Saturday, January 21, 2012

Good read, good call, good courage

My friend Poker Lawyer (PL) is in town. She came with her friend, Texas Cardslinger (Terry). Yesterday all three of us were playing at the same table at Imperial Palace, when an interesting hand came up.

Terry raised preflop to $12. PL called. I folded. Seat 1 was a middle-aged Asian gentleman, extremely quiet, pretty tight. He reraised to $35 on the button--the first three-bet we had seen from him in a couple of hours of play. Both Terry and PL called.

The flop was uncoordinated, 10-high and rainbow. All three players checked. Seat 1 hesitated a while before checking, whereas he usually acted quickly. I thought he was having trouble deciding whether to bet.

Fourth street was the jack of spades, the second spade on the board. Terry led out betting--I'm thinking it was small, like $20, but I don't remember for sure. PL folded. Seat 1 rechecked his hole cards (about which more later), thought about it a while, then called.

Let's pause here a minute and see if we can figure out what Seat 1 has. His very rare three-bet, even with position, drastically narrowed his range: A-K, or pocket jacks or better. Maybe 10s, but that was less likely. The check behind on the flop was a complete surprise. It seemed to me inconceivable that he would check there with any overpair, giving two opponents a free card, when there was already over $100 in the pot that he presumably would like to move to his stack. The only possibilities, really, were A-K that had missed, or, much less likely, pocket 10s slow-playing top set. His call on the turn strongly suggested a draw--either A-K with the gutshot or a flush draw, possibly both.

Fifth street brought a blank--3 of diamonds, or something equally inconsequential. Terry again bet, and this time I'm sure it was $20. Seat 1 again pondered, then raised another $80 to $100.

Huh? That move surprised me even more than his check on the flop had. I was pretty thoroughly convinced that he had A-K and would fold if he failed to hit an ace, king, or spade. What in the world is he doing? And what is he doing it with? And why so much? His bets and raises had usually been on the conservative side--2-3x (though not always--once he had shoved about $55 into a $15 pot). I don't think we had seem him bluff even once, but because I could not come up with any hand with which he would plausibly take this bizarro line, I ended up concluding that my previous assessment was probably still correct, and he was bluffing a busted draw. The sealer on it was a raise size that was screaming "Go away!" rather than "Please call."

I hoped that Terry would call, both because I thought it would be correct with even weakish made hands, but also so that I would get to see whether my read of the situation was correct. I was afraid that she wouldn't, though; what I had seen of her play had been fairly timid unless she had a very strong hand. But after a little thought, she quietly told the dealer, "OK, I call."

Seat 1 said, "Good call," and flipped up his As-Ks. Terry's A-J won the pot.

A little quiet exchange between the three of us (all on one end of the table) confirmed that Terry was not just blindly looking at top pair/top kicker and deciding that was a hand she had to call with. She had followed essentially the same reasoning process I'm describing here. The inconsistency in Seat 1's pattern of betting had definitely caught her attention.

I may be making it sound like an easy call, but I don't think it was. When a player as tight as Seat 1 three-bets for the first time in hours, you naturally assume he has either aces or kings, even if he subsequently plays the hand in a way that would not be expected for those holdings. Also, the flopped top set remained a viable candidate. (10s are, along with jacks, in that range that people hate to have to play post-flop with the nearly inevitable overcards, so some players will three-bet with them in an effort to end the hand early.) A one-pair hand, even if it is TP/TK, is at best a medium-strength hand, and it would take screwing up some courage to call a big river raise. I was genuinely impressed with and proud of Terry's call.

What lessons can we learn from this hand? At least three.

1. A bluff has to tell a consistent story. His river raise was pure desperation, when he realized he had missed his draws and had only one way to win the pot--steal it. He did not stop to consider how his line would look to an opponent. Before you bluff on the river, you have to ask yourself what it is, exactly, that you want your opponent to believe. Once you've settled on that, then you have to review how the hand has been played and see if the hand you want to represent having would have been played this way. Finally, you have to assess whether your opponent is savvy enough to read the (false) clues you are giving to reach the conclusion you wish to suggest. If the story you're telling is not consistent from beginning to end, thinking players will call you from confusion, curiosity, or both. People don't like to fold, especially after they have put a lot of money into a pot. They have to be pretty thoroughly convinced that putting more chips in is throwing good money after bad. An inconsistent story will not accomplish that persuasion.

