From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 42: If you check the river, most players will bet only with very good hands and with bluffs. They'll check down hands that could win a showdown, but that are unlikely to be called by worse hands.
Yup. This is the correct way for your opponents to play the river, and since it's rather basic and intuitive, most players have mastered this strategy. You will see players deviate from this, but it is quite rare. However, I think it is worth noting because you don't want to fall into a pattern where you pay off your opponents if they make huge bets with hands that are only moderately strong.
As Sklansky and Malmuth point out in their discussion, this fact means that a lot of your medium-strength hands become "bluff-catchers." That is, they can only win if your opponent is bluffing. For example, it doesn't matter whether you hold top pair or bottom pair if your opponent will only value bet with two pair or better. Both hands can beat only bluffs. As S+M point out in Concept 44 (which I analyze below), the bigger your opponent's river bet, the less your hand matters; for example, if your opponent is rational and makes a bet of ten times the size of the pot, a wide range of your hands become "bluff-catchers," since your opponent will probably either have the nuts or nothing. On the other hand, if your opponent bets only one-tenth of the pot, the strength of your hand is very relevant.
Concept No. 43: Big bets mean big hands. Don't make or call big bets very often with weak hands.
As uncontroversial as this concept seems, I don't think it's precisely right. Big bets can mean not only big hands; they can also mean big draws or hands worried about draws. When your opponent makes a big bet (or when you make a big bet and get called), his hand range becomes much stronger. Thus, you should only call or make big bets with hands that have the potential to beat your opponent's strong range. This means you should be calling or making big bets only with big hands or big draws.
We can look back at some earlier concepts that touched on this topic to find some more exceptions and possibly gain some interesting insights.
Concept 35 seems to contradict this concept: "Unusually small bets tend to be made either with a big hand ... or with a bluff... With one pair your opponents will usually either check or bet a larger amount." So Concept 35 suggests: Big bets mean one pair.
Concept 40 also somewhat contradicts our current concept. This one suggests: Big bets mean the board has lots of likely draws.
Concept 1: "When in doubt, bet more" suggests: Big bets imply more doubt. Or something. This concept was pretty dumb.
To be fair, several concepts do support our current concept. For example, Concept 11 said "A big bet is the most relevant and accurate information available."
Concept 39 also supports our current concept. This was the one suggesting that you respect bigger bets more than smaller ones.
Concept No. 44: The bigger a bet your opponent makes, the more of your hands that turn into bluff catchers.
As I said above in my Concept 42 analysis, I agree with this one. At least in my games, my opponents are usually smart enough not to bet a huge amount with a hand that is only moderately strong, because they know I will probably only call them with hands that are even stronger.
It occurs to me that if my opponent is especially talented, he might be able to trick me into calling with a weaker hand trying to catch his bluff, but this is a very high-risk maneuver for him if his hand is only moderately strong.
I like to think about these concepts in terms of hand ranges, when possible. When my opponent makes a big bet, his hand range is polarized: usually he'll have near the nuts, sometimes he'll have nothing. My range falls almost entirely between these two poles, so my hands are mostly bluff-catchers. When my opponent makes a small bet, it's usually with a hand that is only somewhat strong, or it could be a bluff. Many of the hands in my range will fall on either side of his value-betting range, so my hand strength is the main factor determining how I react to his bet.
Sunday, November 29, 2009
Wednesday, November 25, 2009
Analyzing NLHE:TAP Concept 41
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 41: When holding a mediocre hand, usually bet enough (but not more) so that a raise means you are almost certainly beaten.
I don't like this one. It falls into the category of advice that stems from the idea that you should avoid putting yourself in a position of having to make a difficult decision. As I've already discussed twice before in this series of concept analyses, I don't buy into this idea. I don't think that you can generally improve your EV by maneuvering like this. However, I have heard that Chris Ferguson espouses this idea in The Full Tilt Poker Strategy Guide, which makes me worry that I might be wrong. A project I'd like to take on when I'm done with these concept analyses is to try to disprove this idea in general (or prove it, as the case may be). I'll have to take a look at what "Jesus" has to say, since he really knows what he's talking about when it comes to game theory.
Concept No. 41: When holding a mediocre hand, usually bet enough (but not more) so that a raise means you are almost certainly beaten.
I don't like this one. It falls into the category of advice that stems from the idea that you should avoid putting yourself in a position of having to make a difficult decision. As I've already discussed twice before in this series of concept analyses, I don't buy into this idea. I don't think that you can generally improve your EV by maneuvering like this. However, I have heard that Chris Ferguson espouses this idea in The Full Tilt Poker Strategy Guide, which makes me worry that I might be wrong. A project I'd like to take on when I'm done with these concept analyses is to try to disprove this idea in general (or prove it, as the case may be). I'll have to take a look at what "Jesus" has to say, since he really knows what he's talking about when it comes to game theory.
Sunday, November 22, 2009
Analyzing NLHE:TAP Concept 40
After this post, I'll be two-thirds of the way through my analysis of the sixty concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 40: Certain flops require certain-size bets. No matter what hand you hold, your flop bets, on average, should be smaller on flops like AhKdKs than they are on flops like Jh9s7h.
I think the authors have overstated the idea here, but I still like this concept. Most players underestimate the importance of flop texture in sizing their bets, but the authors go too far when they imply that flop texture is the only factor worth considering. Also important are position, stack-sizes, number of opponents, your hand, your table image, and your opponents' tendencies. On the AKK flop, for example, the bet should usually be around 1/4 to 1/2 of the pot. Bets outside this range might sometimes be better if stack sizes are quite small or if your opponents are maniacs, but these are unusual circumstances. On the J97 flop, you should almost never bet less than 3/4 of the pot, and sometimes it will be correct to bet 3/2 of the pot.
The reasons for the different bet sizes on different flops are basically those described in the book's discussion. If you are ahead on a AKK flop, your opponents have little chance to outdraw you, and even if they do, they usually won't be able to make much more from you. They have very little implied odds and very little incentive to call a bet of over 1/4 of the pot unless they have you beat. Similarly, your bluffs should be small on this flop to cover your small value bets. On the J97 flop, however, there are all sorts of draws, meaning that your opponent could easily draw to beat your hand, and if he does, he might be able to win quite a bit more from you. Your opponents are likely to have good implied odds in this situation, and so you must bet more to discourage them from calling. Your bluffs must follow suit.
Another way to look at this is that if you have a made hand on the J97 flop, you can make large value bets without worrying about making your opponents fold, but on the AKK flop, the most you can hope to win is a small bet or two.
Concept No. 40: Certain flops require certain-size bets. No matter what hand you hold, your flop bets, on average, should be smaller on flops like AhKdKs than they are on flops like Jh9s7h.
I think the authors have overstated the idea here, but I still like this concept. Most players underestimate the importance of flop texture in sizing their bets, but the authors go too far when they imply that flop texture is the only factor worth considering. Also important are position, stack-sizes, number of opponents, your hand, your table image, and your opponents' tendencies. On the AKK flop, for example, the bet should usually be around 1/4 to 1/2 of the pot. Bets outside this range might sometimes be better if stack sizes are quite small or if your opponents are maniacs, but these are unusual circumstances. On the J97 flop, you should almost never bet less than 3/4 of the pot, and sometimes it will be correct to bet 3/2 of the pot.
The reasons for the different bet sizes on different flops are basically those described in the book's discussion. If you are ahead on a AKK flop, your opponents have little chance to outdraw you, and even if they do, they usually won't be able to make much more from you. They have very little implied odds and very little incentive to call a bet of over 1/4 of the pot unless they have you beat. Similarly, your bluffs should be small on this flop to cover your small value bets. On the J97 flop, however, there are all sorts of draws, meaning that your opponent could easily draw to beat your hand, and if he does, he might be able to win quite a bit more from you. Your opponents are likely to have good implied odds in this situation, and so you must bet more to discourage them from calling. Your bluffs must follow suit.
Another way to look at this is that if you have a made hand on the J97 flop, you can make large value bets without worrying about making your opponents fold, but on the AKK flop, the most you can hope to win is a small bet or two.
Analyzing NLHE:TAP Concept 39
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 39: You must adapt your play to different-sized bets. If you will call a twice-pot bet as often as you call a half-pot bet, you're in trouble.
This is correct both in terms of game theory and in practice. There are probably some players against whom you should not take this advice, but they are rare.
Game theory assumes that your opponents play optimally against your strategy. If you call just as often whether the bet is big or small, your opponents can easily exploit this by making their bluff bets small and their value bets big.
In both theory and practice, I think it's true that big bets are more likely to be bluffs than small bets, and this is why some people are inclined to call big bets liberally. The big bets look suspicious. However, this does not mean you can call these big bets more often, because you need to win a higher percentage of the time for you to come out ahead. The fact is that you have to let yourself be bluffed once in a while, especially when your opponent makes a big bet.
As I said, there are probably some players against whom you should call big bets even more liberally than small bets. This situation can arise if you notice that your opponent always makes small value bets but his big bets tend to be bluffs. You still need to be wary, though, because players are liable to change their strategy at any time!
Concept No. 39: You must adapt your play to different-sized bets. If you will call a twice-pot bet as often as you call a half-pot bet, you're in trouble.
This is correct both in terms of game theory and in practice. There are probably some players against whom you should not take this advice, but they are rare.
Game theory assumes that your opponents play optimally against your strategy. If you call just as often whether the bet is big or small, your opponents can easily exploit this by making their bluff bets small and their value bets big.
In both theory and practice, I think it's true that big bets are more likely to be bluffs than small bets, and this is why some people are inclined to call big bets liberally. The big bets look suspicious. However, this does not mean you can call these big bets more often, because you need to win a higher percentage of the time for you to come out ahead. The fact is that you have to let yourself be bluffed once in a while, especially when your opponent makes a big bet.
As I said, there are probably some players against whom you should call big bets even more liberally than small bets. This situation can arise if you notice that your opponent always makes small value bets but his big bets tend to be bluffs. You still need to be wary, though, because players are liable to change their strategy at any time!
Saturday, November 21, 2009
Analyzing NLHE:TAP Concept 38
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 38: Be more apt to semi-bluff when your draw isn't to the nuts than when it is.
Yes. This was discussed indirectly in my analysis of Concept 34. The key point there was "the higher implied odds your draw has, the less attractive semibluffing with it becomes."A draw to the nuts has much better implied odds when it hits than does a non-nut draw, which makes checking or calling with it a more attractive option relative to semibluffing.
Put another way, semibluffing will often win the hand immediately when your opponent folds. This is the ideal result regardless of whether your draw is to the nuts, but it's more beneficial if your draw has meager implied odds. Nut draws usually have strong implied odds, so they are commonly worth just calling with.
As an aside, when semibluffing with a non-nut draw, I try to bet enough to make better draws consider folding. For example, if I have 8h7h and the flop is Ah6h5c, I will make sure to offer my opponent significantly worse than 2-to-1 odds. This way, someone with a better flush draw will have to consider folding, because if I have a pair of aces or better, calling would be incorrect for him.
Concept No. 38: Be more apt to semi-bluff when your draw isn't to the nuts than when it is.
Yes. This was discussed indirectly in my analysis of Concept 34. The key point there was "the higher implied odds your draw has, the less attractive semibluffing with it becomes."A draw to the nuts has much better implied odds when it hits than does a non-nut draw, which makes checking or calling with it a more attractive option relative to semibluffing.
Put another way, semibluffing will often win the hand immediately when your opponent folds. This is the ideal result regardless of whether your draw is to the nuts, but it's more beneficial if your draw has meager implied odds. Nut draws usually have strong implied odds, so they are commonly worth just calling with.
As an aside, when semibluffing with a non-nut draw, I try to bet enough to make better draws consider folding. For example, if I have 8h7h and the flop is Ah6h5c, I will make sure to offer my opponent significantly worse than 2-to-1 odds. This way, someone with a better flush draw will have to consider folding, because if I have a pair of aces or better, calling would be incorrect for him.
Wednesday, November 18, 2009
Analyzing NLHE:TAP Concept 37
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 37: Bets on the turn should, on average, constitute a smaller percentage of the pot than flop bets.
This is probably true for the reason given by the authors in their discussion, but they neglect an opposing factor that could conceivably refute this concept's claim.
There are some obvious differences between the flop and the turn. After the turn betting round, there is only one more card to come and one more betting round. After the flop, there are two more cards and two more betting rounds. As Sklansky and Miller discuss, both of these factors favor the EV of draws on the flop over the turn. It follows that, in order to make it unprofitable for a drawing hand to call, a made hand would need to bet more on the flop than on the turn.
However, there is another obvious difference between the flop and the turn that can have an effect of the EV of draws, and this one favors the EV of the draw on the turn. I am referring to the fact that while there are only three cards on the board after the flop, there are four after the turn. This extra card means that the board has much more potential to threaten multiple draws, which means that it will be less obvious which draw your opponent has. This, in turn, means that each draw has higher implied odds, because (as Sklansky and Miller pointed out in Concept 31) if your opponent hits his draw, it will be very difficult for you to figure it out. To combat this, you will often have to bet extra on the turn if you have a made hand. This factor is completely ignored by S+M.
So, on the flop, draws have the benefit of an extra round of betting to extract value if they hit on the turn, plus the a possibility of getting two tries to hit the card (although they may have to call another bet on the turn in order to see the river). These factors increase the EV of draws on the flop, and thus demand bigger bets from made hands.
On the turn, draws have the benefit of some extra "cover" because there will often be more draws on the board than on the flop. These factors increase the EV of draws on the turn, and thus demand bigger bets from made hands.
Although I suspect the former factor is more significant, which would make this concept's claim correct, it's not entirely clear. What is clear is that the latter factor was ignored by the authors in the book's discussion.
Concept No. 37: Bets on the turn should, on average, constitute a smaller percentage of the pot than flop bets.
This is probably true for the reason given by the authors in their discussion, but they neglect an opposing factor that could conceivably refute this concept's claim.
There are some obvious differences between the flop and the turn. After the turn betting round, there is only one more card to come and one more betting round. After the flop, there are two more cards and two more betting rounds. As Sklansky and Miller discuss, both of these factors favor the EV of draws on the flop over the turn. It follows that, in order to make it unprofitable for a drawing hand to call, a made hand would need to bet more on the flop than on the turn.
However, there is another obvious difference between the flop and the turn that can have an effect of the EV of draws, and this one favors the EV of the draw on the turn. I am referring to the fact that while there are only three cards on the board after the flop, there are four after the turn. This extra card means that the board has much more potential to threaten multiple draws, which means that it will be less obvious which draw your opponent has. This, in turn, means that each draw has higher implied odds, because (as Sklansky and Miller pointed out in Concept 31) if your opponent hits his draw, it will be very difficult for you to figure it out. To combat this, you will often have to bet extra on the turn if you have a made hand. This factor is completely ignored by S+M.
