Thursday, December 31, 2009

Analyzing NLHE: TAP Concept 60

This is the final installment of my series looking at the sixty concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller. Soon I can get back to telling stories from the felt, but I also want to revisit some of these concepts, as well as the project as a whole, in future posts.

Concept No. 60: If someone bets into several players, and you have a hand that is somewhat likely to be best, but unlikely to improve, you often have to fold.

Yes. Similar ideas have been discussed in topics a few times before (particularly concepts 18 and 32, and to some extent concept 50, while concept 6 sort of argues the other way). I agree with the author's closing statement that this may be the single most profitable concept in the book for many readers. The reason is that our human nature (mostly due to the regret aversion and confirmation biases) begs us to call in this situation. For example, if we call with top pair and a good kicker and someone raises behind us, or if we have to fold on the turn or the river, this is easily written off as bad luck because we couldn't have anticipated those things, and we don't regret our call; hence, we will probably call again next time. Of course, the fact that there were so many possibilities we could not anticipate is what should have made us decide to fold. On the other hand, if we were to fold and it turns out we would have won the hand, it is psychologically devastating, and we are likely to call in such a situation next time.

With several players yet to act, two cards yet to come, and very little chance to improve our hand, we lack all sorts of crucial information that we would need in order to continue the hand profitably. Certainly, it depends on the situation (perhaps your hand is a big favorite, or your opponents are very passive), but many players call far too liberally in such situations.

Wednesday, December 30, 2009

Analyzing NLHE: TAP Concept 59

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 59: Don't help your opponents play correctly.

I agree with this one, and it seems incontrovertible, but there is at least one example of advice that suggests you should help your opponents to play correctly. I think it is worth considering.

Barry Greenstein has a total of eight "play lessons" in his book Ace on the River. Lesson 5 says the following: "If your bets define the strength of your hand, it may make decisions easier for you on later streets." The idea is that if you confuse your opponents too much, you will have trouble interpreting their plays later in the hand, and you won't know how to react. I think Dan Harrington may have similar advice in his book, but I can't quite remember. Anyway, I think Sklansky and Miller's concept holds up despite Greenstein's lesson. You can use Greenstein's advice for your made hands but still make it difficult for your opponents to play correctly if you play bluffs in the same way. However, bluffing actually seems to go against Greenstein's Lesson 5. He suggests you let "your bets define the strength of your hand." This implies that you should not let your bets "equivocate" (my word, not his), which is what happens if your bets could mean either a bluff or a made hand. Bluffing occasionally seems to go against Greenstein's Lesson 5, and so it should not be taken too seriously. I have to take Team Sklansky's side on this one.

Okay, just one more concept to go. Hopefully, I'll finish this year after all!

Sunday, December 27, 2009

Analyzing NLHE: TAP Concept 58

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 58: Any strategy relatively close to a game theoretical strategy is at least almost as good as the optimal strategy, and sometimes it's better.

Sklansky and Miller are a little messy with their terminology here, but I think they get their point across. The way I interpret this concept, I agree with the authors. Here's my interpretation: even if you don't know precisely what the optimal play is (and you probably don't), it's still worth trying to approximate it. Playing a strategy that is close to "optimal" (that is, unexploitable) is almost always better than playing a strategy that is not close to optimal. Furthermore, if you have identified exploitable weaknesses in your opponents, it can, in fact, be better (in terms of EV) to deviate from the optimal strategy. One problem I have with this idea is that there probably is no mathematically "optimal" strategy in games with more than two players... but I think the main idea still holds up.

I disagree with the last two sentences of the books discussion. It says: "If you plan to make a play that will give away your hand, choose a different play occasionally and make the same play sometimes with a different holding. If you do this consistently as you play, you'll usually do even better than the game theoretical strategy."