2. Consider the power of the continuation bet. Had he simply pushed out a stack of, say, $50 when both opponents checked to him on the flop, he would have seen two folds followed by the dealer pushing the pot his way. I would have concluded that he had aces or kings, and I suspect that both PL and Terry would have done the same.

3. The third lesson is about a very specific but useful tell. Note that I said he rechecked his hole cards after the second spade hit. What I didn't tell you was that on two previous occasions I had seen him do the same thing.

The first was soon after I joined the table. I bought the button, paid a little more to call a small raise, and was pleased to see a flop of K-8-8 (with two hearts), which nicely fit my 8-9 in the hole. I bet, original raiser folded, Seat 1 called. He clearly had either a king, an 8, or a flush draw. The turn was another heart. Before I acted, I noticed that he rechecked his hole cards. Aha! That strongly suggested that he had not had a flush draw that had just come in.

Here's why. Most recreational players don't memorize the suits of their cards. They retain a visual memory that they have either two red cards, two black cards, or one red and one black, but they don't know the suits with any confidence. If he had had two red cards, he would have rechecked them on the flop to see if he had a flush draw before deciding whether to call. But he didn't. He only rechecked them on the turn, suggesting that he had one red card and one black card, and wanted to see if he now had a backdoor flush draw that he might make on the river.

I had been planning to check if a heart hit, thinking that the draw was his most likely hand, but when I saw him check his cards again, that changed my mind. I bet and he folded. I think he probably had a king and nothing else.

A while later, PL had been in a hand with him in which she similarly had to decipher whether he had a flush, but again he didn't check his cards until the third spade came, implying that he did not have a made flush at that point. I don't remember enough details of that hand to recount it, and we didn't see his cards, but his subsequent fold again proved that he did not have a flush.

In the hand that is the subject of this post, his check of his hole cards on the turn implied that he had two black cards, though if a third spade had hit the river, it would have been difficult to know whether he had seen two spades or a spade and a club.

This tell is very common among recreational players, but it is by no means universal, so you have to act on it with some care. Some players do memorize their suits, but when their flush comes in they want to reassure themselves that they didn't misread their cards, so will recheck before committing a lot more money to the pot. Others, of course, are aware of this common tell and are smart enough to use it deceptively against you. Such players will check their hole cards when a third suited card hits the board in order to make you think they only now have a draw, when their flush is actually made. Or, conversely, they will check their cards when the flop has two of a suit in order to make you think they have the flush draw, so that they can bluff when the third suited card comes. But both of these species of player are much rarer than people like Seat 1 in low-stakes games.

Poker gems, #447

Dennis Miller, in HBO special "The Raw Feed":

I saw a painting of drug-sniffing dogs playing poker. It was awful. They were all ganging up on this little Shi-Tzu who was trashed on PCP, overplaying small pairs. You hate to see that.

Friday, January 20, 2012

"You shouldn't have bet that"

Aria tonight. I raised to $12 from early position with K-Q offsuit, got four callers. Flop came 9-10-J rainbow. How perfect is that?

I bet $25. Sadly, all three opponents folded in turn. I showed my hand while effecting an obviously fake pout about not having been called. (I think it's truly disgusting when people genuinely whine and complain that the pot they just won was too small, so I wanted it to be clear that my little tantrum was purely in jest.)

Three players simultaneously said variations on the title to this post: I should have slow-played it. I should have given them a free card so that one of them could catch up. I should have let them do the betting for me.

I didn't respond. I almost never discuss at the table how a hand was played; I think it's a terrible idea nearly all the time. I have no interest in giving other players new and different (and maybe better) ideas about how to play. Why would I?

