Stud hi: What's your plan when 3-bet on 3rd by an overpair?

[Spladle]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
So Spladle make a decision, you only have 10 seconds. Call or raise?

[/ QUOTE ]
What's the ante/bring-in?

[/ QUOTE ]


For the purpose of discussion let's say it's 8-handed 100-200 with 20 ante and 30 bring-in.

[/ QUOTE ]
Re-raising in this situation is a no-brainer.
[ QUOTE ]
Also to anyone who is still reading this thread, say you do reraise the 885, rightly or wrongly. What then is your plan if your opponent 3-bets? Would you agree with what I outlined in my OP?

[/ QUOTE ]
I agree with your plan when you know he has AA.

When he could be 3-betting a wide range, I agree that you cannot auto-fold the river. However, auto-calling also seems incorrect. I guess the solution is to fold it "sometimes". Folding fifth just seems very bad to me but I may be wrong.

[Spladle]

[ QUOTE ]
[ QUOTE ]

I am assuming the A will fold when he has no pair, no 3 flush, and no 3 overcards.

[/ QUOTE ]

You really think he's folding AK4? A lot of folks won't even fold a razz hand.

[/ QUOTE ]
In that case the re-raise is a value bet.

[electrical]

[ QUOTE ]
Here is where you're confused, I think. In HU hold 'em, the pot is the same size pre-flop as it is when there are ten players. However, in HU stud, the pot is smaller than when there are eight players. Therefore, when you hold an ace and are looking at an 8 and a 2 left to act, you should raise more often if five players have folded than if none have.

[/ QUOTE ]
Then why do I observe more (apparently) light raising and vbetting in short-handed stud games? Do you not observe that as well? Are all these short-handed players playing incorrectly?
[ QUOTE ]
What you mean to say is that it boils down to the equity our hand has against the stealer's range. I think MM said somewhere earlier in the thread that we are about a 60/40 favorite here.

[/ QUOTE ]
I would love to see that simulation. I can get different categorical equity measures for different classes of hands, but only by using the twodimes calculator, and only one example at a time, and there is no information about frequency, which would tell us about the range as a whole. I don't see .600 as a realistic aggregate result, but I don't mind being proven wrong. Here's what I got:

against:
(AA)A we are at .077
(KA)A -- .349
(4A)A -- .349
(TT)A -- .367
(88)A -- .370
(KsJs)As .489
(66)A -- .549
(55)A -- .574
(2s3s)As .576
(KJ)A -- .586
(6s4s)As .593
(23)A -- .687
(64)A -- .701

[ QUOTE ]
Your thinking here is flawed. So long as you are playing in games that you are adequately bankrolled for, +EV spots should never be turned down.

[/ QUOTE ]
Do you think, in the long run as played, not as a hot-and-cold simulation, these hands win more than they lose? We can basically never raise with two pair, but the Ace can often induce a bet and check-raise with two pair or concealed trips. The only way to defeat that is to play super passively, which guarantees that we win the minimum. Since there is doubt about the villain's holding, and an expressed willingness to fold to a scary board, we become bluffable later in the hand, when it is more incorrect to fold a winner.

The "+EV situation" cachet seems to ignore these playing circumstances and just assumes a passive hot-and-cold call-down, though very few of us are willing to play poker that way. The OP strategy essentially advises giving up on the river (or Fifth) with one small pair, but that defeats the predicate assumption that one small pair is "ahead of the range." The equity sims certainly do not calculate these "incorrect" folds into the figure, and as such are optimistic, if one follows the proposed strategy.

Now, tell me why all that is wrong. The bit before about the clumping was interesting.

[Spladle]

[ QUOTE ]
Then why do I observe more (apparently) light raising and vbetting in short-handed stud games? Do you not observe that as well? Are all these short-handed players playing incorrectly?

[/ QUOTE ]
It's likely that you've never observed good players in a full-ring setting with a high ante.

[ QUOTE ]
I would love to see that simulation. I can get different categorical equity measures for different classes of hands, but only by using the twodimes calculator, and only one example at a time, and there is no information about frequency, which would tell us about the range as a whole. I don't see .600 as a realistic aggregate result, but I don't mind being proven wrong.

[/ QUOTE ]
There is a program that performs such simulations. I forget what it's called. MM uses it I think. He posted this result earlier in the thread and I have been assuming that it is correct.

[ QUOTE ]
Do you think, in the long run as played, not as a hot-and-cold simulation, these hands win more than they lose?

[/ QUOTE ]
Yes.

[ QUOTE ]
We can basically never raise with two pair, but the Ace can often induce a bet and check-raise with two pair or concealed trips. The only way to defeat that is to play super passively, which guarantees that we win the minimum. Since there is doubt about the villain's holding, and an expressed willingness to fold to a scary board, we become bluffable later in the hand, when it is more incorrect to fold a winner.

