This post really belongs in the odds section, but I'm going to post it here anyway because it's just one of my favorite plays.

I'll start with a hand from the other day where I really don't like my play (can't remember exact stack sizes): I have KdTd in MP, and the flop comes something like Ad8d2c. I bet out pot of $25, and LP raises to $100. I badly decide to flat call, miss my flush but do get my turn free card, and we're done when my opponent turns over AQ.

Well, this is a classic information raise from my opponent, who imo plays the hand correctly. I think I should have moved in on that flop in actuality with my nut flush draw. What I'd like to do here is test my hypothesis.

First question is: What re-raise size gives him 2:1 to call? (I'm going to deal with this in general in the odds section) Ok, there's now actually $150 in the pot, so a raise to $350 total would mean that he has to call $250 for a total pot of $475. That's close enough, anyway. With normal stacks, that's pretty much all-in.

Now, the second question is: What fold equity do I need to make this raise to $325 MORE a profitable play? Let's say the probability of a fold is p. Then, here's what happens:

I win a pot of $150 with a probability of p.

If he calls, then I'm a dog here. I lose $325 65% of the time (with 9 outs) but win $400 35% of the time. Ok, so we have: .35*400 - .65*325 = 140 - 211 = -71.

This occurs with a probability of 1-p. So, the break-even point is when

0 = (p*150) - ((1-p)*71) = (p*150) - 71 + (71*p)
71 = p*221
1/3 = p, more or less.

Hence, if there's a 1/3 probability of a fold (I think one can do better than that), it's a profitable move.

Now I'll just use the same numbers for a nut straight draw, where 8 outs yield only a 31.5% chance of winning at showdown. Let's just say my opponent has the same hand, but the board is now A54 rainbow, and I have 76s.

Now, I lose $325 68.5% of the time but win $400 31.5%, so we have on the call: .315*400 - .685*325 = 126 - 223 = -97.

For p we now get:
0 = p*150 - ((1-p)*97)
0 = p*150 - 97 +p*97
97 = p*247
.39 = p

So, basically, in both cases, you're good to go with 40% or better fold equity.

My conclusion: This is a good play against pretty good players (obviously not one to make against calling stations, but they don't make information raises anyway).

The real idea is this: A fairly good player is likely to flat call or minimum raise on a set but make a larger information raise. If someone has that betting tell, he is actually pretty unlikely to call your re-raise with a mere top pair (particularly if you'll make the same move on your own set). It's just not enough hand to do so. Moreover, the same player, with the minimum raise on the set, actually allows you to see the turn at least for cheap enough (maybe even the river). So, you can fairly safely just flat call the minimum raise, obviously letting go completely if the board pairs, and you can semi-bluff hard against the information raise, typical of TPTK or such (also 2 pair, who is going to be much more likely to make the call, but 2 pair is fairly rare).

Why do you think he played it correctly when he checked the turn behind you and gave you a free card? I guess he was afraid you were going to check raise him, but I wouldve bet turn turn if I were him.

Well, the idea of the information raise is to represent a better kicker. My flat call actually represents AQ or AK for me. I'm pretty sure this is what he assumed I had.

I think it actually is correct to check the turn here on TP hands, and his raise was big enough to deny me odds even to the river if I happen to be on a draw.

The idea of the semi-bluff is making it very difficult to put a tight player (such as me) on an exact hand. In that situation, basically I should lay down something like AJ there, flat call the raise with AQ (perhaps, although I don't mind even laying it down either), but move in over the raise both with set and on the semi-bluff.

For it to be fully viable on the semi-bluff, one may want some extra outs--like with my KdTd, a flop of AdTc2d, giving me 5 extra outs against AQ for a total of 14.

Anyhow, this move is pretty much in line with trying to get a uniform (and aggressive) betting strategy together on a certain range of strong hands--thus putting quite a bit of pressure on opponents. I really don't see how a reasonable AQ can actually call for stack at that point--I wouldn't. Basically, AQ, while ahead to the draw, isn't a huge favorite and is completely dead to the set.