First, I don't agree that you should necessarily change anything just because you plan to make a play that will give your hand away. Hopefully, you've chosen this particular play because it maximizes your EV, so changing it can only reduce your EV. S+M suggest either choosing a different play occasionally (but assuming that you've chosen the play that maximizes EV, changing will lose you money), or making the play sometimes with a different holding (I suggest you only do this if you can do so without losing any EV with this other holding). I've discussed before why I don't like the idea of making certain plays "occasionally."

Second, S+M suggest that as long as you think about what your range and your opponent's range is, and you mix up your play, you can do better than the game theoretically optimal strategy. I doubt that this is true unless your opponents are tremendously weak. The problem is that it's impossible to avoid making mistakes, so very few people would ever be able to do better than if they could consistently play an "optimal" strategy. However, if your competetion is weak enough, you can probably find enough opportunities to exploit them to make up for all of your mistakes.

Analyzing NLHE:TAP Concept 57

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 57: Don't be fooled by players who have giant amounts of cash in front of them.

Sklansky and Miller suggest in their discussion that players who buy in for exorbitant amounts often do so in order to trick other players into thinking that they will be willing to lose their entire stacks, thus enticing these other players to play too loosely. This doesn't fit my experience at all. In my experience, players who buy in deep do so either for mundane strategic reasons (they feel they can outplay the other players at the table with deep stacks) or in order to gain a psychological advantage (either they want to intimidate the shorter stacks, or they want to avoid feeling intimidated). I don't think anyone buys in deep in order to trick players into gambling with them more.

The issue of whether it really is strategically advantageous to buy in deep is something I've addressed in posts a few times before. While many people are intimidated by playing against bigger stacks and believe they are at a disadvantage, I believe it is actually an advantage to have a smaller stack at a full table. It's possible that in certain games you will have higher expectation if you have a big stack, but this is only if you are far superior to your opponents. If everyone plays about the same, players with smaller stacks should be expected to do better.

Thursday, December 24, 2009

Analyzing NLHE: TAP Concepts 55-56

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 55: Unlike limit, limping first in on the button is frequently correct.

In a pure NL holdem game, I think this is probably right, mostly for the reasons given by the authors. Specifically, I agree that making the blinds fold is not as profitable in NL, because the blinds are much smaller compared to the normal raise size and final pot size. This probably makes some hands worth limping with, but only against certain opponents.

In practice, however, I practically never limp first on the button. This is largely because of the way pots are raked in my normal game (and most other games in Los Angeles). The casino will take out $1 if there is no flop, but they take $6 if there is a flop ($1 of which is for the jackpot drop). Winning the pot after the flop therefore costs me $5, which is a significant amount when we're looking at a profit of only $10 or $20.

Concept No. 56: Pot odds (as opposed to implied odds) matter a lot less in deep stack no limit than in limit.

Yes. This is central to the difference in the strategy for the two games. Bet sizes in no-limit tend to be much larger in proportion to the pot than in limit, so future bets will have a much larger impact. Since pot odds do not take future betting into account, they aren't as relevant in no limit poker. In no-limit, it's important to try to estimate implied odds rather than relying on pot odds to help make your decisions.

Sunday, December 20, 2009

Analyzing NLHE:TAP Concept 54

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 54: Checking to induce a bluff is a significantly stronger play in no limit than it is in limit.

Instead of taking this concept's wording completely literally (it's ambiguous how we should define the "strength" of a play), let me rephrase it in a way that I think reflects the authors' point more accurately: "Checking to induce a bluff is the best play much more frequently in no limit than it is in limit." I think this is right.

I should point out that Sklansky and Miller are double-dipping here: we've already discussed this idea in Concept 33. As I implied in that post, I personally don't usually think of checking on the turn with a moderate hand as "inducing a bluff." Rather, I think of checking as moving my hand range toward the weaker end, which then forces me to call on the river if my hand is on the stronger end of my range. I will admit that "induce a bluff" is a simpler way to get this idea across.

Saturday, December 19, 2009

Analyzing NLHE:TAP Concept 53

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 53: In heads-up pots, whether you are first or second to act is more likely to affect your decision in no limit than it is in limit.