So, on the flop, draws have the benefit of an extra round of betting to extract value if they hit on the turn, plus the a possibility of getting two tries to hit the card (although they may have to call another bet on the turn in order to see the river). These factors increase the EV of draws on the flop, and thus demand bigger bets from made hands.
On the turn, draws have the benefit of some extra "cover" because there will often be more draws on the board than on the flop. These factors increase the EV of draws on the turn, and thus demand bigger bets from made hands.
Although I suspect the former factor is more significant, which would make this concept's claim correct, it's not entirely clear. What is clear is that the latter factor was ignored by the authors in the book's discussion.
Sunday, November 15, 2009
Analyzing NLHE:TAP Concept 36
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 36: Be more apt to slowplay very good hands that aren't quite the nuts than the nuts itself.
This concept seemed wrong to me from the first time I read it, but the authors bring up an interesting point in the discussion. After considering it, I still think the advice is generally wrong.
The point that Sklansky and Miller make is that one of the benefits of slowplaying a hand that is not quite the nuts is that you will potentially save a lot of money if you happen to be up against the nut hand. The example they use is a J66 flop. Here they say you'd rather slowplay with K6 than with JJ, because with K6 you might save money if you are up against JJ, J6, or A6.
Well, this is true, but it's a minor concern because your opponent will so rarely hold one of these hands. In the extremely likely event that your opponent does not hold one of these hands, you should be more apt to slowplay with the nuts (well, JJ is not the nuts, but we can imagine we hold 66). It's possible that there may be some situations where you would save so much money by slowplaying with K6 when you are behind that it actually is correct to slowplay it in a situation where slowplaying with JJ is not correct. However, I think this would be extremely unlikely in practice.
Sklansky and Miller neglect to mention the bad things that can happen when you slowplay. By slowplaying:
1. You don't get as much money in the pot when your opponent would call you.
2. Your opponent might outdraw you with a hand he would have folded.
3. Your opponent might have called on the flop but be scared off by a turn or river card.
4. A turn or river card might scare you enough that you have to stop betting or raising.
Factors 2 and 4 are a much bigger concern if you hold the near-nuts than if you hold the nuts. These problems are relatively common, at least when compared to the likelihood of finding that you're up against JJ when you hold K6 on a J66 flop.
Concept No. 36: Be more apt to slowplay very good hands that aren't quite the nuts than the nuts itself.
This concept seemed wrong to me from the first time I read it, but the authors bring up an interesting point in the discussion. After considering it, I still think the advice is generally wrong.
The point that Sklansky and Miller make is that one of the benefits of slowplaying a hand that is not quite the nuts is that you will potentially save a lot of money if you happen to be up against the nut hand. The example they use is a J66 flop. Here they say you'd rather slowplay with K6 than with JJ, because with K6 you might save money if you are up against JJ, J6, or A6.
Well, this is true, but it's a minor concern because your opponent will so rarely hold one of these hands. In the extremely likely event that your opponent does not hold one of these hands, you should be more apt to slowplay with the nuts (well, JJ is not the nuts, but we can imagine we hold 66). It's possible that there may be some situations where you would save so much money by slowplaying with K6 when you are behind that it actually is correct to slowplay it in a situation where slowplaying with JJ is not correct. However, I think this would be extremely unlikely in practice.
Sklansky and Miller neglect to mention the bad things that can happen when you slowplay. By slowplaying:
1. You don't get as much money in the pot when your opponent would call you.
2. Your opponent might outdraw you with a hand he would have folded.
3. Your opponent might have called on the flop but be scared off by a turn or river card.
4. A turn or river card might scare you enough that you have to stop betting or raising.
Factors 2 and 4 are a much bigger concern if you hold the near-nuts than if you hold the nuts. These problems are relatively common, at least when compared to the likelihood of finding that you're up against JJ when you hold K6 on a J66 flop.
Saturday, November 14, 2009
Analyzing NLHE:TAP Concept 35
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 35: Unusually small bets tend to be made either with a big hand (a suck-in bet) or with a bluff (a cheap stab at the pot). With one pair, your opponents will usually either check or bet a larger amount.
This is a strange concept because it contains no advice. I think the information is correct, but, to me, it seems only slightly useful.
In my experience, I think it's true that when someone makes a small bet it means it's slightly more likely that he has a big hand, and slightly less likely that he has one pair, but it's not a very significant difference. Depending on the game you are playing in, this concept could be off the mark for a stereotypical player; different types of plays are more or less popular in different venues, and this can also change over time. Also, once you have figured out what a particular opponent likes to do, the generalization made by this concept will be obsolete for that opponent; this concept applies only to players you do not know much about, and for them, bet-size is only a weak indication of hand-type.
Supposing that your opponent's bet-size really were a strong indicator of the type of hand he held, this information would be quite useful, but not devastatingly so. On the one hand, if you knew your opponent either had a big hand or a bluff, you could decide whether to fold right away or just call; raising would never be a good play in this situation. If, on the other hand, you knew your opponent held a pair, it could be a good play for you to raise either for value or as a bluff, but it would still be unclear how your opponent would react to this raise or if he would continue to bet his pair on future streets.
With this concept, Sklansky and Miller are pointing out a very marginal interpretation of how to read an opponent's action, and it's dependent (as they admit in their discussion) on who your opponent is. They offer no insight into how to use this information, although I can accept that they probably consider it outside the scope of this concept (or maybe they just think the answer is obvious).
Concept No. 35: Unusually small bets tend to be made either with a big hand (a suck-in bet) or with a bluff (a cheap stab at the pot). With one pair, your opponents will usually either check or bet a larger amount.
This is a strange concept because it contains no advice. I think the information is correct, but, to me, it seems only slightly useful.
In my experience, I think it's true that when someone makes a small bet it means it's slightly more likely that he has a big hand, and slightly less likely that he has one pair, but it's not a very significant difference. Depending on the game you are playing in, this concept could be off the mark for a stereotypical player; different types of plays are more or less popular in different venues, and this can also change over time. Also, once you have figured out what a particular opponent likes to do, the generalization made by this concept will be obsolete for that opponent; this concept applies only to players you do not know much about, and for them, bet-size is only a weak indication of hand-type.
Supposing that your opponent's bet-size really were a strong indicator of the type of hand he held, this information would be quite useful, but not devastatingly so. On the one hand, if you knew your opponent either had a big hand or a bluff, you could decide whether to fold right away or just call; raising would never be a good play in this situation. If, on the other hand, you knew your opponent held a pair, it could be a good play for you to raise either for value or as a bluff, but it would still be unclear how your opponent would react to this raise or if he would continue to bet his pair on future streets.
With this concept, Sklansky and Miller are pointing out a very marginal interpretation of how to read an opponent's action, and it's dependent (as they admit in their discussion) on who your opponent is. They offer no insight into how to use this information, although I can accept that they probably consider it outside the scope of this concept (or maybe they just think the answer is obvious).
Monday, November 09, 2009
Analyzing NLHE:TAP Concept 34
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 34: If you have a close decision between semibluffing with a draw and checking it, be more inclined to check if you could make your draw with an overcard to the board.
Sklansky and Miller explain that if you make your draw with an overcard to the board, you might be able to win a big pot if your opponent happens to hit this card, too. I agree, but I don't like how this concept is presented. The discussion begins with the observation that "the higher implied odds your draw has, the less attractive semibluffing with it becomes." This is an important idea, and it would have made for a better "concept" than the one the authors chose. By semibluffing, you often eliminate your implied odds by ending the betting right away.
As it stands, Concept 34 describes a specific case of this idea. The authors not only missed this opportunity to put a particularly useful idea in their concepts, but they also failed to emphasize that the reader needs to be careful not to overgeneralize the advice given, which I think only holds in the specific case where you are deciding between semibluffing and checking. If you are deciding between semibluffing and betting, the opposite advice seems to hold: be more inclined to semibluff if you could make your draw with an overcard to the board.
The reason for this subtle distinction is that your implied odds only improve if an overcard is likely to help your opponent. This is generally only the case if your opponent has two overcards, which is unikely if you are thinking about calling (meaning your opponent has bet), because your opponent probably already has a pair. In this case, you would rather your draw be to undercards, because your opponent might be scared away by an overcard, thereby reducing your implied odds. In the case described by S+M, you are considering checking with your draw, meaning your opponent has not bet. So, it's a lot more likely that he his holding two overcards, and indeed it might help your implied odds if your draw included overcards to the board.
This is a little confusing, but I think the important insight is rather clear: "the higher implied odds your draw has, the less attractive semibluffing with it becomes."
Concept No. 34: If you have a close decision between semibluffing with a draw and checking it, be more inclined to check if you could make your draw with an overcard to the board.
Sklansky and Miller explain that if you make your draw with an overcard to the board, you might be able to win a big pot if your opponent happens to hit this card, too. I agree, but I don't like how this concept is presented. The discussion begins with the observation that "the higher implied odds your draw has, the less attractive semibluffing with it becomes." This is an important idea, and it would have made for a better "concept" than the one the authors chose. By semibluffing, you often eliminate your implied odds by ending the betting right away.
As it stands, Concept 34 describes a specific case of this idea. The authors not only missed this opportunity to put a particularly useful idea in their concepts, but they also failed to emphasize that the reader needs to be careful not to overgeneralize the advice given, which I think only holds in the specific case where you are deciding between semibluffing and checking. If you are deciding between semibluffing and betting, the opposite advice seems to hold: be more inclined to semibluff if you could make your draw with an overcard to the board.
The reason for this subtle distinction is that your implied odds only improve if an overcard is likely to help your opponent. This is generally only the case if your opponent has two overcards, which is unikely if you are thinking about calling (meaning your opponent has bet), because your opponent probably already has a pair. In this case, you would rather your draw be to undercards, because your opponent might be scared away by an overcard, thereby reducing your implied odds. In the case described by S+M, you are considering checking with your draw, meaning your opponent has not bet. So, it's a lot more likely that he his holding two overcards, and indeed it might help your implied odds if your draw included overcards to the board.
This is a little confusing, but I think the important insight is rather clear: "the higher implied odds your draw has, the less attractive semibluffing with it becomes."
Sunday, November 08, 2009
Analyzing NLHE:TAP Concepts 32-33
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 32: It can be correct to fold a hand before the river that has a better than 50 percent chance of being the best hand.
This is certainly true, but there are two reasons that each would be sufficient even in the absence of the other. Sklansky and Miller ignore the first of these reasons in their discussion, and only touch on part of the second.
1. It matters how far behind you are when behind, and how far ahead when ahead. For example, TT is ahead 4/7 of the time against a range of AA,KK, or AK, but it only wins about 40% of the time. This is because it is way behind 3/7 of the time and only a little bit ahead 4/7 of the time.
2. With more cards and more betting still to come, other factors come into play when it comes to EV. Your pot equity can take a back seat to the power of having more information about your opponent's hand than he has about yours. The obvious example is that if you are out of position, you will have to act first in later rounds, and thus you will give your opponent information about your hand before each of his actions. Also, if you are a tight player, your opponent will have a better idea of what sort of hand you are likely to be holding. Furthermore, drawing hands have an information advantage on later streets; whether the draw is hit or missed, the player who was drawing can be fairly certain of whether he has the best hand or the worst hand. All of these factors are more important in NL holdem than in Limit, because players can make larger bets after accumulating the new information.
Sklansky and Malmuth are right about this concept, but their analysis neglects all these factors except that position becomes more important in NL, and that "vulnerable" hands (ie, moderately strong hands that can lose to many draws) lose some value in NL.
Concept No. 33: Be willing to risk free cards to manage the pot size and induce bluffs.
This is good advice. With certain types of hands (in particular, "vulnerable" hands), you would prefer to play a small pot. I've never had much use for the concept of "pot control" or "managing pot size," but it's not really such a bad way to conceptualize why it's best not to bet with certain moderately strong hands. I'm not in the habit of using these ideas; I always just think about it in terms of managing my range and trying to maximize my EV. If I bet with a medium-strength hand on a drawish flop, I will tend to be called mostly by better hands than mine or draws; I will lose a lot against the stronger hands, but gain only a little against the draws. Making matters worse, players holding draws will sometimes raise as a bluff, forcing me to fold. If I do decide to bet the flop despite all these dangers, I will probably check on the turn, regardless of whether the draw comes in. If I bet, I will have problems similar to those I had on the flop, only worse. I will have shown strength by betting, but my hand will be among the weakest in my betting range, and I will have to consider folding if there is a substantial bet on the river. If instead I check and show weakness, my hand will be among the strongest in my checking range, and I can pick off lots of bluffs on the next street.
There is way too much information to process exhaustively at a poker table. Even when I sit at home and analyze hands, I often have to generalize and guess at EV in certain situations, because it's just too complicated to approach it more thoroughly. So, it's necessary to conceptualize poker and use principles or heuristics to simplify decisions. The way I think about the game lends itself much better to the EV and hand ranges conceptualization, but there is value in using a higher-level conceptualization such as pot control and bluff-inducing, as Sklansky and Miller recommend. In future analyses, I might try occasionally to approach problems with both conceptualizations if I think there is a chance they will be at odds.
Concept No. 32: It can be correct to fold a hand before the river that has a better than 50 percent chance of being the best hand.
This is certainly true, but there are two reasons that each would be sufficient even in the absence of the other. Sklansky and Miller ignore the first of these reasons in their discussion, and only touch on part of the second.
1. It matters how far behind you are when behind, and how far ahead when ahead. For example, TT is ahead 4/7 of the time against a range of AA,KK, or AK, but it only wins about 40% of the time. This is because it is way behind 3/7 of the time and only a little bit ahead 4/7 of the time.
2. With more cards and more betting still to come, other factors come into play when it comes to EV. Your pot equity can take a back seat to the power of having more information about your opponent's hand than he has about yours. The obvious example is that if you are out of position, you will have to act first in later rounds, and thus you will give your opponent information about your hand before each of his actions. Also, if you are a tight player, your opponent will have a better idea of what sort of hand you are likely to be holding. Furthermore, drawing hands have an information advantage on later streets; whether the draw is hit or missed, the player who was drawing can be fairly certain of whether he has the best hand or the worst hand. All of these factors are more important in NL holdem than in Limit, because players can make larger bets after accumulating the new information.
Sklansky and Malmuth are right about this concept, but their analysis neglects all these factors except that position becomes more important in NL, and that "vulnerable" hands (ie, moderately strong hands that can lose to many draws) lose some value in NL.
Concept No. 33: Be willing to risk free cards to manage the pot size and induce bluffs.
