Strange Pot - Checkraise that flop

[StellarWind]

This was a very enjoyable article on the dynamics of live play.

I don't see the flop decision quite the way Nate does. I think checkraising is absolutely mandatory. It's true as he observes that a checkraise cannot protect him from a good flush draw. But what about a bad flush draw? If you can isolate on Ed there is a very good chance that life will go on after the fourth spade hits because he didn't happen to get one.

You don't want to give the pro a cheap card with 5 [img]/images/graemlins/spade.gif[/img] 5 [img]/images/graemlins/heart.gif[/img]. He has twelve outs versus you counting the backdoor straight. But he will be hard-pressed to call if you confront him with two bets. He's making a gigantic Sklansky mistake by folding the last spade but how could he know?

Then there is Daniel with A [img]/images/graemlins/club.gif[/img] 4 [img]/images/graemlins/spade.gif[/img]. That's twelve outs and another huge Sklansky mistake. By your description he sounds like exactly the sort of player who will probably fold for two bets but call one.

That Ed might make a strange 3-bet is not a reason for concern. It's going to be hard to get away from this hand unless a third player shows strength. Ed should feel free to put in as much action with worse hands as he wants [img]/images/graemlins/grin.gif[/img]. Furthermore, I'm sure Ed's tendencies are not lost on the pro or probably even Daniel. The last thing they want to do is put a loose call in between you and Ed. They know full well this could come back for two more bets.

The best thing about the flop checkraise is you are likely to have the best hand. This is a situation where it is worth a sacrifice to protect your hand but you are probably getting value on this raise. At least it's close.

Principle: Evaluating outs on a monotone flop is extremely difficult. Large Sklansky errors in both directions occur when players are confronted with pressure. Sometimes players sensibly call multiple bets to draw dead. Othertimes obvious folds are actually huge live draws or even the best made hand.

[nate1729]

StellarWind --

I'm glad to see that you read my article and thought about it. I hope you found it worthwhile.

I thought (and hoped) that the flop action would be controversial and thought-provoking. In my opinion it illustrates yet another valuable principle I didn't mention explicitly in the article: often in poker we need to remember that a parlay of individually likely events might not be sufficiently likely.

Why would we want to checkraise the flop? The preliminary answer is that we want to eliminate hands we want to eliminate and extract bets from hands we want to extract bets from. This might take the form of protecting a best made hand, but that requires both that our hand is ahead and that confronting an opponent with two bets will knock him out.

I agree that it is reasonable to guess that I have the best hand. But I am laying big odds by checkraising out of position: I am at an informational disadvantage, and the act of raising involves the laying of odds for an opponent. The main benefit of checkraising would be to make an opponent fold a medium spade incorrectly, but that requires that I be ahead, that the best spade be medium, and that that spade be folded. This parlay, though composed of reasonable events, is not likely enough in combination to warrant the risky play of checkraising. Overcards are another possible hand we'd prefer not to give a Sklansky-profitable call to, but the pro is unlikely to continue on overcard value and Ed isn't folding for one more bet. So the overcard-elimination value is almost entirely the parlay involving the case when Daniel will fold for two bets but not one, which does not contribute a lot of equity to the play.

What makes this hand interesting, I think, is that I would have *loved* to have received the benefits of the flop checkraise: I would have been remiss not to think of all the weak-overcard and medium-spade hands I might want to face with two bets. But the costs of the play were too great. Remember that there is no guarantee I'm ahead; that my hand, if ahead, is very vulnerable; that I'm at an informational disadvantage; and that raising is likely to manipulate the pot size into one that will let me easily make Sklansky mistakes on later streets. Also, remember that hand-protection does not end on the flop: when the turn came good I was able to make a bet that was much more powerful than it would have been had I raised the flop.

Thank you for taking the time and effort to post. I'd love to hear anything else you, and anyone else, might have to say about the hand, especially if it's in disagreement with me.

Also, please note that there is a typo in the article and that Ed did not pair on the flop; the 6[img]/images/graemlins/spade.gif[/img] in his hand should be the 7[img]/images/graemlins/spade.gif[/img]. He was continuing with an overcard and a small one-card spade draw.

--Nate

[Position]

Very neat conversation & article!

[bobman0330]

Another important factor is that the pot is so small that shutting a flush draw out for one street is not all that important. Even if Nate was 75% sure that he was ahead, and that his raise would protect him from a 4th spade, it's still doubtful that it's worth investing another small bet to raise.