[/ QUOTE ]
This simply isn't true. We can raise any time we like. We can check behind with one pair if checked to. Or we can bet and fold to a check-raise if the ace is bad enough to only make that play when we're crushed. We can bluff. The reason we fold with one pair on later streets against scary boards is because our equity will not be favorable against the range of hands that he could hold even considering the money already in the pot. If our opponent holds a weak hand very often in this situation it will be because he made an error on third street by calling our re-raise too loosely. So it's okay to fold in these spots, even though it seems as though you're being exploited, because you're really not. He had to put in too much money as too big a dog to count on making a winner or developing a scary board while we failed to improve often enough for continuing on third to be profitable.

Basically, what I'm saying is that you're right, the hand isn't over. If you play good on the later streets, you can play looser on 3rd. If you play bad on the later streets, you should play tighter on 3rd. However, even if you play bad and expect to lose a little later in the hand, (85)8 is simply too strong in this situation to fold. Ed Miller used to talk about this when trashing other low-limit hold 'em books. Beginning (bad) players should raise more pre-flop because they expect to lose money later in the hand, so they should exploit thin equity edges pre-flop. Stoxtrader mentions in his new book that you should be more inclined to three-bet hands like A9o that suffer from reverse implied odds so as to exploit your equity edge while you have it. Howard Lederer wrote an article for Full Tilt awhile back talking about the same hand, but in a no-limit setting. Late position raiser, short stack, A9o, yadda yadda yadda, arrr-in. A9o is ahead of the raiser's range, but you'll lose money playing it post-flop, so push your edge while you have it.

All of these considerations apply to this hand. I was a little hesitant to "appeal to authority" but I think you'll probably be more likely to believe them than me, and anyway this isn't [censored] debate, it's just a poker forum.

[ QUOTE ]
The "+EV situation" cachet seems to ignore these playing circumstances and just assumes a passive hot-and-cold call-down, though very few of us are willing to play poker that way. The OP strategy essentially advises giving up on the river (or Fifth) with one small pair, but that defeats the predicate assumption that one small pair is "ahead of the range." The equity sims certainly do not calculate these "incorrect" folds into the figure, and as such are optimistic, if one follows the proposed strategy.

[/ QUOTE ]
You've misunderstood. Calling down and giving up on the river is only correct if three bets went in on third and you know for a fact that your opponent has 88 beaten on third.

When you re-raise third, you're usually not going to get three-bet. Much of the time he'll call, check/call fourth, and check/fold fifth. Or maybe he'll check/fold fourth or check/call fifth and sixth, check the river, and 88 will be good. Or maybe he'll check/call the river and 8s up will be good. Or maybe he'll check/raise at some point and you'll call down and get shown a better hand, or maybe a bluff. Or you'll fold to the check-raise. Or you'll call the check-raise but fold sixth because his board improves.

What we're talking about here is how to play the later streets when you're opponent three-bets third with a wider range than just hands that beat 88 (let's say any totally live 3-flush or pocket pair). Now, you're still a dog when he three-bets, but not by so much that auto-folding the river is correct if he'll fire every street unimproved (which he is correct to do if you'll auto-fold the river unimproved).

You're right, the hand isn't over, but unless your opponent is much better than you, he won't make up enough ground on the later streets for folding third to be correct. If you're much better than your opponent, you could even get away with playing some really trashy hands in this situation. However, in order to arrive at what we hope will be close to game-theoretically optimal solutions, we tend to use hot-and-cold equity calculations, since solving perfect play for the later streets is just impossible. These equity calculations ignore implied odds, reverse implied odds, and differences in skill between opponents. Fortunately, when we are aware that these factors will be relevant, we can modify the amount of equity that will be necessary to make certain early-street plays. It's not perfect, but it's the best we can do.

Poker is hard.

[ QUOTE ]
Now, tell me why all that is wrong. The bit before about the clumping was interesting.

[/ QUOTE ]
The thing is, you're not exactly wrong. When playing against exploitably tight (bad) players, or with a small ante, folding third can be correct. That just isn't the situation we're talking about here.

[electrical]

Dude, incredible. Very nice read. I have a lot to think about here.

[SGspecial]

[ QUOTE ]
[ QUOTE ]

I am assuming the A will fold when he has no pair, no 3 flush, and no 3 overcards.

[/ QUOTE ]

You really think he's folding AK4? A lot of folks won't even fold a razz hand.

[/ QUOTE ]

A lot of folks wouldn't even fold AK4 IN razz [img]/images/graemlins/tongue.gif[/img]

[SGspecial]

[ QUOTE ]
Howard Lederer wrote an article for Full Tilt awhile back talking about the same hand, but in a no-limit setting. Late position raiser, short stack, A9o, yadda yadda yadda, arrr-in. A9o is ahead of the raiser's range, but you'll lose money playing it post-flop, so push your edge while you have it.

[/ QUOTE ]

Good point Howard, I mean Spladle. One thing Howard is not particularly good at is boiling down his points to a rule of thumb, so let me try. By going all-in here, you are likely ahead and you become unbluffable. By raising against the A in our example you're likely ahead and become essentially unbluffable since you've bloated the pot enough to play on except when crushed.

[Poker CPA]

Mic Man:

“Also to anyone who is still reading this thread, say you do reraise the 885, rightly or wrongly. What then is your plan if your opponent 3-bets? Would you agree with what I outlined in my OP?”