This is good advice. With certain types of hands (in particular, "vulnerable" hands), you would prefer to play a small pot. I've never had much use for the concept of "pot control" or "managing pot size," but it's not really such a bad way to conceptualize why it's best not to bet with certain moderately strong hands. I'm not in the habit of using these ideas; I always just think about it in terms of managing my range and trying to maximize my EV. If I bet with a medium-strength hand on a drawish flop, I will tend to be called mostly by better hands than mine or draws; I will lose a lot against the stronger hands, but gain only a little against the draws. Making matters worse, players holding draws will sometimes raise as a bluff, forcing me to fold. If I do decide to bet the flop despite all these dangers, I will probably check on the turn, regardless of whether the draw comes in. If I bet, I will have problems similar to those I had on the flop, only worse. I will have shown strength by betting, but my hand will be among the weakest in my betting range, and I will have to consider folding if there is a substantial bet on the river. If instead I check and show weakness, my hand will be among the strongest in my checking range, and I can pick off lots of bluffs on the next street.
There is way too much information to process exhaustively at a poker table. Even when I sit at home and analyze hands, I often have to generalize and guess at EV in certain situations, because it's just too complicated to approach it more thoroughly. So, it's necessary to conceptualize poker and use principles or heuristics to simplify decisions. The way I think about the game lends itself much better to the EV and hand ranges conceptualization, but there is value in using a higher-level conceptualization such as pot control and bluff-inducing, as Sklansky and Miller recommend. In future analyses, I might try occasionally to approach problems with both conceptualizations if I think there is a chance they will be at odds.
Saturday, November 07, 2009
Poker Riddle
Yesterday, a guy named Jesse, his friend (can't remember his name), and the actor Michael Muhney were in the 500 NL game discussing probabilities and IQs and other nerdy things. I learned that Michael is in MENSA (actually I knew this from his IMDb page), Jesse has a 168 IQ, and his friend went to a school that required an IQ of at least 150. They had a discussion of how many times you'd need to double up to get $1 million if you started with $3. Jesse concluded that the probability was 1-(1/2)^18. He was emphatic on this point for a while even after it was suggested to him that the real answer was more likely (1/2)^18.
Anyway, Jesse had a riddle that he was sure nobody could figure out: you are heads-up in a holdem game, and you're ahead preflop, on the flop, and on the turn, but you have no chance to win on the river (only fold or chop). Jesse offered Michael 3-1 odds on a bet that he couldn't figure it out in 15 minutes, but Michael refused because, as Jesse put it, he is a "life nit," meaning that he is unwilling to gamble on things outside of poker. (A "nit" in poker is someone who is unwilling to gamble much, mostly only playing the nuts.) Michael confirmed that yes, he is a "life nit," if, by that, Jesse meant "a responsible person with two kids who doesn't want to have to explain to his wife in 15 years that he can't pay for their kids to go to college because he gambled all his money away to some guy at the casino with a 168 IQ who can't even do simple algebra."
I asked Jesse's friend, who had heard the riddle and was sitting next to me, how they define "being ahead on the turn." My definition would be that "being ahead" means you have the best chance to win the pot. Clearly, they had a different definition, since the problem is set up such that the hand that is "ahead" supposedly has no chance to win the pot. Jesse's friend said that a hand was defined as being "ahead" if it would win without any more cards coming. Jesse offered me the deal as well ($25 to $75), but, being a life nit myself, I turned him down.
I thought about the problem for a minute but did not figure it out. I eliminated the possibility that the answer involved flushes, and decided that it probably involved low cards. With low cards, it's much easier for kickers to get counterfeited on the river. Anyway, after playing for another half hour, I asked Jesse's friend what the answer was.
"You have to figure it out for yourself, man. Jesse! He wants us to just tell him the answer!"
"Okay, let me think about it," I said.
I thought about 32 against 42, but it didn't quite work. I thought about 43 against 42, with a flop of QQ4. Then a 2 on the turn. So far, so good. 43 is ahead preflop, on the flop, and on the turn, but I can't think of any river card that would result in a win for 43. If the river is a 2, then 42 wins with a full house. Any X higher than 2, and both players have QQ44 with a X kicker. Any Q or 4 and both players have the same full house. After about 1 minute, I told Jesse's friend I thought I figured it out.
"Did you think through all the possibilities? Just think about it." Jesse told me he still wanted to bet me 3-1, and he would give me twelve minutes and let me discuss it with anyone in the casino except his friend, who knew the answer. I hesitated, partly due to his confidence (although he had been similarly confident about the 1-(1/2)^18 formulation), but mostly just because I do not like making proposition bets, especially at poker tables. After a minute, I took the bet anyway. I asked Tony, another prop at the 500NL table, to confirm my answer. We discussed it for about five minutes, and he thought it looked good. Meanwhile, Jesse, having heard my discussion with Tony, proclaimed that I was "a million miles off" and wanted to double the stakes and give me a hint and let me call someone on the phone. I didn't want to escalate the situation any further, so I declined. I gave him my answer. After studying it incredulously for ten or fifteen minutes (and briefly trying to argue that 43 was not "ahead" on the turn), he conceded that it looked right and gave me my $75. I gave $15 to Tony for helping. Jesse said he had seen this question in a magazine and thought there was only one answer, which he told me. He said he had been asking that question at poker tables for five years and nobody had figured it out. I suspect he left something important out of the question, but I can't think what it would be. Anyway, Tony now thinks I'm a genius.
My answer can actually be generalized quite a bit. Instead of 43 and 42, I think any X3 and X2 will work, except X=2 or 3. Similarly, in addition to QQ on the flop, any YYX flop will work, unless it gives X3 a backdoor straight draw. So, X=Ace and Y=4 does not work, but X=7 and Y=4 does. The key is that a 2 has to come on the turn in all cases.
Jesse's answer does not fall into this category. Can you think of it? I'll give clues in the comments if anyone asks. I haven't thought of any solutions besides Jesse's answer and mine.
Anyway, Jesse had a riddle that he was sure nobody could figure out: you are heads-up in a holdem game, and you're ahead preflop, on the flop, and on the turn, but you have no chance to win on the river (only fold or chop). Jesse offered Michael 3-1 odds on a bet that he couldn't figure it out in 15 minutes, but Michael refused because, as Jesse put it, he is a "life nit," meaning that he is unwilling to gamble on things outside of poker. (A "nit" in poker is someone who is unwilling to gamble much, mostly only playing the nuts.) Michael confirmed that yes, he is a "life nit," if, by that, Jesse meant "a responsible person with two kids who doesn't want to have to explain to his wife in 15 years that he can't pay for their kids to go to college because he gambled all his money away to some guy at the casino with a 168 IQ who can't even do simple algebra."
I asked Jesse's friend, who had heard the riddle and was sitting next to me, how they define "being ahead on the turn." My definition would be that "being ahead" means you have the best chance to win the pot. Clearly, they had a different definition, since the problem is set up such that the hand that is "ahead" supposedly has no chance to win the pot. Jesse's friend said that a hand was defined as being "ahead" if it would win without any more cards coming. Jesse offered me the deal as well ($25 to $75), but, being a life nit myself, I turned him down.
I thought about the problem for a minute but did not figure it out. I eliminated the possibility that the answer involved flushes, and decided that it probably involved low cards. With low cards, it's much easier for kickers to get counterfeited on the river. Anyway, after playing for another half hour, I asked Jesse's friend what the answer was.
"You have to figure it out for yourself, man. Jesse! He wants us to just tell him the answer!"
"Okay, let me think about it," I said.
I thought about 32 against 42, but it didn't quite work. I thought about 43 against 42, with a flop of QQ4. Then a 2 on the turn. So far, so good. 43 is ahead preflop, on the flop, and on the turn, but I can't think of any river card that would result in a win for 43. If the river is a 2, then 42 wins with a full house. Any X higher than 2, and both players have QQ44 with a X kicker. Any Q or 4 and both players have the same full house. After about 1 minute, I told Jesse's friend I thought I figured it out.
"Did you think through all the possibilities? Just think about it." Jesse told me he still wanted to bet me 3-1, and he would give me twelve minutes and let me discuss it with anyone in the casino except his friend, who knew the answer. I hesitated, partly due to his confidence (although he had been similarly confident about the 1-(1/2)^18 formulation), but mostly just because I do not like making proposition bets, especially at poker tables. After a minute, I took the bet anyway. I asked Tony, another prop at the 500NL table, to confirm my answer. We discussed it for about five minutes, and he thought it looked good. Meanwhile, Jesse, having heard my discussion with Tony, proclaimed that I was "a million miles off" and wanted to double the stakes and give me a hint and let me call someone on the phone. I didn't want to escalate the situation any further, so I declined. I gave him my answer. After studying it incredulously for ten or fifteen minutes (and briefly trying to argue that 43 was not "ahead" on the turn), he conceded that it looked right and gave me my $75. I gave $15 to Tony for helping. Jesse said he had seen this question in a magazine and thought there was only one answer, which he told me. He said he had been asking that question at poker tables for five years and nobody had figured it out. I suspect he left something important out of the question, but I can't think what it would be. Anyway, Tony now thinks I'm a genius.
My answer can actually be generalized quite a bit. Instead of 43 and 42, I think any X3 and X2 will work, except X=2 or 3. Similarly, in addition to QQ on the flop, any YYX flop will work, unless it gives X3 a backdoor straight draw. So, X=Ace and Y=4 does not work, but X=7 and Y=4 does. The key is that a 2 has to come on the turn in all cases.
Jesse's answer does not fall into this category. Can you think of it? I'll give clues in the comments if anyone asks. I haven't thought of any solutions besides Jesse's answer and mine.
Friday, November 06, 2009
Analyzing NLHE:TAP Concepts 30-31
After this post, I'll be more than half-way through my analysis of the sixty concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 30: Implied odds are a critically important decision-making tool, but always be aware that different opponents offer different odds.
I like this one. Players often overestimate their implied odds by assuming they will win their opponent's entire stack if they make their hand. In reality, of course, sometimes the other player will fold. Other times, you will make your hand only to see it lose to a better one. The probability of either of these two things happening can be approximated without any knowledge of your opponent. However, as Sklansky and Miller point out, if you do have any prior knowledge about your opponent, you'd better take it into account. It can make the difference when deciding whether to call with a drawing hand.
I also like the authors' observation (in their discussion) that the players who offer the least implied odds (because they will fold if you hit your draw) will also be the easiest to bluff. This slightly increases the EV of calling a player who offers low implied odds, but it also influences how you should play your draws and the types of draws you should be calling with. First of all, against such players you should be much more inclined to raise as a semi-bluff. Not only are they more likely to fold, but you're also giving up less in the way of implied odds had you just called. Second of all, you'll want to look for opportunities disguise one draw as another. For example, if you're heads-up against such a player on a board of Ad Th 8d, it's probably better to have the 97 straight draw than the diamond flush draw. You have fewer "outs," but with the straight draw, you not only have decent implied odds for the times you hit your straight, but you'll also probably win the pot by bluffing if another diamond comes. On the other hand, if you have the flush draw, your only chance to win is probably to make your flush, and if you do, your draw is so obvious that you have little chance of getting paid off. This leads directly to the next concept, so let's move right along...
Concept No. 31: Your implied odds with any draw will be better the less obvious the draw is.
Since you just read my analysis of Concept 30, you can guess that I agree with this one. If your opponents are worried about a particular draw and it comes in, they are not likely to pay you off if you hit it. Against some players, you'll probably want to make only a tiny bet if you hit such a draw, because that is your only hope to get paid. In other words, if your draw is obvious, you have very little implied odds. S+M use the example of the nut flush draw on a flop of three diamonds. Everyone is worried about the flush. On the other hand, if you decided to play 53 and the flop comes K42 rainbow, you will have very good implied odds for drawing to your straight.
Concept No. 30: Implied odds are a critically important decision-making tool, but always be aware that different opponents offer different odds.
I like this one. Players often overestimate their implied odds by assuming they will win their opponent's entire stack if they make their hand. In reality, of course, sometimes the other player will fold. Other times, you will make your hand only to see it lose to a better one. The probability of either of these two things happening can be approximated without any knowledge of your opponent. However, as Sklansky and Miller point out, if you do have any prior knowledge about your opponent, you'd better take it into account. It can make the difference when deciding whether to call with a drawing hand.
I also like the authors' observation (in their discussion) that the players who offer the least implied odds (because they will fold if you hit your draw) will also be the easiest to bluff. This slightly increases the EV of calling a player who offers low implied odds, but it also influences how you should play your draws and the types of draws you should be calling with. First of all, against such players you should be much more inclined to raise as a semi-bluff. Not only are they more likely to fold, but you're also giving up less in the way of implied odds had you just called. Second of all, you'll want to look for opportunities disguise one draw as another. For example, if you're heads-up against such a player on a board of Ad Th 8d, it's probably better to have the 97 straight draw than the diamond flush draw. You have fewer "outs," but with the straight draw, you not only have decent implied odds for the times you hit your straight, but you'll also probably win the pot by bluffing if another diamond comes. On the other hand, if you have the flush draw, your only chance to win is probably to make your flush, and if you do, your draw is so obvious that you have little chance of getting paid off. This leads directly to the next concept, so let's move right along...
Concept No. 31: Your implied odds with any draw will be better the less obvious the draw is.
Since you just read my analysis of Concept 30, you can guess that I agree with this one. If your opponents are worried about a particular draw and it comes in, they are not likely to pay you off if you hit it. Against some players, you'll probably want to make only a tiny bet if you hit such a draw, because that is your only hope to get paid. In other words, if your draw is obvious, you have very little implied odds. S+M use the example of the nut flush draw on a flop of three diamonds. Everyone is worried about the flush. On the other hand, if you decided to play 53 and the flop comes K42 rainbow, you will have very good implied odds for drawing to your straight.
Thursday, November 05, 2009
Analyzing NLHE:TAP Concept 29
I'm working my way through the sixty concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller. Almost halfway done!
Concept No. 29: It's okay to make small raises (2-3x the big blind) to build the pot or to set up a future play.
My standard raises are already rather small, about 3-4x the big blind, so I have to agree that there's probably nothing wrong with raising only 2-3x the big blind instead. In fact, I will sometimes raise as little as 2x the blind, but this is almost always in order to deal with awkward stack-sizes. For example, if the effective stack sizes are around 8-12 times the big blind, I think 3-4x raises will commit me to the pot, but going all-in is too much to risk for just the blinds. In this scenario, I might make a tiny raise pre-flop. Another conceivable reason for a minimum-raise would be to manipulate your opponents to either reraise you or just call you (as S+M suggest in Concept 24), but unless you are extremely attuned to your opponents, it will be hard to convince them to react exactly how you intended. I don't think I personallyhave the talent to make this play work. In any case, Sklansky and Miller have an entirely different reason for making these small raises, and I am not impressed with it.