I understand the impulse to play possum and pretend like you missed the flop when you actually have the nuts. And there are times when it's clearly the right thing to do. But I think that should not be the default play in that spot. Let's see how many reasons I can list for why a lead-out bet is defensible, and arguably even the best play.

1. I have a big hand, so I want a big pot. The most reliable way to build a big pot is to put money into it at every opportunity, hoping that somebody with a second-best hand will do the same.

2. If it checks around and fourth street is a queen or an 8, it vastly increases the chance that I will have to split the pot with somebody else, and/or that it will kill any shot at being called by somebody who flopped top pair or even two pair, because the straight will be both more obvious and more likely for somebody to be holding.

3. Waiting for the turn to bet means that I'll now have a flush draw on the board about 3/4 of the time. My decisions will then become increasingly complicated.

4. In just a short session, this was already the third time I had raised pre-flop and followed it up with a continuation bet, because I had gotten lucky and hit top pair or better each time. This meant that any players paying attention might think, "He can't have it every time," and decide to draw a line in the sand, calling or even raising with something like top pair, thinking that I most likely whiffed.

5. I know perfectly well that most recreational $1-3 NLHE players can't resist slow-playing monsters like this. Players tend to assume that others think and play the same way they do, which means that the nuts is pretty much the last hand they will think I have when I lead out looking so strong. They will put me on top pair, an overpair, a straight draw, or a pure bluff. Put another way, this is a situation in which I'm being tricky by being straightforward. If I'm lucky enough that one of them flopped two pair or a set, we play for stacks with the odds greatly in my favor, because they will not believe that I have them crushed.

6. Metagame considerations: The typical player in these games will tend to check both his complete misses and his strongest hands, which means that he's leading out betting with a fairly narrow range--mostly one-pair hands. I want to give the impression to anybody paying attention that they cannot reliably put me on a range of hands based on whether or not I c-bet, i.e., that I don't follow the pattern they have come to expect from other players. This is also the primary reason that I showed it after winning the pot (which I only rarely do). Showing a lead-out bet with the nuts means that a continuation bet with air later in the session is more likely to succeed, because opponents will remember that they can't exclude the strongest hands from my range based on the fact that I'm betting.

7. Because several of them will believe that slow-playing is clearly the superior play, they may conclude that this was a beginner's mistake and that I'm less experienced than I actually am. They may decide that I'm straightforward and not tricky, that I bet when I have a good hand and check when I don't. I hope to exploit, dash, and confound all of those expectations later in the session.

8. If I give a free card that pairs the board or puts a flush draw out there that comes in on the river, and I end up losing a big pot to some awful runner-runner combo, I'm going to kick myself down the nearest staircase for misplaying it, and I can't afford the resulting hospital bill.

9. One of Mike Caro's most frequent admonitions is to do the obvious thing. For example, here's a bit from Caro's Most Profitable Hold'em Advice, pages 119-120:
The simplest choice of strategy is usually the best. Exceptions are exceptions for a reason....

The reason I'm telling you this is, once you become skilled at hold'em, it's easy to justify doing the unusual. But the most obvious decision is usually correct. You should make occasional exceptions to keep observant opponents off-guard and to earn extra profit.

But, if you stray too often from what are the simplest and most obvious decisions, you're sure to sacrifice profit.
Of course, as this particular hand played out, it's likely that I could have made more money by checking. Had I known their exact cards--say, that I was up against a pair of deuces, a pair of threes, and a pair of fours--then I might have come to a different conclusion. Letting one of them catch a set on the turn would be about the only way to swell the pot under such circumstances.

But given the range of hands that people usually play against a raiser, and the fact that I had three opponents, there was an excellent chance that that flop hit one of them hard enough that he or she could justify a call or raise. The fact that it wasn't true this time does not dissuade me from the view that a continuation bet was the best move, all things considered.

I'll never live it down

You ever do something on an impulse that seems just silly and goofy and fun, and then about five seconds later think, "That's probably going to turn out to have been a really bad idea"?

(Photo credit--if you can call it that-- goes to PokerLawyer.)