I think Sklansky and Miller are having another issue with semantics (or perhaps it's my fault for taking them too literally), but any way I look at it, I can't see the use of this concept. It's also quite awkward trying to judge whether it's true. I happen to think it is true as stated, but not for the reasons the authors were probably thinking of. This post is an almost purely academic analysis of this concept's claim, without much poker insight, so let me apologize. I personally think it's rather interesting, though. First, let's look at the example Sklansky and Miller give.

Their example is that if you have the nut flush draw in limit holdem, you will probably want to bet regardless of whether you are in position. In no-limit, though, you should be much more willing to bet if you are first to act. If you check, they say, you are likely to have to face a bet anyway, so by betting out, you can set the bet-size yourself. If you are second to act, you might want to check to give yourself a free card and a chance to win a big pot.

Well, that is only a specific example. Let's take a step back and try to answer the question: Should your position be more likely to affect your decision in limit or in no limit? Of course, your position is only one of many factors affecting the decision of whether to check or bet. Your hand range, your opponent's range, and your opponent's tendencies are usually the most relevant. Let's imagine holding all of these other factors constant, and distill your "position" down to the following. If you are "first to act" and you check, your opponent has the opportunity to bet. If you are "second to act" and you check, the betting round ends. In either case, if you bet, your opponent has the opportunity to raise. So, in this formulation, you should always (whether you are playing limit or no limit) be more inclined to check if you are second to act, because you gain the added benefit of denying your opponent the opportunity to bet. So, the question simply becomes: Is this added benefit more valuable in no limit or in limit? I think the answer must be "no limit." Giving your opponent the opportunity to bet tends to be much riskier in no limit than in limit, because a bet can be much larger in proportion to the pot. So, denying your opponent this opportunity has greater value in no limit. Thus, I would have to agree with this concept's claim that, in no limit, your position is somewhat more likely to affect your decision than in limit.

In practice, though, we are more likely to bet when we are second to act than when we are first to act. When we are second to act, this means our opponent has checked, which means his range is weaker. When our opponent's range is weaker, we should be more inclined to bet. In my analysis in the previous paragraph, however, we held these hand ranges constant, which led us to the awkward and counter-intuitive conclusion that you would rather check if you are second to act than if you are first to act. This conclusion led me to agree with this concept's claim. However, this is, of course, not the point the authors were trying to make with the concept. This is made abundantly clear from their example about how you would play a nut flush draw, which incorporates plenty of good ideas about poker, but nothing that really addresses the concept's claim. It attempts only to show one specific case where the concept holds, and it fails even that because it does not hold the opponent's ranges constant (the opponent's range weakens after he checks).

I won't address the ideas that Sklansky and Miller were trying to get across in this concept, which they revealed in their example, except to say that I essentially agree with them.

Friday, December 18, 2009

Analyzing NLHE:TAP Concept 52

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 52: The play of check-raising to knock people out, an important tool in limit, should rarely be used in no limit.

I think this is true. Sometimes when I check-raise on the flop in no limit, I do so with the intention of knocking out some middle-position players, but this is almost never the primary objective of my check-raise. I will either have a very strong hand or a semi-bluff.

In limit, it's sometimes better to check with a low or medium pair, and then raise if a late position player bets. This play is very likely to force players who have overcards to your pair to fold, even if they may have called had you bet out instead of check-raising. If the late position player was betting without a pair, getting these players out substantially improves your chances of winning the pot. In limit poker, as we discussed in Concept 9, winning the pot is the main focus.

In no limit holdem, this play makes far less sense. Here, you don't need to count on someone to bet so that you can check-raise in order to confront them with two bets at once. If you want to deny a middle-position opponent the right odds to call, you can just bet more yourself right out of the gate. Also, in no limit, a check-raise tends to be much more expensive in relation to the pot. Defending a relatively small pot with a large bet should not generally be your main objective (concept 9 again). When you do want to knock out middle position players with your check raise, it's either because you are semi-bluffing with a small flush draw and you want possible bigger draws to fold, or you have a made hand and you want strong draws to fold. It's not like in limit where you are hoping to make players with overcards fold. In no limit, these players will already be folding to the original bet (unless the bet is unusually small), and you won't need to raise it.