The authors say, "often you should make this sort of raise with 'brave' hands - pocket pairs, suited connectors, and suited aces - hands that play well after the flop." While I do like calling such hands "brave," I think it's because they are capable of going up against bigger hands and beating them for a big pot. You can be "brave" and call a raise with 98s because you have a chance of winning a big pot with a small investment. However, intentionally increasing the size of this initial investment defeats the whole purpose! By doubling the bet, you are essentially cutting your implied odds in half. If the effective stack size was 100 times the big blind, a 2x raise means you can now only win 50 times your investment if you get lucky and hit your straight. Usually all that will happen is you will fold on the flop and lose two blinds instead of one.
Note that I'm not suggesting that you should never raise with "brave" hands. In fact, I think raising with them is an important part of a balanced strategy. However, that doesn't work if you raise an abnormally small amount with them. You need to raise your normal 3-4x amount in order to disguise your hand.
Concept No. 29: It's okay to make small raises (2-3x the big blind) to build the pot or to set up a future play.
My standard raises are already rather small, about 3-4x the big blind, so I have to agree that there's probably nothing wrong with raising only 2-3x the big blind instead. In fact, I will sometimes raise as little as 2x the blind, but this is almost always in order to deal with awkward stack-sizes. For example, if the effective stack sizes are around 8-12 times the big blind, I think 3-4x raises will commit me to the pot, but going all-in is too much to risk for just the blinds. In this scenario, I might make a tiny raise pre-flop. Another conceivable reason for a minimum-raise would be to manipulate your opponents to either reraise you or just call you (as S+M suggest in Concept 24), but unless you are extremely attuned to your opponents, it will be hard to convince them to react exactly how you intended. I don't think I personallyhave the talent to make this play work. In any case, Sklansky and Miller have an entirely different reason for making these small raises, and I am not impressed with it.
The authors say, "often you should make this sort of raise with 'brave' hands - pocket pairs, suited connectors, and suited aces - hands that play well after the flop." While I do like calling such hands "brave," I think it's because they are capable of going up against bigger hands and beating them for a big pot. You can be "brave" and call a raise with 98s because you have a chance of winning a big pot with a small investment. However, intentionally increasing the size of this initial investment defeats the whole purpose! By doubling the bet, you are essentially cutting your implied odds in half. If the effective stack size was 100 times the big blind, a 2x raise means you can now only win 50 times your investment if you get lucky and hit your straight. Usually all that will happen is you will fold on the flop and lose two blinds instead of one.
Note that I'm not suggesting that you should never raise with "brave" hands. In fact, I think raising with them is an important part of a balanced strategy. However, that doesn't work if you raise an abnormally small amount with them. You need to raise your normal 3-4x amount in order to disguise your hand.
Wednesday, November 04, 2009
Analyzing NLHE:TAP Concept 28
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 28: With strong hands, generally raise either a small, pot-building amount or a large, hand-defining amount. Don't raise an amount in the middle that both tells your opponents that you have a good hand and offers them the right implied odds to try to beat you.
I think it's better to just always bet a normal amount. I don't think you should you ever raise a "hand-defining amount." This will enable good players to play correctly against you, which means they will probably just fold unless they have you beat. This is clearly a terrible scenario for your strong hands. Bad players might call you with worse hands, but they even bad players are usually more likely to call a smaller bet. Then you can outplay them post-flop.
I'm also not a fan of "pot-building" bets, but these could conceivably work. Personally, I'd rather bet a little more and risk making a few players fold. Usually the result will be a pot of around the same size with fewer opponents, meaning you have a better chance to win.
Concept No. 28: With strong hands, generally raise either a small, pot-building amount or a large, hand-defining amount. Don't raise an amount in the middle that both tells your opponents that you have a good hand and offers them the right implied odds to try to beat you.
I think it's better to just always bet a normal amount. I don't think you should you ever raise a "hand-defining amount." This will enable good players to play correctly against you, which means they will probably just fold unless they have you beat. This is clearly a terrible scenario for your strong hands. Bad players might call you with worse hands, but they even bad players are usually more likely to call a smaller bet. Then you can outplay them post-flop.
I'm also not a fan of "pot-building" bets, but these could conceivably work. Personally, I'd rather bet a little more and risk making a few players fold. Usually the result will be a pot of around the same size with fewer opponents, meaning you have a better chance to win.
Monday, November 02, 2009
Analyzing NLHE:TAP Concepts 26-27
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 26: When there's an ante, your opening raises should be larger than if there were no ante. But they shouldn't be larger in the same proportion that the size of the initial pot increases; they should be somewhat smaller than that.
I don't know of any NL Holdem cash games that actually use an ante, but in theory I think Sklansky and Miller are right that this would be the correct way to adjust. With more money in the pot, it becomes more profitable to "steal"pots, and it becomes correct for you and your opponents to play looser because you have better pot odds. This involves not only raising more, but also widening your raising and calling ranges.
Concept No. 27: When semi-bluffing before the flop, usually do it those times you have one of the best hands that you'd otherwise fold. However, when you are in the blinds in an unraised pot, you should usually do it when you have one of your worst hands.
This one is complicated. I used to like this strategy, but now I think it's probably wrong. I now think the best hands to "semi-bluff" with are the hands that play the best against your opponent's likely calling hands. These include primarily suited aces, but also pocket pairs and suited connectors. Note that I don't think it is relevant whether you would "otherwise fold" with these hands, as Sklansky and Miller suggest. Many of these "semi-bluffing" hands are hands that you would likely have otherwise called with.
Sklansky and Miller's advice seems to be based on game-theoretically optimal river strategy, when there are no cards left to come. On the river, your bluff bets should indeed be with your worst possible hands, and your bluff raises should be with the best of the hands you would otherwise fold. This mirrors S+M's advice that you should bluff when "you have one of your worst hands" from the blinds, and that if you are not in the blinds you should bluff when "you have one of the best hands that you would otherwise fold."
The problem is that this optimal river bluffing strategy does not apply preflop. Preflop, the logic is much different because there are such things as drawing hands and semibluffs. Preflop, you do not need to bluff with terrible hands in order to induce your opponents to play hands weaker than your range; they will already be calling with some hands that they know are behind your range because they are "drawing hands." Instead, you should be bluffing with moderately strong hands as "semibluffs." The best hands to semibluff with preflop have two qualities. 1. They contain an ace, thereby reducing the chances you are up against AA. 2. They have a decent chance of making a big hand like a flush or a set. If your opponent is very tight, or if there have already been a couple of raises, point 1 is reasonably important. Otherwise, point 2 is much more important, because your opponent is unlikely to have AA anyway. In any case, this is why I suggested that the best hands to semibluff with preflop are suited aces, suited connectors, and pocket pairs.
Besides these types of hands, if you want to add some "bluffs" to your range, I think the best approach is to add some more hands that you would otherwise have called with. For example, try reraising with AJ or 77 on the button if you feel the need to bluff. Although these hands do have a lot of equity in just calling, they will also fare okay when your bluff is called. In the end, it's just my opinion, but I think this is a better approach than taking S+M's advice and raising with hands you would otherwise with fold, such as K7s. Although you are not losing any "calling equity" with K7s (since you would be folding otherwise), against normal opponents, these hands are just too unlikely to win if you are called preflop.
Concept No. 26: When there's an ante, your opening raises should be larger than if there were no ante. But they shouldn't be larger in the same proportion that the size of the initial pot increases; they should be somewhat smaller than that.
I don't know of any NL Holdem cash games that actually use an ante, but in theory I think Sklansky and Miller are right that this would be the correct way to adjust. With more money in the pot, it becomes more profitable to "steal"pots, and it becomes correct for you and your opponents to play looser because you have better pot odds. This involves not only raising more, but also widening your raising and calling ranges.
Concept No. 27: When semi-bluffing before the flop, usually do it those times you have one of the best hands that you'd otherwise fold. However, when you are in the blinds in an unraised pot, you should usually do it when you have one of your worst hands.
This one is complicated. I used to like this strategy, but now I think it's probably wrong. I now think the best hands to "semi-bluff" with are the hands that play the best against your opponent's likely calling hands. These include primarily suited aces, but also pocket pairs and suited connectors. Note that I don't think it is relevant whether you would "otherwise fold" with these hands, as Sklansky and Miller suggest. Many of these "semi-bluffing" hands are hands that you would likely have otherwise called with.
Sklansky and Miller's advice seems to be based on game-theoretically optimal river strategy, when there are no cards left to come. On the river, your bluff bets should indeed be with your worst possible hands, and your bluff raises should be with the best of the hands you would otherwise fold. This mirrors S+M's advice that you should bluff when "you have one of your worst hands" from the blinds, and that if you are not in the blinds you should bluff when "you have one of the best hands that you would otherwise fold."
The problem is that this optimal river bluffing strategy does not apply preflop. Preflop, the logic is much different because there are such things as drawing hands and semibluffs. Preflop, you do not need to bluff with terrible hands in order to induce your opponents to play hands weaker than your range; they will already be calling with some hands that they know are behind your range because they are "drawing hands." Instead, you should be bluffing with moderately strong hands as "semibluffs." The best hands to semibluff with preflop have two qualities. 1. They contain an ace, thereby reducing the chances you are up against AA. 2. They have a decent chance of making a big hand like a flush or a set. If your opponent is very tight, or if there have already been a couple of raises, point 1 is reasonably important. Otherwise, point 2 is much more important, because your opponent is unlikely to have AA anyway. In any case, this is why I suggested that the best hands to semibluff with preflop are suited aces, suited connectors, and pocket pairs.
Besides these types of hands, if you want to add some "bluffs" to your range, I think the best approach is to add some more hands that you would otherwise have called with. For example, try reraising with AJ or 77 on the button if you feel the need to bluff. Although these hands do have a lot of equity in just calling, they will also fare okay when your bluff is called. In the end, it's just my opinion, but I think this is a better approach than taking S+M's advice and raising with hands you would otherwise with fold, such as K7s. Although you are not losing any "calling equity" with K7s (since you would be folding otherwise), against normal opponents, these hands are just too unlikely to win if you are called preflop.
Saturday, October 31, 2009
Analyzing NLHE:TAP Concept 25
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 25: The button is the true bread and butter position in no limit. In many games you can play an extremely wide range of hands from the button, even for a raise.
Sklansky and Miller say that if effective stack sizes are at least 200 times the big blind, then, as long as you have at least one opponent who plays too loose after the flop, it is correct to limp on the button with over 50% of hands ("probably any two suited cards, any big offsuit cards, any ace, and any offsuit connector down to at least five-four"), and possibly with 100%. They also suggest that you call with over 30% of hands "if the raise represents only a few percent of the stacks (e.g., no more than maybe $50 with $1000 stacks)."
This is another one that is tough to analyze precisely, but in general, I think this advice is encouraging the reader to play too loosely. Unless you are much, much better than your opponents, this advice is probably -EV. I would be especially wary of the advice to call a raise to $50 with KTo in a $2-5 blind game, even if the stacks are over $1000. First of all, this advice completely neglects to consider whether the raise is coming from early or late position, which can have a drastic effect on your opponent's range. Even if the player is in late position and is rather aggressive, though, I would not be happy calling with KTo in this spot. You need to flop a straight or have the flop to contain KT or TT in order to be confident with your hand in there is a lot of action. Otherwise, a hand like this is going to have no implied odds, or worse, negative implied odds.
I do think the general idea that the button is a powerful position is correct, and indeed there are a lot more hands that can be played profitably from here than any other position (except the BB with no raise). However, in my humble opinion, S+M take the idea a bit to far here. If I were an online player, or if I took careful track of my results during poker sessions, such data might be helpful in coming to a well-informed conclusion. As it stands, all I (and also S+M, I suspect) can really do is guess.
Concept No. 25: The button is the true bread and butter position in no limit. In many games you can play an extremely wide range of hands from the button, even for a raise.
Sklansky and Miller say that if effective stack sizes are at least 200 times the big blind, then, as long as you have at least one opponent who plays too loose after the flop, it is correct to limp on the button with over 50% of hands ("probably any two suited cards, any big offsuit cards, any ace, and any offsuit connector down to at least five-four"), and possibly with 100%. They also suggest that you call with over 30% of hands "if the raise represents only a few percent of the stacks (e.g., no more than maybe $50 with $1000 stacks)."
This is another one that is tough to analyze precisely, but in general, I think this advice is encouraging the reader to play too loosely. Unless you are much, much better than your opponents, this advice is probably -EV. I would be especially wary of the advice to call a raise to $50 with KTo in a $2-5 blind game, even if the stacks are over $1000. First of all, this advice completely neglects to consider whether the raise is coming from early or late position, which can have a drastic effect on your opponent's range. Even if the player is in late position and is rather aggressive, though, I would not be happy calling with KTo in this spot. You need to flop a straight or have the flop to contain KT or TT in order to be confident with your hand in there is a lot of action. Otherwise, a hand like this is going to have no implied odds, or worse, negative implied odds.
I do think the general idea that the button is a powerful position is correct, and indeed there are a lot more hands that can be played profitably from here than any other position (except the BB with no raise). However, in my humble opinion, S+M take the idea a bit to far here. If I were an online player, or if I took careful track of my results during poker sessions, such data might be helpful in coming to a well-informed conclusion. As it stands, all I (and also S+M, I suspect) can really do is guess.
Friday, October 30, 2009
Analyzing NLHE:TAP Concept 24
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 24: If you have a hand that you would limp with in a passive game, consider making a small raise (two to three times the big blind) in an aggressive game instead of limping.
The idea here is that by raising, you reduce the chances an aggressive player will put in a big raise because he will be afraid of you. As long as your raise is rather small, this should allow you to see a flop more cheaply than if you just limp.
In my experience, this doesn't really work. After I first read this concept, I tried this strategy a few times with no success. I found that if I was in a game that was so aggressive that I felt like trying this strategy, I very often found myself facing a large reraise. Players who are hyper-aggressive preflop tend to be interested in gambling, and they will not be satisfied with your 2-3x raise. They will reraise you. Some players will even (correctly) see your abnormally small raise as a sign of weakness and raise with hands with which they may otherwise have limped or folded. In fact, in such situations you might want to consider making these small raises when you have strong preflop hand in order to trap your aggressive opponents when they reraise you.
If you are in a game where the aggressive players are not maniacs, this "blocking raise" strategy could conceivably work, but this is rarely the case. As I said, I personally have never utilized the strategy successfully, but I've only tried it a few times. If you do find yourself in this situation and want to try this "blocking raise" strategy, it's probably a good idea to also make small raises with big hands like JJ+ in order to balance your range, as S+M suggest in the discussion of this concept. Without at least some balance, a strong, aggressive opponent won't take long to figure out that your small raises indicate weakness.