[StellarWind]

[ QUOTE ]
Another important factor is that the pot is so small that shutting a flush draw out for one street is not all that important. Even if Nate was 75% sure that he was ahead, and that his raise would protect him from a 4th spade, it's still doubtful that it's worth investing another small bet to raise.

[/ QUOTE ]
Let's not lose perspective. If Nate is 75% likely to be ahead then this is a tremendous value raise versus three opponents and increasing our winning chances is merely a very nice bonus.

I suspect 75% is a rather high estimate however.

The pot is not so small when you consider implied odds. There are currently 2.5 BB in the pot. This will increase to 3.5 BB after Hero and (assume) one more player call the flop. Hero intends to bet the turn which if called in two spots will increase the pot to 6.5 BB. Toss in 2 BB for a river bet and one caller and we have 8.5 BB. Obviously it could be less but it could easily also be more.

This is going to be a 7-10 BB pot when someone wins it. That's why it's not a big deal to invest an extra 0.5 BB on the flop to improve our chances.

It's also important to realize that "invest" is not synonymous with "spend" or "lose". If everyone were expected to call then we would be getting 3-1 odds on our raise. We would only need to win 25% of the time for our raise to be free and anything beyond 25% is immediate profit.

I think our pot equity at least meets this 25% threshold. I'm not concerned about the high cost of raising because there is no cost. This is a value raise.

The obvious objection is that we may not get three callers. But our likely best hand is so precarious that virtually any fold is more valuable to us than a caller.

Example: we raise, BB folds as he always intended, Daniel would have called one but folds to two, and Ed calls. Our raise cost us 1 SB and we get these benefits:

1. The pot is 1 SB bigger than it would have been (1 from us, 1 from Ed, less 1 that Daniel keeps instead of calling just one). We are the heads-up favorite and a big piece of that extra SB is ours.

2. Our chance to win is substantially increased by Daniel's departure. This is a process that favors weak made hands over nut draws. If Ed is banking on A [img]/images/graemlins/spade.gif[/img] 3 [img]/images/graemlins/heart.gif[/img] then his nut flush and pair draw give him a very good chance to win. But removing Daniel does almost nothing for that winning chance. Almost all the times Daniel would have won will go to us. Ed would be much better served by Daniel staying in the pot and donating money.

Overall in this example we lose less than 1/2 SB of immediate EV and gain a substantial increase in our chance of winning a medium-size final pot. That is a very favorable trade.

This is a classic example of the Fundamental Theorem of Poker breaking down in a 3-way pot. The junky best hand wants Daniel to fold his outs while the flush draw wants Daniel to stay and donate his money. Whether Daniel is making a mistake by calling or folding is beside the point. [Lest anyone get the wrong idea, Sklansky clearly states that his theorem does not always apply multiway.]

Another perspective on checkraising the flop is that from a purely protection view it is by far the best. The pot is as small as it is ever going to be. Ed is still live behind the potential coldcallers and will also be acting last after our threatened follow-up turn bet. If the middlemen won't fold for two bets now, then they can never be made to fold. This checkraise is the biggest bomb at our disposal and if it won't work then our hand is beyond protection.

Remember that any turn protection is inherently only half as good because it only works against one card instead of two.

[nate1729]

[ QUOTE ]
Let's not lose perspective. If Nate is 75% likely to be ahead then this is a tremendous value raise versus three opponents and increasing our winning chances is merely a very nice bonus.

I suspect 75% is a rather high estimate however.

[/ QUOTE ]

When we're ahead we prefer to gain value, of course. The problem is that just as effective odds differ from pot odds, the weighted payoffs when we're ahead and behind require a greater degree of certainty than the immediate numbers would indicate. If we're behind, betting can be very costly, and lead to a precarious pot size. Meanwhile, if we're ahead, we're not ahead by too much, in the weighted average case. Even if we're winning currently, we do not have the same equity that top pair normally has.

[ QUOTE ]
The pot is not so small when you consider implied odds. There are currently 2.5 BB in the pot. This will increase to 3.5 BB after Hero and (assume) one more player call the flop. Hero intends to bet the turn which if called in two spots will increase the pot to 6.5 BB. Toss in 2 BB for a river bet and one caller and we have 8.5 BB. Obviously it could be less but it could easily also be more.

This is going to be a 7-10 BB pot when someone wins it. That's why it's not a big deal to invest an extra 0.5 BB on the flop to improve our chances.