Wednesday, December 16, 2009

Analyzing NLHE:TAP Concept 51

If I'm going to get through these by the end of the year, I'd better try to get a few done during the week. So, here is my analysis of the next concept from the end of No Limit Hold 'em: Theory and Practice. Sorry to those of you who are more interested in my narratives from the poker tables! I'll try to get one of those out in the next week.

Concept No. 51: In tournaments, other things being relatively equal, prefer small river value bets that will often be called to large river value bets that will seldom be called. Put another way, if a smaller bet has a bit less EV, it is still right to make it in most tournament situations.

I haven't played tournaments in years, so I'm no expert, but this concept is seems obviously correct. It's a specific example of a more general tournament concept: it can be worth giving up a little bit of EV in exchange for less volatility. The reasoning behind this idea is that since you can get paid in a tournament even if you do not come in first, it's worthwhile to hold on longer than other players even if your chances of coming in first are diminished slightly. This more general idea would have made for a much better concept idea (ie, more insightful, useful, and generalizable), except perhaps that most tournament poker players already understand this idea.

I'm not sure why, but in these concepts, Sklansky and Miller often forgo more general concepts like this in favor of specific examples that the reader might have difficulty generalizing and incorporating into his game. Maybe using specific examples is more effective in marketing poker books; Sklansky is certainly experienced at writing poker books that are popular.

Also, does anyone else find it strange that the authors are bringing up tournaments in this one concept for this rather obscure point? I guess tournaments are fair game since they are a common form of "No Limit Hold 'em," but I had thought they were focusing on cash games.

Tuesday, December 15, 2009

Analyzing NLHE:TAP Concept 50

In which I analyze the fiftieth concept at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 50: If someone bets the flop and gets two or more calls, anyone who bets a significant amount on the turn should get respect.

This seems to be happening a lot, but once again, I don't like this concept even though I do agree with it. A big bet into a multiway pot after significant action on the flop is a very strong play, and the bettor should have a very strong hand if he expects to make a profit.

I have two main gripes with the way this concept is presented. First, the concept is a very specific example of a more general idea, which is that when someone bets with the expectation of being raised or called, his hand range must be very strong. I think the book's "concepts" are more useful when they are more general, because they can be applied to various situations.

Second, the example in the discussion is very unconvincing, a sort of "straw man" argument. If the authors felt confident in the generality of their concept, they would set up an example where all factors would suggest that you should not respect the bettor's bet, except for the reason given in the concept. On the contrary, in the example in the book there are many reasons why you should respect the bettor's bet. Sklansky and Miller try to use this to their advantage, saying, "you are out of position with a hand that is unlikely to improve..." However, this just muddles their argument. Thus, their conclusion that "you ... should frequently fold" to the bet on the turn is true, but it doesn't bolster their main argument in the least. They have completely failed to make a logical argument in support of their concept.

Recently, I haven't been doing EV evaluations like I did at the beginning of this project, and that is partly do to my own laziness. However, I think Sklansky and Miller also were lazy in formulating these concepts and discussions. It's frustrating, but enough of them are thought-provoking that I think this is still a worthwhile project. Only ten more to go!

Monday, December 14, 2009

Analyzing NLHE:TAP Concept 49

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 49: If someone makes a big bet into multiple players, typically he will have a good, but not great, hand.

This is very similar to concept 35, and once again, I don't have anything very insightful to say. I think the claim made by this concept is usually true, but it's not very reliable. Different people play differently at different times.

Saturday, December 12, 2009

Analyzing NLHE:TAP Concept 48

From No Limit Hold 'Em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 48: Often make small bluffs (about one-third the size of the pot) in multiway pots when it appears no one has hit the flop. Balance those bluffs by also sometimes making small bets with good hands.

This is too general. It's also rather vague as to how it can "appear" that no one has hit the flop, but I'll take this to mean that most players have checked (as in the discussion in the book).

