Here is a scenario that comes up in one form or another fairly frequently -

10/20 NL, I've got a full stack. I get AA in the SB. 3 limpers, I raise to 120. Called by one of the three limpers, who is a shortstack.

Flop comes with no possible straights or flushes. Say T54. Pot is 300, and you bet the pot. Your one opponent, whom you have no read on, raises all-in. How much does he have to have for you to consider folding? When is the "shortstack" no longer short? Should I take the philosphy that so long as I dont lose a total of 10x my preflop raise, then I'm OK? So anything less than a $1200 loss is OK here?

The pot will be 900+ the amount of the raise. How much is too much?

D

Your guideline sounds pretty healthy to me. If you really want to get picky, you could set it at 7.5 times the raise, set being 7.5:1. But, as has been pointed out, you also do have chances to overset them on the turn and river.

I really wish we had at least a "bluffy/non-bluffy" read on this one, though.

You may also be up against someone putting you on unimproved AK. This is a rather difficult situation, I think.

Honestly, with no read at all, I'm fairly tempted to call against stacks up to around $1500 (or put them in if they don't bet the whole thing), but I'm not terribly happy about it. Against unknown players, I don't think a statement of that sort is a bad move, although it's a rather expensive statement at those stakes if you're in fact behind.

This is actually a thread I hope we can pursue further, as I think it's enormously important.

Hitherto, here's what I've considered just "correct ABC": You bet your AA as you say, then call the first raise, then you're done with the hand without improvement. But quantities here are certainly extremely important, and this question gives rise to a number of others even with a lot bigger stack depth (like "How to play the set?" and "How and when to lay down your aces?"). The moreso for me, since I've started buying in at $600 in my 2/5 as a rule (at least against the players whom I already know). I want ideally to have anybody I know at the table covered. I actually don't like having much more than 100xBB against unfamiliar big stacks (one thing I really don't like with huge stacks are KK vs. AA situations, but that's another issue).

I agree with rdale, by the way, at really labeling more like 50xBB or less as short-stacked, but I'd like to analyse the situation regardless of stack depth.

Ok, against truly unknown player, the call up to $1,200 in your example obviously has nothing to do with drawing to the additional ace. It's really more an indicator of toughness than anything else (by the way, the raise to $120 seems a little large to me, unless there's some reason at the table--I do like the "I want 2 callers" rule in determining raising quantity).

Now, out of position, what is happening here is that the limper calls only $100. He makes a profit by setting if he can pull $750 out of this pot. And if you're on aces and don't grant him this action, I think the penalty you suffer against a perceptive player is getting no action on your aces. Departing a bit from pot odds and factoring in the setting chances for the aces on the turn and river, I actually think you only need to call an unknown deep stack up to a raise of around $900--i.e., an immediate raise with set to $1,200 is excessive against a good player willing to bet out full pot on the flop, as you did. If the losses are limited, I think I'd call an all-in here only up to right around $1,000 (with regard to your original short-stack question).

From the perspective of the set, raising to $900 here gives you $600 to call with a pot of $1,500. Raising to $1,200 gives you $900 to call on a pot of $1,800 (i.e., the old 2:1). To an aggressive bettor with the big pair, I think the $900 is actually plenty.

Now, there are several dimension to this whole scenario: First, no one knows you didn't have AK, in which case you're obviously laying down to any raise. But looking at it from both sides here, an AA (if one can put you on AA or really any overpair here) that doesn't call the $900 raise is in my opinion selectively bluffable. If I'm playing against this player regularly, I'll definitely start playing some suited connectors and one-gappers to this raiser with the intention of making the same semi-bluff (and I might even try just a blatant bluff if I feel like it's set up properly--anyhow, failing to call the $900 with AA or overpair is imo just weak-tight).

But I'm just going to say (again assuming deep rather than short stacks) that raising to more than $900 is playing the set simply too fast. Anyone care to agree or disagree? I'm not saying that you may not WANT to play the set too fast against certain players (those who are too cally with their big pairs), but I think that's just the "right bet" against a good, aggressive player (and I'm not quibbling here about anything in the $900-$1,000 range).

On the other hand, let's also look at it with positions reversed: Say you're on the button rather than in SB and make the same raise with the same result. Now someone bets into you for $300. I think you're also obligated to raise your aces to at least $900 (against a bluffy big stack, I'm probably raising more, but, again, the raise and this betting quantity just seems to me to be roughly "correct").

At this point, with set, I think you need to move in (or re-raise to $2,700 or so, if both players have stacks that deep), and a good player with aces can in good conscience fold against a credible opponent. I don't see much reason for the set to give a free card to the turn here. If AA is going to still call a sizeable turn bet unimproved, then AA is going to still call on the flop, I would think. I really see no reason not to play the set pretty fast here.

So, I guess that's my analysis. I'd be tempted to put the stack-depth cap actually a little smaller against an unknown (around $1,000) for calling a potential set.

I wouldnt really call $1200 in a $2000 buy in game..a short stack.

I think that "short stack" is just meaningless semantics for the purposes of this discussion. The question is this:

What pot odds do you need to call with AA on an uncoordnated board on the flop after showing strength, provided that the implied odds on the turn and river are exactly zero?

Here are some steps you could take towards answering that question:
(1) Under what circumstances would you think that the opponent has less than a 60% chance of having a set?
(2) If the player is raising more than the pot size, does a larger raise indicate a higher or a lower probability of a set?

If you can answer those two questions, you ought to be able to pretty much deduce when you should call and when you should fold in this specific situation.