[/ QUOTE ]

I'm glad that you look at the big picture of a hand, but I fear you might be assuming some things you shouldn't. Part of the reason you predict a big pot is that you're planning to raise the flop. Bets beget bets; once the money's in there the pot's more worth fighting for. That the pot might get big is bad news, not good news, for a weak made hand suffering from a severe information defecit. And, of course, some of the predicted 7-10 BB is invested by us, and (more to the point) if the pot gets that big and we stay in it we're unusually likely to have gotten some money in badly. T9o on a 9-high monochrome board is not a hand that can confidently claim a good equity share of the entire projected size of a pot.

[ QUOTE ]
It's also important to realize that "invest" is not synonymous with "spend" or "lose". If everyone were expected to call then we would be getting 3-1 odds on our raise. We would only need to win 25% of the time for our raise to be free and anything beyond 25% is immediate profit.

I think our pot equity at least meets this 25% threshold. I'm not concerned about the high cost of raising because there is no cost. This is a value raise.

[/ QUOTE ]

I like this thought; this is one consideration that makes this decision difficult. As I've said, however, I think the real weighted equity we're facing, and the stark contrast between immediate and effective odds here, makes it misguided to project enough of an immediate equity gain to swing the decision.

[ QUOTE ]
The obvious objection is that we may not get three callers. But our likely best hand is so precarious that virtually any fold is more valuable to us than a caller. [...]

[/ QUOTE ]

The same precariousness is what makes us hesitant about raising for immediate equity. Remember also that the range of hands to which we can denying a profitable call is smaller than it might seem: bad spades and overcards in Daniel's hand would be our targets, but this is not likely enough to support the risk of raising. (Also note that the difference between one and two bets is less likely to cause Daniel to make a Sklansky-incorrect fold than other players. He's a hand-reader, not an odds man.)

[ QUOTE ]
Another perspective on checkraising the flop is that from a purely protection view it is by far the best. The pot is as small as it is ever going to be. Ed is still live behind the potential coldcallers and will also be acting last after our threatened follow-up turn bet. If the middlemen won't fold for two bets now, then they can never be made to fold. This checkraise is the biggest bomb at our disposal and if it won't work then our hand is beyond protection.

Remember that any turn protection is inherently only half as good because it only works against one card instead of two.

[/ QUOTE ]

I think that your argument applies very well to many common hold'em situations, but one thing that makes this hand interesting to me is that the turn bet is comparatively more powerful than it often is. Many hands have proper odds to call a double bet on the flop, but if the turn blanks off, the turn bet will be more powerful if we're confronting opponents with 9:2 odds instead of something more attractive. (Remember that we're squeezing Daniel between our bet and Ed on the turn, so 9:2 might be less attractive than it seems.)

Also, we're not betting every turn. Just calling makes it easier to get away from a hand we're pretty sure is beaten if the turn comes ugly.

Even though flop protection might prevent an opponent from seeing two cards instead of one, turn protection is often superior. In this case, we can exploit the increased betting limit and the fact that our hand might be strong enough to protect on the turn (and not on the flop.) And, of course, Daniel and Ed aren't the only ones dodging those cards; we are too.

Thanks again for taking the time to think about my article. Your comments have helped me consider new aspects of the situation. If you continue to disagree with me (and even if you don't) I hope you share any more thoughts you might have about the hand.

--Nate

[StellarWind]

This is getting long with many tangents. Let me summarize my big picture view and then just comment on a few specific points from your post:

1. Driving players out on the flop is twice as likely to prevent a suckout as driving players out on the turn. This creates a huge incentive for attacking now.

2. Checkraising confronts the two remaining players with immediate 7-2 pot odds AND the threat of an Ed 3-bet/Hero cap AND the threat of a Hero turn bet toward Ed on the Button. If this doesn't dislodge the middlemen then it is unlikely that anything can.

3. I'm fully aware that our pot equity does not nearly match up to our immediate chance of having the best hand. We will probably lose more often than we win. But it's still a top pair, two players checked the flop behind us, and the Button-bettor is a known nitwit. We will manage to win a fair amount, sometimes by improving our hand. Given that we are getting 3-1 odds on our raise we should win often enough for the immediate value to be roughly neutral or a little in our favor. The cost burden of protecting our hand with a checkraise simply doesn't exist. The best protection play money can buy is free.

[ QUOTE ]
Part of the reason you predict a big pot is that you're planning to raise the flop.

[/ QUOTE ]
This is an incorrect restatement of my previous post. My estimate of 3.5 BB for the flop reflects four limpers, Ed's flop bet, and two flop callers including Hero. This usually leads to a 7-10 BB final pot depending on how the further play goes.