Concept No. 24: If you have a hand that you would limp with in a passive game, consider making a small raise (two to three times the big blind) in an aggressive game instead of limping.
The idea here is that by raising, you reduce the chances an aggressive player will put in a big raise because he will be afraid of you. As long as your raise is rather small, this should allow you to see a flop more cheaply than if you just limp.
In my experience, this doesn't really work. After I first read this concept, I tried this strategy a few times with no success. I found that if I was in a game that was so aggressive that I felt like trying this strategy, I very often found myself facing a large reraise. Players who are hyper-aggressive preflop tend to be interested in gambling, and they will not be satisfied with your 2-3x raise. They will reraise you. Some players will even (correctly) see your abnormally small raise as a sign of weakness and raise with hands with which they may otherwise have limped or folded. In fact, in such situations you might want to consider making these small raises when you have strong preflop hand in order to trap your aggressive opponents when they reraise you.
If you are in a game where the aggressive players are not maniacs, this "blocking raise" strategy could conceivably work, but this is rarely the case. As I said, I personally have never utilized the strategy successfully, but I've only tried it a few times. If you do find yourself in this situation and want to try this "blocking raise" strategy, it's probably a good idea to also make small raises with big hands like JJ+ in order to balance your range, as S+M suggest in the discussion of this concept. Without at least some balance, a strong, aggressive opponent won't take long to figure out that your small raises indicate weakness.
Wednesday, October 28, 2009
Analyzing NLHE:TAP Concept 23
From No Limit Hold 'Em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 23: It's ok to limp in, planning to fold to a raise. It's sometimes ok even when you think a raise is likely.
This piece of advice is meant for limit players, who are used to automatically calling a raise after they limp in. I certainly agree with this advice; the only thing possibly controversial is that some would argue that you should never limp in, you should always open with a raise. I think limping is fine, though, and I do it often.
Sklansky and Miller give a good, simple example to support their stronger claim that limp-folding is "sometimes ok even when you think a raise is likely." Basically, the example says that if you think your limp for $2 will return $6 on average when nobody raises, then even if there is a 60% chance someone will make a raise and you fold, it's still profitable to limp. EV= (.6)*(-$2) + (.4)*($4) = $0.40. Although this ignores the possibility that raising could be even more profitable than limping, it proves their point that limping can be better than just folding even if you are likely going to fold anyway.
Concept No. 23: It's ok to limp in, planning to fold to a raise. It's sometimes ok even when you think a raise is likely.
This piece of advice is meant for limit players, who are used to automatically calling a raise after they limp in. I certainly agree with this advice; the only thing possibly controversial is that some would argue that you should never limp in, you should always open with a raise. I think limping is fine, though, and I do it often.
Sklansky and Miller give a good, simple example to support their stronger claim that limp-folding is "sometimes ok even when you think a raise is likely." Basically, the example says that if you think your limp for $2 will return $6 on average when nobody raises, then even if there is a 60% chance someone will make a raise and you fold, it's still profitable to limp. EV= (.6)*(-$2) + (.4)*($4) = $0.40. Although this ignores the possibility that raising could be even more profitable than limping, it proves their point that limping can be better than just folding even if you are likely going to fold anyway.
Tuesday, October 27, 2009
Analyzing NLHE:TAP Concept 22
Another installment in my analysis of the concepts at the end of No Limit Hold 'Em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 22: Ace-king is a powerful "move-in" hand and frequently moving in preflop is by far the best play with it.
I think this is a powerful idea that is underutilized by many of the good players I play with. Sklansky won me over to this idea when he introduced it in Tournament Poker for Advanced Players.
AK has a negative reputation as being kind of a "sucker" hand, since it is rarely the favorite if another player puts lots of chips in before the flop, because the other player usually has a pair or another AK. I think this reputation is undeserved. While it's true that AK has less than 50% chance against many other strong hands, it is only a big underdog against AA (6.5%-12%) and, to a lesser extent, KK (~30%). Against other pairs, AK has 43%-50%. There are three points that I think are misunderstood by players who think it's a sucker play to move all-in with AK preflop.
1. Players put way too much emphasis on having 50% equity when all-in. Whether a play is "correct" is determined by calculating (or guessing at) the play's expected value (EV). Actually heads-up winning percentage is only one factor in an EV calculation. When you are raising all-in, your EV depends only on the size of your raise, your pot equity (the amount in the pot after your raise is called times the probability you will win), and your fold equity (the amount in the pot before your raise times the probability your opponents will fold). Because fold equity is always positive, the required equity (when you are called) to make raising better than folding is always less than 50%. Having a higher probability of winning when you are all-in is nice, but it is only one factor to consider, and the difference between 50% equity and 43% equity is pretty small. Using the criterion of having a 50% or better chance against a certain hand or range of hands is wrong. It misleads people into thinking AK is worse than it really is because AK often falls just below this threshhold.
2. If you hold AK, this reduces the likelihood an opponent could hold AA or KK, which are really the only two hands you need to worry about. Players do understand this, but I think the effect is underrated. There are only half as many ways to make AA or KK when only three of each are left in the deck. If you hold AK and your opponent's range is AA-JJ, there is only a 1/3 chance he holds AA or KK, and your equity is about 35% (38% with AK suited). Note that when your opponent's range is this small, your fold equity is likely quite large.
3. Many players tend to view a big preflop raise by a good player as AA. They will often fold JJ or QQ, and some will even fold KK in certain situations. Although this means you are usually way behind when you are called, your fold equity is enormous.
Good players tend to play very tight ranges when there is a lot of action preflop and not much money left to play with after the flop. Players often fall into the habit of playing only AA or KK in these situations, sometimes QQ. AK is the perfect hand with which to balance your big-pot play before the flop. In some situations where your opponent's range is especially strong, it can certainly be correct to fold AK before the flop, but I think this is done way too often. Players are missing out on lots of situations where moving all-in with AK (and especially AK suited) is a +EV play.
Concept No. 22: Ace-king is a powerful "move-in" hand and frequently moving in preflop is by far the best play with it.
I think this is a powerful idea that is underutilized by many of the good players I play with. Sklansky won me over to this idea when he introduced it in Tournament Poker for Advanced Players.
AK has a negative reputation as being kind of a "sucker" hand, since it is rarely the favorite if another player puts lots of chips in before the flop, because the other player usually has a pair or another AK. I think this reputation is undeserved. While it's true that AK has less than 50% chance against many other strong hands, it is only a big underdog against AA (6.5%-12%) and, to a lesser extent, KK (~30%). Against other pairs, AK has 43%-50%. There are three points that I think are misunderstood by players who think it's a sucker play to move all-in with AK preflop.
1. Players put way too much emphasis on having 50% equity when all-in. Whether a play is "correct" is determined by calculating (or guessing at) the play's expected value (EV). Actually heads-up winning percentage is only one factor in an EV calculation. When you are raising all-in, your EV depends only on the size of your raise, your pot equity (the amount in the pot after your raise is called times the probability you will win), and your fold equity (the amount in the pot before your raise times the probability your opponents will fold). Because fold equity is always positive, the required equity (when you are called) to make raising better than folding is always less than 50%. Having a higher probability of winning when you are all-in is nice, but it is only one factor to consider, and the difference between 50% equity and 43% equity is pretty small. Using the criterion of having a 50% or better chance against a certain hand or range of hands is wrong. It misleads people into thinking AK is worse than it really is because AK often falls just below this threshhold.
2. If you hold AK, this reduces the likelihood an opponent could hold AA or KK, which are really the only two hands you need to worry about. Players do understand this, but I think the effect is underrated. There are only half as many ways to make AA or KK when only three of each are left in the deck. If you hold AK and your opponent's range is AA-JJ, there is only a 1/3 chance he holds AA or KK, and your equity is about 35% (38% with AK suited). Note that when your opponent's range is this small, your fold equity is likely quite large.
3. Many players tend to view a big preflop raise by a good player as AA. They will often fold JJ or QQ, and some will even fold KK in certain situations. Although this means you are usually way behind when you are called, your fold equity is enormous.
Good players tend to play very tight ranges when there is a lot of action preflop and not much money left to play with after the flop. Players often fall into the habit of playing only AA or KK in these situations, sometimes QQ. AK is the perfect hand with which to balance your big-pot play before the flop. In some situations where your opponent's range is especially strong, it can certainly be correct to fold AK before the flop, but I think this is done way too often. Players are missing out on lots of situations where moving all-in with AK (and especially AK suited) is a +EV play.
Sunday, October 25, 2009
Standard Work Week
Starting this coming week, I'll be working Monday-Friday for the first time since I started playing poker professionally. This isn't much of a shift; I was already working 11AM-7pm Tuesday-Friday and noon-8pm on Saturday. It used to be the case that poker games were juicier on the weekends, but for some reason, things have been slower recently on Saturdays than the rest of the week. Today, the $500+ NL game never even got started. This is not uncommon, which is the reason my supervisor decided to change my schedule.
I noticed last week that a local bar has poker every Saturday. I might check it out, but I doubt it will be worth my time. They have some sort of point system and at the end of the year someone gets a prize. I was also invited to a house game supposedly run by Koreans in downtown LA. I'm told it runs 8pm-4am on Tuesdays and Thursdays and that there are incredible amounts of money to be won. I don't think I'll go. Even if it felt safe to me, I have to work both those days, and I would find it difficult to play for 8 hours at the Bike and then head off to another game.
I've been seeing and meeting some actors and other entertainment industry types recently at the Bike. One of the producers of Zombieland plays the $500NL sometimes. Teri Hatcher came and played the $300-$500 NL game at least twice this month. Michael Muhney, who was in Veronica Mars and now stars in The Young and the Restless played $500NL with us for a few hours yesterday. There was a guy who claimed to have been a drummer for Parliament Funkadelic, but I forgot his name. He bragged of a patent he has for some denim jeans design. Bo Koster of the band My Morning Jacket plays quite often. He's a good guy and a good player. I've chatted with him quite a bit, but I haven't seen him at the Bike for the past couple weeks.
I noticed last week that a local bar has poker every Saturday. I might check it out, but I doubt it will be worth my time. They have some sort of point system and at the end of the year someone gets a prize. I was also invited to a house game supposedly run by Koreans in downtown LA. I'm told it runs 8pm-4am on Tuesdays and Thursdays and that there are incredible amounts of money to be won. I don't think I'll go. Even if it felt safe to me, I have to work both those days, and I would find it difficult to play for 8 hours at the Bike and then head off to another game.
I've been seeing and meeting some actors and other entertainment industry types recently at the Bike. One of the producers of Zombieland plays the $500NL sometimes. Teri Hatcher came and played the $300-$500 NL game at least twice this month. Michael Muhney, who was in Veronica Mars and now stars in The Young and the Restless played $500NL with us for a few hours yesterday. There was a guy who claimed to have been a drummer for Parliament Funkadelic, but I forgot his name. He bragged of a patent he has for some denim jeans design. Bo Koster of the band My Morning Jacket plays quite often. He's a good guy and a good player. I've chatted with him quite a bit, but I haven't seen him at the Bike for the past couple weeks.
Thursday, October 22, 2009
Analyzing NLHE:TAP Concept 21
Here's my analysis of the next concept from No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 21: Sometimes you can try for a deep check-raise with the nuts (or close to it).
By "deep" the authors mean that you are in middle-late position in a large field. I have seen it recommended that you not try a check-raise in such a situation, because some of the reasons for a check-raise are diminished in a large field, while the risk of being outdrawn is greatly increased. Still, I think Sklansky and Miller are right about this, and I would actually say you often should try for a deep check-raise with the nuts. It's very difficult to do a proper EV analysis of this situation, but here's my subjective opinion, for what it's worth.
In a small field, check-raising with a strong hand is standard (although there are situations where you would want to bet out even with the nuts). One reason for this is that a late position player is more likely to bet behind you if he is facing a small field. It's very common for the player in last position to bet if the first player or two check, because he will often be able to take the pot with a bet in this situation. A second reason is that the downside risk of checking is not as great when there is a small field; even if nobody bets on the flop, it's very unlikely that anyone will outdraw you, because you only have one or two opponents. A third reason to check-raise with strong hands when there is a small field is that it disguises your checks when you miss a flop. If your opponent knows you might check-raise, he'll be slightly more likely to check the flop and let you see the turn for free.
These reasons for check-raising do not hold up as well when you are in middle-late position in a large field. You need a stronger hand to check-raise in this situation. Still, with hands that are nearly the nuts, I think it's often correct to try it. A flopped set or straight is only a little vulnerable to being outdrawn, and the reward for a successful check-raise can be quite huge if there are multiple callers. If nobody bets, often another player will make two pair or three of a kind on the turn, which is likely to result in a very big win.
Concept No. 21: Sometimes you can try for a deep check-raise with the nuts (or close to it).
By "deep" the authors mean that you are in middle-late position in a large field. I have seen it recommended that you not try a check-raise in such a situation, because some of the reasons for a check-raise are diminished in a large field, while the risk of being outdrawn is greatly increased. Still, I think Sklansky and Miller are right about this, and I would actually say you often should try for a deep check-raise with the nuts. It's very difficult to do a proper EV analysis of this situation, but here's my subjective opinion, for what it's worth.
In a small field, check-raising with a strong hand is standard (although there are situations where you would want to bet out even with the nuts). One reason for this is that a late position player is more likely to bet behind you if he is facing a small field. It's very common for the player in last position to bet if the first player or two check, because he will often be able to take the pot with a bet in this situation. A second reason is that the downside risk of checking is not as great when there is a small field; even if nobody bets on the flop, it's very unlikely that anyone will outdraw you, because you only have one or two opponents. A third reason to check-raise with strong hands when there is a small field is that it disguises your checks when you miss a flop. If your opponent knows you might check-raise, he'll be slightly more likely to check the flop and let you see the turn for free.
These reasons for check-raising do not hold up as well when you are in middle-late position in a large field. You need a stronger hand to check-raise in this situation. Still, with hands that are nearly the nuts, I think it's often correct to try it. A flopped set or straight is only a little vulnerable to being outdrawn, and the reward for a successful check-raise can be quite huge if there are multiple callers. If nobody bets, often another player will make two pair or three of a kind on the turn, which is likely to result in a very big win.
Sunday, October 18, 2009
Analyzing NLHE:TAP Concept 20
After this post, I'll be one third 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. 20: Sometimes you should limp behind limpers with pocket aces.
I agree with this. It does sound a little like it's recommending that you "randomize" your play, which I've argued against previously. However, this one is a little different for two reasons.
