead lock to make the roster.

So you might go with fewer pitchers if you don't think some of your starters will help you. I do think you have to bring at least 3 SPs, though, one to start, one to match-up if need be, and one to be first/last man out of the bullpen and eat innings. (That's first if the starter gets killed, last if you're going 15 innings.)

Depending on who’s healthy and who pitches games 161 and 162, the Orioles could potentially be forced to do a “bullpen game” with no true starter. They may want as many as 10 pitchers for that strategy, with ~5 guys who can go two innings, and the other 5 for situational/extra innings.

I’d never take a third catcher in a game like this. It’s far more valuable to have another pinch hitter, pinch runner or defensive replacement.

This also could lead to some interesting pitching strategies at the end of the regular season.

Let’s say the Orioles have clinched a wild card spot, but are 2 games back of the Yankees going in to game 161. I think you could make a pretty strong case that the Orioles should save their best arms for the play-in game.

I’d never take a third catcher in a game like this. It’s far more valuable to have another pinch hitter, pinch runner or defensive replacement.

That's probably true but if you've got catchers with heavy platoon advantages you might find it useful. My other thinking is I want my manager comfortable pulling the trigger and running or hitting for the starting catcher. I'd hate to hear a manager say "I didn't want to leave myself with one catcher, what happens if he takes a foul ball off the finger?"

Let’s say the Orioles have clinched a wild card spot, but are 2 games back of the Yankees going in to game 161. I think you could make a pretty strong case that the Orioles should save their best arms for the play-in game.

Generally I think you have to keep going for it. The benefit of not playing a winner take all game is too great. The only type of team that I think would make sense playing it the way you suggest is a team with two great pitchers. For example, the 2004 Red Sox would probably have felt good going with Pedro in the Wild Card game and Schilling in Game One of an LDS (or vice versa) but even then a one game situation should be ducked as much as reasonably possible I think.

I would enjoy this rule exploitation this year; it's just a matter of time before teams have to submit a 28-man active roster before each game after the September 1st expansion.

Let’s say the Orioles have clinched a wild card spot, but are 2 games back of the Yankees going in to game 161. I think you could make a pretty strong case that the Orioles should save their best arms for the play-in game.

Generally I think you have to keep going for it. The benefit of not playing a winner take all game is too great. The only type of team that I think would make sense playing it the way you suggest is a team with two great pitchers. For example, the 2004 Red Sox would probably have felt good going with Pedro in the Wild Card game and Schilling in Game One of an LDS (or vice versa) but even then a one game situation should be ducked as much as reasonably possible I think.

But in this scenario, wouldn't the O's at best still have to play a 1-game playoff with the Yankees to determine the division winner? In that case, that playoff game is the one they would be saving their best arms for I would think.

Okay, but what if they're 1.999999999 games back?

But in this scenario, wouldn't the O's at best still have to play a 1-game playoff with the Yankees to determine the division winner? In that case, that playoff game is the one they would be saving their best arms for I would think.

Yeah that’s was my thought.

The only way to avoid a one-game play-in is a bit of a long shot - they’d have to win two while the Yankees lose two, and then win the division tiebreak game. It’s equally (un)likely that they’d end up in the worst case scenario of losing the division tiebreaker and then having to play the play-in game after spending all their ammo chasing the division.

I wouldn’t throw the games or anything, but I’d skip my best starter and go easy on my bullpen aces to make sure they were available for the play-in.

#4 is an interesting idea. Although I don't think anybody's ever done it in a 1-game playoff (or deciding game), if you had a good, rested bullpen, an all-star game like approach to the pitching might be your best option. I tend to doubt anybody would have a rested bullpen coming into a 1-game playoff nor one so deep you'd really trust 6-7 relievers more than all of your starters so it's probably more theoretical. Still, unless their ace was starting, I can easily see the starter pulled after 4-5 innings to bring on a parade of 1-inning guys.

Sheehan and Rany have discussed this issue (rosters for the one-game wild card playoff) EXTENSIVELY on their last 3-4 podcasts.

With Up and In no longer in production, I strongly recommend following them.

Using the plan in #1 for the A's:

Catchers -- Norris, Kottaras
Infield -- Carter, Moss, Barton, Weeks, Rosales, Pennington, Drew, Donaldson
Outfield -- Cespedes, Crisp, Reddick, Smith, Gomes, Cowgill
Pitchers -- Blackley (starter), Balfour, Cook, Doolittle, Blevins, Neshek, Figueroa, Scribner, Milone

Injured/Suspended -- McCarthy, Norberto, Colon, Devine, Braden, Anderson, Inge, Carignan, Sizemore, Sogard
Unavailable Starting Pitchers -- Straily, Griffin, Parker
Not Selected -- Chavez, Ross, Miller, Hicks

The only position player not on the active roster would be Brandon Hicks and his main skill is being able to play all IF positions badly. This could come in handy if Melvin were to PH for 3 other IF. I could see keeping Hicks instead of Scribner or Figueroa.