Estimates of the final pot size are inherently imprecise but that's OK because precision is not need to play the flop. It's sufficient to recognize that 8 BB is a reasonable rough estimate and that Hero's raise/call decision can't change that very much on average. Sometimes a flop raise will add a BB or so, but if the flop raise drives out players the final pot will probably become *smaller*.

[ QUOTE ]
Remember also that the range of hands to which we can denying a profitable call is smaller than it might seem: bad spades and overcards in Daniel's hand would be our targets

[/ QUOTE ]
Protection is not just about Daniel; there is also the BB. I'm a "solid professional" and I assure you that the list of dangerous hands with which I will call one bet but not two is very long. Most hands with two useful cards including a spade will do nicely. Q [img]/images/graemlins/club.gif[/img]T [img]/images/graemlins/spade.gif[/img] for example. Or my previous example 5 [img]/images/graemlins/spade.gif[/img]5 [img]/images/graemlins/heart.gif[/img]. It's worth a SB to try and get lucky and have a shot at the 8 BB, especially playing against Ed. I strongly consider that on this action the turn may check through and give me an invaluable free card.

[ QUOTE ]
Also note that the difference between one and two bets is less likely to cause Daniel to make a Sklansky-incorrect fold than other players.

[/ QUOTE ]
This is a crucial point from a theoretical perspective. The Fundamental Theorem of Poker does not apply here. I very much want to help Daniel make correct folds of hands with a few outs. When he sucks out with some random overcard, the person who loses the pot is almost always me because I'm the guy trying to win with a pair of nines and no spade. But when Daniel calls and misses I only see a fraction (maybe 50%) of his money. The rest goes to all of the big hands Ed either has or is drawing toward. So Daniel's incorrect calls cost him money but they often cost me money too. Ed is the guy collecting the money. The third player in the pot causes the theorem to not apply.

[ QUOTE ]
That the pot might get big is bad news, not good news, for a weak made hand suffering from a severe information defecit.

[/ QUOTE ]
This general concept is referred to in several places and it doesn't seem to apply very much. Calling doesn't usually help us play the rest of the hand better. It doesn't affect the pot size enough to matter and it doesn't teach us anything we can use.

Actually the opposite is true. Driving players out now makes it easier to play the big streets. Getting coldcalled would also be more informative than simply watching them call one bet. Raising creates some clarity concerning BB and Daniel.

As for Ed, I really feel that as long as it's just Ed sitting on the Button and the remaining players pose no threat, we have to call him down almost no matter what. We may have the best hand, we may have outs, and even a spade wouldn't convince me not to take a look heads up. If that means we make losing bets on the turn and river, so be it. Couldn't be avoided by any logical play.

[nate1729]

I like many of your comments, even though we disagree on the conclusion. A few things I'd like to mention:

I don't think it's fair to say that flop aggression is twice as likely to prevent an outdraw as turn aggression. On the turn, we can utilize the larger bet size and the potential to be a bigger favorite.

Also, remember that we would prefer to eliminate the big blind if:
-He would call one bet but not two;
-The one-bet call takes money away from us.

This might be true if, say, he held a medium spade and there were no spades in Daniel's or Ed's hands, but that itself is quite a parlay. Estimating the probabilities of various cards being in various places will show, I think, that confronting the pro with an extra bet here gives us very little equity due to extra protection.

While much of your analysis seems well-founded, I might caution you about one thing: at times you seem to be hoping for folds and at others you seem to be hoping for calls. Depending on the holdings, sometimes we'll want calls and sometimes we'll want folds, and there are even some perfectly plausible distributions of hands for which it's better to checkraise, but I remain confident that, given the information we have, calling and evaluating the turn is better.

Thanks again for the comments. I hope you share any more thoughts you have on the hand. Also, I hope other people weigh in, but that might be wishful thinking.

--Nate

Excellent thread. One of the best I've seen in the two years I've posted on the forums (not reflected in my member date since I just changed my name).

I have little to add, except to articulate one idea more clearly (perhaps) than it has been, although it's certainly implicit in many of the things that have already been said. That is, that by c/ring the flop and driving out opponents -- and I do think there's a good chance of driving out 1, 2, or even all 3 opponents(note description of the flop bettor) -- the hand becomes much easier to play on later streets. With 3 or 4 opponents, half the deck are scare cards. If one of those hits, the hand becomes very hard to play, particularly OOP. Of course, this value depends, like much of the rest of the points made in the debate, on how successful a flop c/r will be in thinning the field.

[RaiNz]

Nice article and great continuation thread. =)