First, Sklansky and Miller are not actually recommending that you apply this advice in a random manner. Instead, they say, "you'd do this if you have opponents yet to act who like to raise a series of limpers with weak hands." I agree, but I think there are also other situations where you may want to limp behind limpers with AA. Basically, any situation where you think it likely that someone behind you will raise is a good time to limp with AA. In fact, sometimes you should be limping with weaker hands, as well. This week I was sitting to the right of a maniac, and I literally stopped raising preflop with any hands because it was so likely that the maniac would reopen the betting for me if I just limped. This was a great situation because I got the best relative position before the flop, meaning I got to see how everyone else reacted to the maniac's raises before I had to decide how to proceed. This is an extreme situation, but even if the player to your left is only somewhat maniacal, limp-raising with JJ+ (or even weaker) may be correct.
Second, although randomizing your play is not usually a good idea, it can be theoretically correct to randomize your play in certain situations with the best possible hand. Before the flop, AA is the best possible hand, and it can be useful to include it in your limp-raising range. Even if you are never in a game with a maniac or with players that like to try to steal before the flop after several limpers, it still might be a good idea to limp randomly sometimes with AA behind other limpers. This play is probably only worthwhile if your opponents do not expect you would try such a thing. If they already do expect you'd try this play, it's probably not worth actually doing; its value is in its deceptiveness.
Concept No. 20: Sometimes you should limp behind limpers with pocket aces.
I agree with this. It does sound a little like it's recommending that you "randomize" your play, which I've argued against previously. However, this one is a little different for two reasons.
First, Sklansky and Miller are not actually recommending that you apply this advice in a random manner. Instead, they say, "you'd do this if you have opponents yet to act who like to raise a series of limpers with weak hands." I agree, but I think there are also other situations where you may want to limp behind limpers with AA. Basically, any situation where you think it likely that someone behind you will raise is a good time to limp with AA. In fact, sometimes you should be limping with weaker hands, as well. This week I was sitting to the right of a maniac, and I literally stopped raising preflop with any hands because it was so likely that the maniac would reopen the betting for me if I just limped. This was a great situation because I got the best relative position before the flop, meaning I got to see how everyone else reacted to the maniac's raises before I had to decide how to proceed. This is an extreme situation, but even if the player to your left is only somewhat maniacal, limp-raising with JJ+ (or even weaker) may be correct.
Second, although randomizing your play is not usually a good idea, it can be theoretically correct to randomize your play in certain situations with the best possible hand. Before the flop, AA is the best possible hand, and it can be useful to include it in your limp-raising range. Even if you are never in a game with a maniac or with players that like to try to steal before the flop after several limpers, it still might be a good idea to limp randomly sometimes with AA behind other limpers. This play is probably only worthwhile if your opponents do not expect you would try such a thing. If they already do expect you'd try this play, it's probably not worth actually doing; its value is in its deceptiveness.
Friday, October 16, 2009
Analyzing NLHE:TAP Concept 19
Here is another in my series analyzing the concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 19: Don't call in protected pots without a very good hand.
By "protected pot," Sklansky and Miller mean a pot where bluffing will obviously not work. The most common example is when one player is all-in. Then, there is no point in bluffing because the all-in player will probably win the pot even if you get the other player to fold.
I think this should be good advice in theory, but in practice, it is astonishing how often players will bluff into protected pots. I generally follow S+M's advice and fold my moderate hands in such situations, but, in so doing, I've been bluffed out of many pots. In fact, I've recently decided to relax my calling standards slightly except against thoughtful opponents.
Sklansky and Miller extend this advice to pots that are "protected" by a loose player in the field or by a player who is nearly all-in. Unless you know your opponent is an alert and logical player, I would not suggest taking this advice very seriously. Players still sometimes bluff in these situations. Many will fail to notice that these factors are in play, and many who do notice it will still fail to realize that this means the pot is protected and that they should not try bluffing. This is my personal experience, so take it with a grain of salt.
Surprisingly, the seemingly obvious idea that you should not bluff when a player is all-in (and there is no side-pot) is not supported by game theory. Bill Chen and Jerrod Ankenman show in Chapter 29 of The Mathematics of Poker that it actually can be theoretically advantageous to bluff into protected pots. This is basically because it forces your opponent to play much more loosely. In "theory," all players always know the strategy of their opponents. In practice, of course, this is not the case, and your opponent probably will not expect you to bluff into a protected pot. So, bluffing in these situations is probably never a good idea after all. Actually, if you can get your opponent to suspect that you might bluff into a protected pot, you can forgo ever actually making such bluffs and still gain the benefit of making your opponent play too loosely. I'm not sure how you convince your opponents of this. Maybe pretend to be very drunk?
Concept No. 19: Don't call in protected pots without a very good hand.
By "protected pot," Sklansky and Miller mean a pot where bluffing will obviously not work. The most common example is when one player is all-in. Then, there is no point in bluffing because the all-in player will probably win the pot even if you get the other player to fold.
I think this should be good advice in theory, but in practice, it is astonishing how often players will bluff into protected pots. I generally follow S+M's advice and fold my moderate hands in such situations, but, in so doing, I've been bluffed out of many pots. In fact, I've recently decided to relax my calling standards slightly except against thoughtful opponents.
Sklansky and Miller extend this advice to pots that are "protected" by a loose player in the field or by a player who is nearly all-in. Unless you know your opponent is an alert and logical player, I would not suggest taking this advice very seriously. Players still sometimes bluff in these situations. Many will fail to notice that these factors are in play, and many who do notice it will still fail to realize that this means the pot is protected and that they should not try bluffing. This is my personal experience, so take it with a grain of salt.
Surprisingly, the seemingly obvious idea that you should not bluff when a player is all-in (and there is no side-pot) is not supported by game theory. Bill Chen and Jerrod Ankenman show in Chapter 29 of The Mathematics of Poker that it actually can be theoretically advantageous to bluff into protected pots. This is basically because it forces your opponent to play much more loosely. In "theory," all players always know the strategy of their opponents. In practice, of course, this is not the case, and your opponent probably will not expect you to bluff into a protected pot. So, bluffing in these situations is probably never a good idea after all. Actually, if you can get your opponent to suspect that you might bluff into a protected pot, you can forgo ever actually making such bluffs and still gain the benefit of making your opponent play too loosely. I'm not sure how you convince your opponents of this. Maybe pretend to be very drunk?
Analyzing NLHE:TAP Concept 18
Continuing my project of analyzing each of the sixty concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 18: Don't get trapped with a fourth street top pair in multiway checked pots.
Wow, Sklansky and Miller are really going out on a limb with this one! "Don't get trapped" is pretty hard to argue with. The question is whether this advice is likely to help anyone.
Actually, I think this is marginally good advice, but it's too specific. There are plenty of circumstances where you can get trapped by underestimating the chance that someone in a large field of opponents is lurking with a big hand. For example, top pair is also a treacherous hand to call with in multiways pots on the flop. Still, this doesn't invalidate the more specific point made in this concept.
It's true that you need a much stronger hand to call a bet when there are several players behind you than when you are heads-up, especially if there are various draws on the board. A top pair hand with a decent kicker is much weaker than it looks if you're used to having only one or two opponents on a flop, and thus it's pretty common for players to get trapped with such hands. In their discussion, Sklansky and Miller say "there's a decent chance you have the best hand, yet this isn't reason enough to call." This is a good point. This idea seems paradoxical at first, and it would make for a more interesting Concept topic, in my opinion.
Concept No. 18: Don't get trapped with a fourth street top pair in multiway checked pots.
Wow, Sklansky and Miller are really going out on a limb with this one! "Don't get trapped" is pretty hard to argue with. The question is whether this advice is likely to help anyone.
Actually, I think this is marginally good advice, but it's too specific. There are plenty of circumstances where you can get trapped by underestimating the chance that someone in a large field of opponents is lurking with a big hand. For example, top pair is also a treacherous hand to call with in multiways pots on the flop. Still, this doesn't invalidate the more specific point made in this concept.
It's true that you need a much stronger hand to call a bet when there are several players behind you than when you are heads-up, especially if there are various draws on the board. A top pair hand with a decent kicker is much weaker than it looks if you're used to having only one or two opponents on a flop, and thus it's pretty common for players to get trapped with such hands. In their discussion, Sklansky and Miller say "there's a decent chance you have the best hand, yet this isn't reason enough to call." This is a good point. This idea seems paradoxical at first, and it would make for a more interesting Concept topic, in my opinion.
Monday, October 12, 2009
Analyzing NLHE:TAP Concept 17
From No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 17: If your preflop raise is called behind you, check a lot of flops.
The standard play is to continuation-bet in these situations, in order to maintain the initiative and keep the pressure on your opponent(s). I think this standard c-bet is probably overused, so in this sense Sklansky and Miller's advice to check a lot of flops is good. Still, I disagree with the reasoning they offer in their discussion of this concept. They focus too much on your hand and on randomizing your play, and not enough on the texture of the flop. I think better advice would be, "... often check on certain types of flops." This is mostly based on my intuition, and I won't be presenting any quantitative evidence.
I often do like to continuation bet even if I miss a flop because many players like to call preflop raises with small pocket pairs or suited connectors, hoping to catch a set, two pair, or a big draw and win a big pot against the preflop raiser. Usually, these hands will miss the flop, and they will fold to a continuation bet. For example, if I have AT and the flop comes Q43, I have a decent chance to win with a continuation bet even though my opponents often have a small pair or a healthy six outs. On this flop it's also a good idea to bet with many of your other likely preflop raising hands for similar reasons. With a good hand like KK, you can bet for value to balance all your weaker hands. However, there are certain types of flops where the continuation bet seems counter-productive. Generally, the flops where you don't want to bet are ones where there's a good chance you have the best hand and not too much risk of being outdrawn. For example, I think KK on a board of A92 rainbow is often better off just checking. On this flop, you're probably okay checking lots of your other likely hands, too. This includes strong hands such as AT. By waiting until the turn to bet with these hands, you are probably more likely to get called by a small pocket pair.
There are also more extreme examples of flops that don't have many draws, such as AA6 rainbow, or, to a lesser extent, 66T. On paired flop such as these I will often make a small bet. If my opponent has something, they'll call or raise, and I can react according to the strength of my hand. Often, my opponents will have completely missed these flops and just fold.
This seems like it could make for a bigger project if I want to come back to it later and do more quantitative analysis. For example, I'd like to test whether it's better to just alter my bet-sizes based on the texture of a flop rather than changing my betting frequency.
Concept No. 17: If your preflop raise is called behind you, check a lot of flops.
The standard play is to continuation-bet in these situations, in order to maintain the initiative and keep the pressure on your opponent(s). I think this standard c-bet is probably overused, so in this sense Sklansky and Miller's advice to check a lot of flops is good. Still, I disagree with the reasoning they offer in their discussion of this concept. They focus too much on your hand and on randomizing your play, and not enough on the texture of the flop. I think better advice would be, "... often check on certain types of flops." This is mostly based on my intuition, and I won't be presenting any quantitative evidence.
I often do like to continuation bet even if I miss a flop because many players like to call preflop raises with small pocket pairs or suited connectors, hoping to catch a set, two pair, or a big draw and win a big pot against the preflop raiser. Usually, these hands will miss the flop, and they will fold to a continuation bet. For example, if I have AT and the flop comes Q43, I have a decent chance to win with a continuation bet even though my opponents often have a small pair or a healthy six outs. On this flop it's also a good idea to bet with many of your other likely preflop raising hands for similar reasons. With a good hand like KK, you can bet for value to balance all your weaker hands. However, there are certain types of flops where the continuation bet seems counter-productive. Generally, the flops where you don't want to bet are ones where there's a good chance you have the best hand and not too much risk of being outdrawn. For example, I think KK on a board of A92 rainbow is often better off just checking. On this flop, you're probably okay checking lots of your other likely hands, too. This includes strong hands such as AT. By waiting until the turn to bet with these hands, you are probably more likely to get called by a small pocket pair.
There are also more extreme examples of flops that don't have many draws, such as AA6 rainbow, or, to a lesser extent, 66T. On paired flop such as these I will often make a small bet. If my opponent has something, they'll call or raise, and I can react according to the strength of my hand. Often, my opponents will have completely missed these flops and just fold.
This seems like it could make for a bigger project if I want to come back to it later and do more quantitative analysis. For example, I'd like to test whether it's better to just alter my bet-sizes based on the texture of a flop rather than changing my betting frequency.
Sunday, October 11, 2009
Analyzing NLHE:TAP Concept 16
Okay, let's keep the ball rolling. Here's the next installment in my analysis of the concepts at the end of No Limit Hold 'Em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 16: Occasionally overbet with moderate hands to disguise your overbets with excellent hands.
No; this seems like terrible advice. By "occasionally," I think Sklansky and Miller mean something like "randomly," and I've already discussed in my Concept 3 analysis why I think this idea is vastly overrated. In their discussion, S+M say "as long as you don't do it too often, these overbets won't cost you too much, and they will support you those times you make big bets with excellent hands." I think this is wrong. If a play costs you anything in the long run, it should be considered "too much," because you can just fold and lose nothing more. If you determine that a certain play is +EV, you should do it every time, not just occasionally. If it's -EV, never do it (except maybe in some extremely rare instances if you really know what you're doing). Something I forgot to mention about randomizing your play in my Concept 3 analysis is that if you ever make mistakes when you play poker (ie, make a play that is -EV), you are already randomizing your play. Don't make matters worse by adding extra mistakes to your game! Of course, I assume everyone makes mistakes, so this advice should apply to everyone; to some extent, human error automatically disguises your hands.
There's another, more obvious, explanation for why this concept's advice is bad. If you want to "disguise your overbets with excellent hands," the best way to do this is by overbetting with draws as semibluffs. The example in the book suggests occasionally overbetting with KQ on a board of Ks9s7c. This is going to force your opponent to fold most of the hands you can beat. When you are behind, you will be called or raised and have only about three outs. I would suggest simply value-betting KQ while overbetting with hands like T8 or flush draws as well as with your monster hands. Semibluffs are great because your opponents are likely to fold better hands than yours, but you have lots of outs if you get called. Neither of these advantages exist when you overbet with mediocre hand like KQ on this flop. Moreover, semibluffs actually do a better job of "supporting" your overbets with excellent hands, because if your opponent has KJ or KT, he might think about calling if he suspects you are on a draw. If he thinks you have at worst KQ, he will fold right away.
Concept No. 16: Occasionally overbet with moderate hands to disguise your overbets with excellent hands.