Yes, there are situations where a small bluff into a multiway pot can be profitable, but you will probably lose money if you apply this advice indiscriminately. Whether such bluffs are profitable depends on your opponents and on the flop texture. There are lots of players who will call small bets with weak hands or will check-call with relatively strong hands when they are out of position. Also, bluffing small on flops with straight draws usually won't work, because people can call a small bet with a gutshot. In general, your best chance will be on boards with the fewest draws, such as paired boards or A- or K-high boards. Boards with these textures are also the best for making small bets with your good hands, so it will be difficult for observant players to exploit you.

You also might want to use your hand as a guide to when you should bluff. If you have a backdoor draw or overcards, your bluff will have some of the benefits of a semi-bluff, since you will have at least a slim chance to outdraw your opponent if you get called.

Monday, December 07, 2009

Analyzing NLHE:TAP Concepts 46-47

From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 46: Don't just think about what you put your opponents on. Think about what they put you on also.

Yeah, this is what I'm always talking about. You need to think about what your range is and what your opponent's range is. That is the basis of game theory.

Concept No. 47: If it's clear your opponent has a hand at least worth a call, but he raises instead, it's almost never a bluff.

I think this is an overstatement to the extent that it is misleading. Also, it is very difficult to tell when your opponent has a hand "at least worth a call." However, this does occasionally arise if your opponent played in a way that was probably a made hand but if not, must have been a draw. If the draw comes in, you might want to bet to get value from his previously-made hands, but you'll have to fold if he raises. I would argue that your opponent could still be bluffing in this situation, but it's unlikely enough that you should usually still fold.

Suppose that you identified a situation where your opponent certainly had a hand that was at least worth a call, and he has raised you. Sklansky and Miller say it is almost never a bluff. However, it could still be a bluff if your opponent is overly aggressive. He would likely be making a mistake, but mistakes like this are not that rare. Moreover, if he knows you will fold to this bluff (which S+M are recommending to you), it means his bluff was actually the correct play after all.

This last idea is a bit convoluted, but I have used this against my more astute opponents. For example, I played a hand that went something like this: suppose you have AsJh and the board comes Jd 5s 4s. There is a lot of action on the flop. The turn is the Ace of clubs and you get check-raised. I think you are probably behind here, but you might still want to call because you might be ahead, and you do have four outs against a small set. If the river is a spade and your opponent bets, you can raise as a bluff. Your opponent will figure you would only do this with a flush, and fold his set. He might even fold a small flush or straight.

Friday, December 04, 2009

Analyzing NLHE:TAP Concept 45

After this post, I'll be three-quarters of the way through the concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.

Concept No. 45: Know when a hand (even a good one) has more value as a bluff catcher.

This is good advice (after all, it's always good to know things), but the Sklansky and Miller don't give a good explanation of how to identify this type of situation.

The authors merely point out that sometimes you should check with your good hands, especially if you think it's likely that your opponent has a hand that's too weak to call with. This is true, but it's a meager assessment of the circumstances that make it profitable to check-call with good hands rather than to value bet with them. It would be helpful to discuss these circumstances.

For one thing, we need to consider our opponent's style. As I summed up in my analysis of Concept 10, if you've identified any weakness in your opponent's play, try to put him in situations that will allow you to exploit this. There are two types of mistakes your opponents might make that would turn some of your strong hand into bluff-catchers.

The most common weakness that you will be able to exploit is if your opponent bluffs too often. Against such opponents, you should be much more willing to check with your strong hands and use them as bluff-catchers. Another type of opponent is so tight that he will not pay you off if you bet. These players are great to bluff against, but if you hold a strong hand, you want to get him to put more money in the pot. So, it may be better to just check and call against extremely tight players. Often very tight players will be very passive, though, so this opportunity seldom arises.

