[tewall]

I did an ICM calculation assuming a resteal of an opponent who would steal with anything and call with the best 25% of those hands, which would include hands as bad as K6s and K9o. (I used pokerstove, setting at 25%). I had to make up chip amounts, etc., but assuming this scenario restealing with Q7s was clearly profitable. In the scenario I did, the resteal would be just over 25% of hero's stack.

The calucaltion was roughly

a.Just under .14 equity to start with
b.Equity rises to .155 if resteal works
c.Equrity rises to over .23 if called and win
d.Equity goes to 0 if called and lose

Probability of winning is 1/3 if called.
Probability of opponent folding is 3/4.
Total equity rose from just under .14 to about .19, a huge gain.

The precise numbers aren't important, of course, but it gives a feel to what is happening. Most of the equity gain comes from the fold equity, but a good chunk comes from 1/3 of the time when called.

I used a serious simplication here which favor the restealing calculation, which is that no one behind will call. Also I assumed the stealer would steal with any two. However, the assumption that the stealer would called with hands as bad as K9o is a conservative one.

I would be happy to do other calculations with any given assumptions, but based on a preliminary trial, it appears to me that if one is correct in the assumption that the opponent will steal with any two, restealing with Q7s in a scenario similar to the one Colin laid out would be clearly profitable.

Out of curiosity, I redid the calculation assuming restealing with 32o. This was also profitable, raising the equity from just under .14 to almost .16.

It would probably be more accurate to weight the calculation. Something like x% of the time it's a steal, and y% of the time it's a value steal. That is, villain might not have been intending to steal, but woke up with a hand, and our guess that his raise was a steal was incorrect.

I redid the calculation assuming that villain would be raising with a legitimate hand (top 10% hand) 1/3 of the time, and any two 2/3 of the time. Under this assumption, the resteal with Q7s was still profitable, raising the equity from a little under .14 to .15.

So one's guess that the villain was restealing would have to be way off for the resteal with Q7s not to be profitable. If one were absolutely certain that villain would raise with any two, and fold hands worse than K6s, then restealing with 32o would be profitable in the scneario I ran. Having a hand as strong as Q7s allows one to be wrong by quite a bit in the guess that villain is stealing, and still have that be a profitable play.

Of course, even though villain is stealing, he might not steal with any two. He might steal with top 50% or something like that, so there's many different assumptions that could be made. But it appears to me that regardless of the assumptions one makes, Q7s will be a profitable resteal, given that your guess that villain's raise is a steal is reasonably accurate.