No; this seems like terrible advice. By "occasionally," I think Sklansky and Miller mean something like "randomly," and I've already discussed in my Concept 3 analysis why I think this idea is vastly overrated. In their discussion, S+M say "as long as you don't do it too often, these overbets won't cost you too much, and they will support you those times you make big bets with excellent hands." I think this is wrong. If a play costs you anything in the long run, it should be considered "too much," because you can just fold and lose nothing more. If you determine that a certain play is +EV, you should do it every time, not just occasionally. If it's -EV, never do it (except maybe in some extremely rare instances if you really know what you're doing). Something I forgot to mention about randomizing your play in my Concept 3 analysis is that if you ever make mistakes when you play poker (ie, make a play that is -EV), you are already randomizing your play. Don't make matters worse by adding extra mistakes to your game! Of course, I assume everyone makes mistakes, so this advice should apply to everyone; to some extent, human error automatically disguises your hands.
There's another, more obvious, explanation for why this concept's advice is bad. If you want to "disguise your overbets with excellent hands," the best way to do this is by overbetting with draws as semibluffs. The example in the book suggests occasionally overbetting with KQ on a board of Ks9s7c. This is going to force your opponent to fold most of the hands you can beat. When you are behind, you will be called or raised and have only about three outs. I would suggest simply value-betting KQ while overbetting with hands like T8 or flush draws as well as with your monster hands. Semibluffs are great because your opponents are likely to fold better hands than yours, but you have lots of outs if you get called. Neither of these advantages exist when you overbet with mediocre hand like KQ on this flop. Moreover, semibluffs actually do a better job of "supporting" your overbets with excellent hands, because if your opponent has KJ or KT, he might think about calling if he suspects you are on a draw. If he thinks you have at worst KQ, he will fold right away.
Saturday, October 10, 2009
Analyzing NLHE:TAP Concept 15
After this post, I'll be one quarter of the way through the sixty concepts at the end of No Limit Hold 'em: Theory and Practice by Sklansky and Miller.
Concept No. 15: Bet more than usual when your opponent likely has a hand that he thinks might be good.
Taken literally, it seems to me that Concept 15 is useless. A recurring problem I'm having with these analyses is that the concepts are worded so vaguely or figuratively as to lose any clear meaning. In these cases, in order to do my analysis, I feel obliged to parse the concepts into something that both makes some sense and could plausibly be what Sklansky and Miller were trying to say.
Let's look closely at concept 15 to see if we can glean something from it. It starts out "Bet more than usual when..." Okay, so this suggests that in "usual" cases, Sklansky and Miller assume that we will bet within a certain range of amounts, but that in some unusual cases (to be revealed at the end of the concept's sentence) we should be betting more. So far, so good. Let's look at the end of the sentence to find out what these unusual cases are: "... when your opponent likely has a hand that he thinks might be good." Hmm. This doesn't seem like an unusual circumstance. In fact, since we almost never know what our opponent is holding or thinking, it seems to me that it's almost always the case that it's at least somewhat likely that he thinks his hand might be good. Sklansky and Miller's unusual circumstance is actually the norm.
As worded, I don't think the concept makes enough sense to be analyzed. Let's try to salvage it by giving it some plausible interpretation that can be analyzed. Using clues from S+M's analysis of this concept, I think they are trying to say something like: "If you hold the nuts and you somehow know your opponent has a made hand (as opposed to a drawing hand), bet more than you would if you knew he was on a draw." I think this must be what S+M are trying to say; I'm looking back at the book at the chapter on "bet sizing" (p54), and this is essentially the advice given.
The standard here is still too vague because not all draws are alike. If I know my opponent has 21 outs going to the river, he has a 21/44 chance of winning. Then, ignoring implied odds, the correct play is to bet over 10.5 times the size of the pot and hope he makes a bad call (if you don't have this much, just go all-in). On the other end of the spectrum, my opponent might have only 1 out. Here (again ignoring implied odds), you need only bet over 1/43 of the pot. Suppose the pot is $430. In the first case you should be betting over $4515. In the second case you should be betting over $10. So there isn't really a "usual" amount to bet if you somehow knew your opponent was on a draw unless you knew how many outs he had.
Let's refine the advice a little further and say, "If you hold the nuts and you somehow know your opponent has a made hand (as opposed to a drawing hand), make a big bet of around the size of the pot." In the book, one example of a "normal" bet against a draw is 1/3 of the pot, so a pot-sized bet seems like it should qualify as big.
I still don't think this advice is good in general. First of all, you almost never actually know what sort of hand your opponent is holding, so it makes the whole argument moot. Supposing you could know that your opponent didn't have a drawing hand hand, the correct bet size would still depend on your opponent's hand range and your opponent himself. Some players get very suspicious of very big bets because they suspect they are bluffs. Against such players, it really is a good idea to be big with the nuts, but not all players are like this. You also need to consider your opponent's hand range. The stronger his range, the more you should bet. In particular, if your opponent's range is very weak, you should usually be betting almost nothing. This invalidates the advice that you should generally be making a big bet. Sometimes yes, sometimes no.
Concept No. 15: Bet more than usual when your opponent likely has a hand that he thinks might be good.
Taken literally, it seems to me that Concept 15 is useless. A recurring problem I'm having with these analyses is that the concepts are worded so vaguely or figuratively as to lose any clear meaning. In these cases, in order to do my analysis, I feel obliged to parse the concepts into something that both makes some sense and could plausibly be what Sklansky and Miller were trying to say.
Let's look closely at concept 15 to see if we can glean something from it. It starts out "Bet more than usual when..." Okay, so this suggests that in "usual" cases, Sklansky and Miller assume that we will bet within a certain range of amounts, but that in some unusual cases (to be revealed at the end of the concept's sentence) we should be betting more. So far, so good. Let's look at the end of the sentence to find out what these unusual cases are: "... when your opponent likely has a hand that he thinks might be good." Hmm. This doesn't seem like an unusual circumstance. In fact, since we almost never know what our opponent is holding or thinking, it seems to me that it's almost always the case that it's at least somewhat likely that he thinks his hand might be good. Sklansky and Miller's unusual circumstance is actually the norm.
As worded, I don't think the concept makes enough sense to be analyzed. Let's try to salvage it by giving it some plausible interpretation that can be analyzed. Using clues from S+M's analysis of this concept, I think they are trying to say something like: "If you hold the nuts and you somehow know your opponent has a made hand (as opposed to a drawing hand), bet more than you would if you knew he was on a draw." I think this must be what S+M are trying to say; I'm looking back at the book at the chapter on "bet sizing" (p54), and this is essentially the advice given.
The standard here is still too vague because not all draws are alike. If I know my opponent has 21 outs going to the river, he has a 21/44 chance of winning. Then, ignoring implied odds, the correct play is to bet over 10.5 times the size of the pot and hope he makes a bad call (if you don't have this much, just go all-in). On the other end of the spectrum, my opponent might have only 1 out. Here (again ignoring implied odds), you need only bet over 1/43 of the pot. Suppose the pot is $430. In the first case you should be betting over $4515. In the second case you should be betting over $10. So there isn't really a "usual" amount to bet if you somehow knew your opponent was on a draw unless you knew how many outs he had.
Let's refine the advice a little further and say, "If you hold the nuts and you somehow know your opponent has a made hand (as opposed to a drawing hand), make a big bet of around the size of the pot." In the book, one example of a "normal" bet against a draw is 1/3 of the pot, so a pot-sized bet seems like it should qualify as big.
I still don't think this advice is good in general. First of all, you almost never actually know what sort of hand your opponent is holding, so it makes the whole argument moot. Supposing you could know that your opponent didn't have a drawing hand hand, the correct bet size would still depend on your opponent's hand range and your opponent himself. Some players get very suspicious of very big bets because they suspect they are bluffs. Against such players, it really is a good idea to be big with the nuts, but not all players are like this. You also need to consider your opponent's hand range. The stronger his range, the more you should bet. In particular, if your opponent's range is very weak, you should usually be betting almost nothing. This invalidates the advice that you should generally be making a big bet. Sometimes yes, sometimes no.
Tuesday, October 06, 2009
Analyzing NLHE:TAP Concept 14
I'm offering analysis of each of the concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller.
Concept No. 14: Raise less often than you would in limit, because raising reopens the betting, and that's riskier to do in no limit.
I think this is good advice, but not entirely for the reason given. The advice actually holds even if you know your opponent will not re-raise you, such as when you are raising all-in. This is because in no limit your raise is almost certainly going to be larger (in proportion to the pot size) than in a limit game, which means your opponent will need a stronger hand in order to call you. If your opponent is only calling with stronger hands, then you can only profitably bet with even stronger hands. Thus, you need to be raising less often. (You may protest that since your opponent is folding more often, you should be bluffing more often. In fact, you do need to be bluffing more often, but only as a proportion of your raises. Since you are value-raising less often, your bluffing proportion automatically increases. In order to achieve a reasonable bluffing rate when making big bets, it's probably not actually necessary to bluff more often.)
The reason Sklansky and Miller give, namely that it's riskier to reopen the betting in no-limit than in limit, is also true. If you hold a hand that has some outs against possible raising hands in your opponent's range, it's sometimes best to just call in no-limit even if it would be better to raise in a limit game. This is partly because it's simply more expensive to be caught raising with a second best hand in a no-limit game, but another reason is that if you are raised, you can often call a small bet (such as in a limit game) with a drawing hand for +EV, but if you're forced to fold to a larger raise (such as in a no-limit game), this is necessarily 0 EV (actually -EV if you count the loss of the amount you just raised). Even if you can profitably call a large raise, it will not be as profitable as it would be if you only had to call a smaller amount.
Concept No. 14: Raise less often than you would in limit, because raising reopens the betting, and that's riskier to do in no limit.
I think this is good advice, but not entirely for the reason given. The advice actually holds even if you know your opponent will not re-raise you, such as when you are raising all-in. This is because in no limit your raise is almost certainly going to be larger (in proportion to the pot size) than in a limit game, which means your opponent will need a stronger hand in order to call you. If your opponent is only calling with stronger hands, then you can only profitably bet with even stronger hands. Thus, you need to be raising less often. (You may protest that since your opponent is folding more often, you should be bluffing more often. In fact, you do need to be bluffing more often, but only as a proportion of your raises. Since you are value-raising less often, your bluffing proportion automatically increases. In order to achieve a reasonable bluffing rate when making big bets, it's probably not actually necessary to bluff more often.)
The reason Sklansky and Miller give, namely that it's riskier to reopen the betting in no-limit than in limit, is also true. If you hold a hand that has some outs against possible raising hands in your opponent's range, it's sometimes best to just call in no-limit even if it would be better to raise in a limit game. This is partly because it's simply more expensive to be caught raising with a second best hand in a no-limit game, but another reason is that if you are raised, you can often call a small bet (such as in a limit game) with a drawing hand for +EV, but if you're forced to fold to a larger raise (such as in a no-limit game), this is necessarily 0 EV (actually -EV if you count the loss of the amount you just raised). Even if you can profitably call a large raise, it will not be as profitable as it would be if you only had to call a smaller amount.
Sunday, October 04, 2009
Analyzing NLHE:TAP Concept 13
Okay, time to get back in the saddle and analyze another one of the concepts at the end of No Limit Hold 'em: Theory and Practice by Sklansky and Miller.
Concept No. 13: If you're thinking about raising, but you wouldn't know how to respond to an all-in reraise, usually you should either move in yourself or raise a smaller amount (that would allow you to fold easily to a reraise).
Here in Concept 13, Sklansky and Miller advise that we should try to avoid situations where we would not know how to respond to a reraise. This is essentially the same as saying that we should avoid a situation if it forces us into a difficult decision. As I said at the end of my analysis of Concept 7, "I think conventional Sklansky-wisdom is overly concerned with playing hands in a way that avoids difficult decisions." I've often heard the advice on TV, in writing, and from other players that it's good to put the other player "to the test" by betting in such a way that you expect they have a difficult situation. I consider this advice to be pretty dubious (players will often pass this test!), and I think that Sklansky is applying a corollary: if it's good to force your opponent into a difficult decision, it must be bad to force yourself into one.
As poker players, our primary job is to make decisions that maximize our Expected Value at every point. Thus, we can define a "difficult decision" as one where we have two or more options (eg calling or folding) that have approximately equal EV, thus making it hard to choose which is best. Since folding always has EV=0, any difficult decision that includes folding as a candidate will have EV of approximately 0. We cannot make money in situations where we have EV=0, and I suppose this is why Sklansky advises to avoid them. However, I disagree with the notion that these situations should be avoided. We have to make our decisions with incomplete information, and when we initially bet, we do not know that we are going to be raised. When the raise comes, we must re-assess the situation, accounting for the fact that our opponent's range is now stronger than it was a moment ago, before we knew he would raise. Sure, it's unfortunate your opponent has a stronger hand than you realized, but that does not invalidate the decision to bet (except in hindsight).
To take an example from limit poker, if you bet and are raised and aren't sure how to respond, this implies that your hand fell solidly into your range of "betting hands." You should certainly not be checking with such hands just to avoid a difficult situation.
The situation described here in Concept 13 is a bit different. It may even be technically true. Sklansky and Miller use the term "usually," so it's hard to disprove unless I can show that it is never correct. In fact, S+M contrive a situation in the book in which the advice does indeed hold; my question is whether their example is representative of the "usual" situation. This involves making arguments about how our opponents are most likely to respond to different sized raises with particular hands, and so, really, it's just a matter of opinion. It is probably possible to find the game-theory answer (where we assume our opponent plays optimally). Even this will not give a definitive answer as to what is the best play against actual human opponents, but I might come back to this later if I feel like doing some more hard-core game theory.
Concept No. 13: If you're thinking about raising, but you wouldn't know how to respond to an all-in reraise, usually you should either move in yourself or raise a smaller amount (that would allow you to fold easily to a reraise).
Here in Concept 13, Sklansky and Miller advise that we should try to avoid situations where we would not know how to respond to a reraise. This is essentially the same as saying that we should avoid a situation if it forces us into a difficult decision. As I said at the end of my analysis of Concept 7, "I think conventional Sklansky-wisdom is overly concerned with playing hands in a way that avoids difficult decisions." I've often heard the advice on TV, in writing, and from other players that it's good to put the other player "to the test" by betting in such a way that you expect they have a difficult situation. I consider this advice to be pretty dubious (players will often pass this test!), and I think that Sklansky is applying a corollary: if it's good to force your opponent into a difficult decision, it must be bad to force yourself into one.
As poker players, our primary job is to make decisions that maximize our Expected Value at every point. Thus, we can define a "difficult decision" as one where we have two or more options (eg calling or folding) that have approximately equal EV, thus making it hard to choose which is best. Since folding always has EV=0, any difficult decision that includes folding as a candidate will have EV of approximately 0. We cannot make money in situations where we have EV=0, and I suppose this is why Sklansky advises to avoid them. However, I disagree with the notion that these situations should be avoided. We have to make our decisions with incomplete information, and when we initially bet, we do not know that we are going to be raised. When the raise comes, we must re-assess the situation, accounting for the fact that our opponent's range is now stronger than it was a moment ago, before we knew he would raise. Sure, it's unfortunate your opponent has a stronger hand than you realized, but that does not invalidate the decision to bet (except in hindsight).