I would call but your question just doesn’t happen in a real poker game. If you’re afraid of a 3 bet, then don’t raise. Per Sklansky, this rarely happens against an Ace, if at all. You win this pot on 4th and 5th with cards 9 thru 12. Your comfort level is much greater (which I think is SG's point) with a call against his range then a raise, and its cheaper. And its been my experience that the Ace is not folding to a RR, no matter what he has. Again I think Sklansky agrees with this too. A call here does create a “free card” situation. So you asking, IMO, “Do we compound our mistake by calling a 3 bet”. I would only put myself in this situation against a real clown, and if that was the case, I CAP.

Electrical:

Based on your previous posts from other threads, I find this comment bizarre.

“That five of the eight opponents at a full table have chosen not to play makes this presumed steal a reverse Monte Hall puzzle. Your group of opponents is still much more likely to have a good hand among them, and that group chance is condensed down to the last two opponents, one of whom has just raised into your Eights with an Ace showing.

Ignoring playing history and other information (which, as others have noted is more important), you are going to run into a decent hand there more often than at a table which is seated short-handed, because among the larger number of opponents at the table, you are more likely to run into a good hand somewhere.

If that is incorrect, please someone tell me why”

So do you think the chances of the Ace being pair are greater because of this situation? This is insane. If you were one of the 5 hands, and had an Ace, you’re folding more times than not. The chances of an Ace in the hole doesn’t change, its still 17% (I think). Its pure Randomness.

Howard:

Raising a pair of 8s, with a 5 kicker, into a Ace on 3rd is not a value bet. IMO

[PokrLikeItsProse]

If the eight raises and the ace just calls, how do you think the betting goes on fourth street?

If someone pairs their door card, they will presumably bet.

If no one pairs their door card and the ace catches a big (let's say face) card, will the ace-big usually bet out? If the ace checks, should the eight bet? What does the eight do if he bets and is check-raised?

If no one pairs their door card and the ace catches a small card and checks, should the eight bet? What does the eight do if he bets and is check-raised?

[electrical]

[ QUOTE ]
Electrical:

Based on your previous posts from other threads, I find this comment bizarre.

“That five of the eight opponents at a full table have chosen not to play makes this presumed steal a reverse Monte Hall puzzle. Your group of opponents is still much more likely to have a good hand among them, and that group chance is condensed down to the last two opponents, one of whom has just raised into your Eights with an Ace showing.

Ignoring playing history and other information (which, as others have noted is more important), you are going to run into a decent hand there more often than at a table which is seated short-handed, because among the larger number of opponents at the table, you are more likely to run into a good hand somewhere.

If that is incorrect, please someone tell me why”

So do you think the chances of the Ace being pair are greater because of this situation? This is insane. If you were one of the 5 hands, and had an Ace, you’re folding more times than not. The chances of an Ace in the hole doesn’t change, its still 17% (I think). Its pure Randomness.


[/ QUOTE ]
I'll assume you know what I mean by the "Monte Hall" puzzle, but some people may not. You are on The Price is Right, and Monte hall offers you a prize hidden behind one of three doors. Behind the doors not holding the prize are goats, and if you open a door with a goat behind it, you don't get the prize. Monte lets you pick a door, then opens one of the other doors, showing you a goat. Then he asks you if you want to switch your selection with the one remaining door. You should switch.

Your original choice of a door had 1-in-3 chance of having the prize, so Monte's original group of doors had a 2-in-3 chance of holding the prize. Monte has now condensed that likelyhood into his one remaining door. Remember that he was 66 percent to be holding the prize door at the beginning of the game. If you keep your original door, it will have a prize behind it one-third of the time. If you switch, the remaining door from Monte's original group will have a prize behind it two-thirds of the time.

Villain's hand doesn't have to be a pair of Aces, it can be any hand we don't want to play against. Let's say you have a 90 percent chance of running into such a hand at a full table (I am making this number up for the purpose of discussion), somewhere among the group of your opponents. Individually they only have a 13 percent chance of picking up such a hand, but as a group, there is a 90 percent chance that such a hand is out against you. Since 5 of the opponents have chosen not to play, we can assume they didn't hold such a hand, but when the cards were dealt, there was a 90 percent chance that you would have to play against such a hand, and until you get a walk, you are still looking at that 90 percent. Your tablemates are Monte Hall, and they have shown you Five goats.

This is related to "clumping" mentioned in Spladle's post. I believe clumping refers to the individual cards, but I was referring to the three-card hand as a unit. I am content to have been wrong in my reasoning, but I have not taken it to heart yet. I am mulling it over, as they say.

This question is like the Monte Hall puzzle in another way, in that some answers "seem" right, and even have long-winded reasoning behind them, but turn out to be wrong. Many people intuitively think that after exposing one goat, Monte's remaining door is equally likely to have a goat behind it as the contestant's door. At the beginning of the game, that was so, but after eliminating a goat, it is twice as likely to hold a prize. My equating the poker table to the Monte Hall puzzle may be one of those "makes sense but is wrong" lines of thought.