There are also circumstances that have nothing to do with your opponents weaknesses that can make many of your hands bluff-catchers. Generally, this will occur if your opponent likely knows whether he has your hand beaten. This can happen if your hand-range is extremely narrow (usually because of the type of board and the betting to this point, or maybe you accidentally exposed your cards), or if your opponent's range is polarized (that is, he probably holds either a very strong hand or a very weak one). In such a situation, your only viable options are to check or to put in a small blocking bet. Let's look at some situations where your opponents range might be polarized.

In my discussion of Concept 44, I noted that when your opponent's hand range becomes polarized when he bets on the river: either he has a reasonably strong hand or he is bluffing. If you hold a hand in between these ranges, it doesn't matter much exactly what you hold; all that matters is your opponent's bluffing frequency. If he bluffs too much, all of your hands in this middle range are good to call with. If he bluffs too little, they should all be folded (or, possibly, raised as a bluff).

Another common situation where your opponent's range is polarized is if he is very likely to have had a drawing hand, but you're not sure if he has hit his draw on the river. Your opponent knows if he has you beaten because his hand is polarized. Either he hit his draw or he didn't. Unless you can beat your opponent even if he hit his draw, your hands now turn into bluff-catchers.

There are other probably other situations where your hands might be best used as "bluff-catchers," but these are the most common ones I can think of.

Thursday, December 03, 2009

Diane's Final Hand at the Bike?

There have been some complaints that I have been focusing too much on my poker analyses recently and have neglected telling stories about my experiences at the poker tables. So, here's something that happened last night.

In the 5-10 blind $500+ buyin NL game, I was on the button and somebody straddled for $20 (he gets last action, like another big blind). One player limped, and Diane, to my right, raised to $85. Diane is an attractive Asian woman who prides herself on her ability to loosen up a table, encouraging people to gamble with weak hands. Just a few hands earlier, she had executed a successful bluff with a $250 flop bet followed by a $800 turn bet with ten-high, no draw. She had about $5000 in play.

I had about $900 and looked down at AKo, much better than what Diane was likely playing with. I raised to $275.

To my left (in the small blind) was one of the tightest NL players I've played with. He had about $2700 and called my $275 raise. Everyone folded to Diane, who raised to $1000, putting me all-in. I have to call about $625 more. I need about 30% equity to make the call correct. I call. Even against KK I have enough equity here. I'm hoping the tight player to my left will fold, or that he'll push all-in and make Diane fold (in which case there will be more money for me to win, and I'll only need to win 23% of the time).

While the player to my left was deciding what to do, Diane asked me if she could look at my cards. I smiled at her incredulously and shook my head no. She hadn't bothered to wait for my reply, though, and looked at my cards.

The player who had been in the big blind was quite upset by this, and the floorman was called over. It's rather unusual for a player who isn't directly involved in a situation to call the floorman, and some of the other players (most notably Diane and "Corporation" Mike) were quite upset with this guy. However, as the guy said, "I'm not going to just sit here if I see something fishy going on at the table! I'm supposed to just sit back and watch cheating going on?" I'm sure Diane didn't mean to cheat, but I still think it was right for this guy to call the floorman.

Anyway, I think you can see where this is going. The floorman misunderstood the first explanation and said "well, I can't kill her hand just because someone showed his hand to her!" Of course, she had taken it upon herself to look at my hand. Diane started shouting about how this wasn't fair and if they decided to kill her hand, she would never come back to the Bike. The floorman called over the floor supervisor, who, amidst a growing crowd of onlookers, took about two minutes to finally call Diane's hand dead. Her $1000 stayed in the pot. On the verge of tears, Diane left, loudly cursing the floormen and explaining that this is why nobody comes to the Bike anymore.

In my opinion, this was a relatively clear decision for the floormen. Although Diane was not trying to cheat, she did break the rules and give herself an unfair advantage. Technically, her hand would be dead even if she were also all-in already (even though looking at my cards would give her no advantage because she would have had no more decisions to make). However, it's unfortunate to lose a friendly player who was generally very good for the game.

With Diane's hand out of the way and her money in the pot, the tight player made an easy call with QQ and won the whole pot.