To take an example from limit poker, if you bet and are raised and aren't sure how to respond, this implies that your hand fell solidly into your range of "betting hands." You should certainly not be checking with such hands just to avoid a difficult situation.
The situation described here in Concept 13 is a bit different. It may even be technically true. Sklansky and Miller use the term "usually," so it's hard to disprove unless I can show that it is never correct. In fact, S+M contrive a situation in the book in which the advice does indeed hold; my question is whether their example is representative of the "usual" situation. This involves making arguments about how our opponents are most likely to respond to different sized raises with particular hands, and so, really, it's just a matter of opinion. It is probably possible to find the game-theory answer (where we assume our opponent plays optimally). Even this will not give a definitive answer as to what is the best play against actual human opponents, but I might come back to this later if I feel like doing some more hard-core game theory.
Saturday, October 03, 2009
Propping the $500 Minimum Buy-In 5-10 NL
I'm back from my honeymoon, and I started my new job propping the $500 NL game at the Bike. When I left three weeks ago, the game was seldom being played, so I had the impression that I wouldn't really have much to do. To the contrary, I've played almost 30 hours already in four days. The game is pretty good - most players are too loose preflop, as usual, but other than that they play mostly okay. Of course, there are also some very good players and one or two terrible ones.
I'm a little unused to playing deep-stack no-limit. My job for the past year and a half had me playing mostly limit games. When I did play no-limit, it was always at a small buy-in for short periods of time, meaning I rarely had time to build up a deep stack to play with. So, for this week at least, I've been buying in for less than $1000 every day. There are some built-in strategic advantages to having a short stack, and I think it's necessary to be much better than your opponents in order to overcome that. Plus, I think having a short stack significantly reduces volatility. (I recently read an article in a magazine suggesting that having a short stack actually increases volatility! The claim was unsubstantiated, and I don't see how it could be right.)
I've lost three out of the four days so far, but the losses were minimal. Overall, I'm up about $600, plus my table hit a jackpot on Thursday and I won a $950 table share. (Yes, even the $500+ buy-in game has a jackpot now, to my chagrin. This has been the case for at least six months now.) I think once I get over my jet lag and become more comfortable with the game, I will have a big enough advantage on most of my opponents that I can start buying in for $1000, maybe more. On the other hand, I sometimes have issues with concentration, and that is likely to get worse as I get more comfortable with a game. I also think I may get sloppy about giving off tells if I get too comfortable. I'll have to keep an eye on this.
I hope to write my analysis of Concept 13 tomorrow. I basically disagree with it, but the analysis is complicated.
I'm a little unused to playing deep-stack no-limit. My job for the past year and a half had me playing mostly limit games. When I did play no-limit, it was always at a small buy-in for short periods of time, meaning I rarely had time to build up a deep stack to play with. So, for this week at least, I've been buying in for less than $1000 every day. There are some built-in strategic advantages to having a short stack, and I think it's necessary to be much better than your opponents in order to overcome that. Plus, I think having a short stack significantly reduces volatility. (I recently read an article in a magazine suggesting that having a short stack actually increases volatility! The claim was unsubstantiated, and I don't see how it could be right.)
I've lost three out of the four days so far, but the losses were minimal. Overall, I'm up about $600, plus my table hit a jackpot on Thursday and I won a $950 table share. (Yes, even the $500+ buy-in game has a jackpot now, to my chagrin. This has been the case for at least six months now.) I think once I get over my jet lag and become more comfortable with the game, I will have a big enough advantage on most of my opponents that I can start buying in for $1000, maybe more. On the other hand, I sometimes have issues with concentration, and that is likely to get worse as I get more comfortable with a game. I also think I may get sloppy about giving off tells if I get too comfortable. I'll have to keep an eye on this.
I hope to write my analysis of Concept 13 tomorrow. I basically disagree with it, but the analysis is complicated.
Friday, September 11, 2009
Analyzing NLHE:TAP Concept 12
This will be my last post until at least September 28. I'm heading to Paris for my honeymoon.
Concept No. 12: Be wary of overcallers.
Yes! Good idea. Overcallers are often on a draw, but they are also often slowplaying a big hand. This is especially true if there are no draws on the board, such as the JJ4 rainbow flop in the example in the book.
That's all I have to say for this one. See you in a few weeks!
Concept No. 12: Be wary of overcallers.
Yes! Good idea. Overcallers are often on a draw, but they are also often slowplaying a big hand. This is especially true if there are no draws on the board, such as the JJ4 rainbow flop in the example in the book.
That's all I have to say for this one. See you in a few weeks!
Monday, September 07, 2009
Analyzing NLHE:TAP Concept 11
I'm commenting on the sixty concepts at the end of No Limit Hold 'em: Theory and Practice by David Sklansky and Ed Miller. Let me apologize in advance: this one is a little more mathematical and a little less conclusive than most.
Concept No. 11: A big bet is the most relevant and accurate information at the table.
Once again, I think Sklansky and Miller's advice here is probably bad, but this one is particularly difficult to analyze because it's unclear what is meant by "relevant and accurate information." In any case, this concept's claim is far too broadly stated, and only holds in some circumstances.
If we take the claim in the most literal sense, this concept is not always true. Infinite counterexamples exist. Perhaps you know that your opponent always makes big bets; in such a case, the big bet provides essentially no information at all. Or, sometimes you may have the nuts, which is probably the most "relevant" factor if you're trying to figure out how to proceed with the hand. As far as "accurate" information goes, usually preflop play is most accurate, since you will have observed your opponent in many more preflop situations than postflop situations. Of course, all of this varies from player to player. It seems overly bold for S+M to make a claim that a big bet is (presumably, always) the most relevant and accurate.
If we're not to take this concept literally, let's consider whether it is at least good advice in a figurative sense. This concept seems suggest that players should respect the big bet more than they do, and thus fold to it more. Do players tend to underestimate the importance of a big bet when trying to determine their opponent's possible hand? Human nature actually suggests the exact opposite; there is a phenomenon called the "base rate fallacy" that seems like it could apply here. It has been established that people tend to discount base rates (which could correspond to your read of your opponent before the big bet), and focus only on the most immediate and salient information (such as the big bet) when estimating likelihoods. Are poker players similarly likely to focus to much on the most immediate information? I think sometimes yes, sometimes no. There are certainly a lot of players who are not sensitive enough to bet sizes, and for these players, taking this concept to heart is probably a good idea. I'm not sure that most players fall into this category, though.
In their explanation of this concept, Sklansky and Miller say "all information from the past takes a back seat to the fact that they've made a big bet." I think a fair way to interpret this is "the probability your opponent holds the nuts, or nearly the nuts, given that he just made a big bet, is greater than the likelihood that he did NOT hold the nuts before he made that big bet." Mathematically, we can write this as:
Suppose we are facing a big bet and we think our opponent is either bluffing or holding the nuts. Before he made the big bet, suppose we thought that the way he played the hand meant he had only a 0.05% chance of holding the nuts. This seems reasonable in S+M's example where your opponent is very tight, raised pre-flop, and the nuts is 74 (actually, .05% seems a bit too high, but let's give S+M the benefit of the doubt). Now let's say we think our opponent would make a big bet 80% of the time if he held 74. Otherwise, he would make the same big bet 2% of the time as a bluff. The base rate fallacy would be to think that your opponent would now be holding the nuts 80/82 times, or about 97.6% of the time. Of course, this is way too high because it ignores base rates. In this example, we can use Bayes' theorem to find the actual probability that you are up against the nuts. Bayes gives us the following equation:
P(big bet|nuts) = .8
P(nuts) = .0005
P(big bet|nuts') = .02
P(nuts') = .9999
and so:
P(big bet|nuts)*P(nuts) = .0004
P(big bet|nuts')*P(nuts') = .019998
Then,
If you don't like the rates I chose, I guess it's most likely because you think that even very tight players tend to hold 74 on the river more than .05% of the time in this situation (the situation: he raised preflop, and 74 is the nuts on the river). I admit, this is a tricky one to estimate, but we can do another Bayesian inference to examine my .05% claim. That is, we want to know if .05% is a reasonable approximation for how often a tight, pre-flop raiser holds 74 on the river in this situation. That is, we want to find P(74|saw river).
First of all, let's assume our "tight" player raises preflop with 74 only 1% of the time when dealt to him. Also, assuming we don't have a 7 or a 4 in our hand, our opponent will have been dealt 74 only 1.3% of the time. So, P(74) = .013*.01 = .00013.
Second of all, we need to approximate how often 74 will see the river in this situation. Remember, the situation is: 74 has already raised preflop and seen a flop that gives him a draw to the nuts. IN order to be conservative and to simplify the math, let's assume he will never fold before the river in this situation. So, P(saw river|74) = 1.
Third of all, we need to approximate how often a player would have seen the river in this situation given that he raised preflop holding anything. This is another tough one. I think this is likely over 50%, but let's be conservative and choose 26%, which will also help out with the arithmetic. So, assume P(saw river) = .26. Bayes theorem then gives us:
Concept No. 11: A big bet is the most relevant and accurate information at the table.
Once again, I think Sklansky and Miller's advice here is probably bad, but this one is particularly difficult to analyze because it's unclear what is meant by "relevant and accurate information." In any case, this concept's claim is far too broadly stated, and only holds in some circumstances.
If we take the claim in the most literal sense, this concept is not always true. Infinite counterexamples exist. Perhaps you know that your opponent always makes big bets; in such a case, the big bet provides essentially no information at all. Or, sometimes you may have the nuts, which is probably the most "relevant" factor if you're trying to figure out how to proceed with the hand. As far as "accurate" information goes, usually preflop play is most accurate, since you will have observed your opponent in many more preflop situations than postflop situations. Of course, all of this varies from player to player. It seems overly bold for S+M to make a claim that a big bet is (presumably, always) the most relevant and accurate.
If we're not to take this concept literally, let's consider whether it is at least good advice in a figurative sense. This concept seems suggest that players should respect the big bet more than they do, and thus fold to it more. Do players tend to underestimate the importance of a big bet when trying to determine their opponent's possible hand? Human nature actually suggests the exact opposite; there is a phenomenon called the "base rate fallacy" that seems like it could apply here. It has been established that people tend to discount base rates (which could correspond to your read of your opponent before the big bet), and focus only on the most immediate and salient information (such as the big bet) when estimating likelihoods. Are poker players similarly likely to focus to much on the most immediate information? I think sometimes yes, sometimes no. There are certainly a lot of players who are not sensitive enough to bet sizes, and for these players, taking this concept to heart is probably a good idea. I'm not sure that most players fall into this category, though.
In their explanation of this concept, Sklansky and Miller say "all information from the past takes a back seat to the fact that they've made a big bet." I think a fair way to interpret this is "the probability your opponent holds the nuts, or nearly the nuts, given that he just made a big bet, is greater than the likelihood that he did NOT hold the nuts before he made that big bet." Mathematically, we can write this as:
P(nuts|big bet) > p(nuts' before the big bet). That is: P(nuts|big bet) > p(nuts').I'm fairly confident that this is not always true, although it is might be true if your opponent is a very good player. I think that Sklansky and Miller might have succumbed to the base rate fallacy and are underestimating the importance of their initial read of their opponent's preflop tendencies. Let's do some Bayesian analysis of the example they give in their explanation of this concept.
(Note that nuts' means "not the nuts.")
Suppose we are facing a big bet and we think our opponent is either bluffing or holding the nuts. Before he made the big bet, suppose we thought that the way he played the hand meant he had only a 0.05% chance of holding the nuts. This seems reasonable in S+M's example where your opponent is very tight, raised pre-flop, and the nuts is 74 (actually, .05% seems a bit too high, but let's give S+M the benefit of the doubt). Now let's say we think our opponent would make a big bet 80% of the time if he held 74. Otherwise, he would make the same big bet 2% of the time as a bluff. The base rate fallacy would be to think that your opponent would now be holding the nuts 80/82 times, or about 97.6% of the time. Of course, this is way too high because it ignores base rates. In this example, we can use Bayes' theorem to find the actual probability that you are up against the nuts. Bayes gives us the following equation:
P(nuts|big bet) = P(big bet|nuts)*P(nuts) / (P(big bet|nuts)*P(nuts)From our assumptions above, we have:
+ P(big bet|nuts')*P(nuts'))
P(big bet|nuts) = .8
P(nuts) = .0005
P(big bet|nuts') = .02
P(nuts') = .9999
and so:
P(big bet|nuts)*P(nuts) = .0004
P(big bet|nuts')*P(nuts') = .019998
Then,
P(nuts|big bet) = .0004 / (.0004 + .019998) = .0004/.020398So, under these assumptions, our opponent holds the nuts less than 2% of the time when he makes a big bet! You may quibble with the rates I chose, but even if our calculations are off by a factor of 10, our opponent will still only be holding a winning hand 20% of the time, and bluffing 80%, making it correct to call any sized bet.
= .0196
If you don't like the rates I chose, I guess it's most likely because you think that even very tight players tend to hold 74 on the river more than .05% of the time in this situation (the situation: he raised preflop, and 74 is the nuts on the river). I admit, this is a tricky one to estimate, but we can do another Bayesian inference to examine my .05% claim. That is, we want to know if .05% is a reasonable approximation for how often a tight, pre-flop raiser holds 74 on the river in this situation. That is, we want to find P(74|saw river).
First of all, let's assume our "tight" player raises preflop with 74 only 1% of the time when dealt to him. Also, assuming we don't have a 7 or a 4 in our hand, our opponent will have been dealt 74 only 1.3% of the time. So, P(74) = .013*.01 = .00013.
Second of all, we need to approximate how often 74 will see the river in this situation. Remember, the situation is: 74 has already raised preflop and seen a flop that gives him a draw to the nuts. IN order to be conservative and to simplify the math, let's assume he will never fold before the river in this situation. So, P(saw river|74) = 1.
Third of all, we need to approximate how often a player would have seen the river in this situation given that he raised preflop holding anything. This is another tough one. I think this is likely over 50%, but let's be conservative and choose 26%, which will also help out with the arithmetic. So, assume P(saw river) = .26. Bayes theorem then gives us:
P(74|saw river) = P(saw river|74)*P(74) / P(saw river)Well, this is precisely the number I used above, and I was using conservative numbers in this example. Of course, you could still contend that my numbers or other assumptions are unreasonable, but I have at least convinced myself that in Sklansky and Miller's example, our opponent's big bet usually will not indicate that he has the nuts.
= 1*.00013/.26 = .0005 = .05%
Friday, September 04, 2009
Analyzing NLHE:TAP Concept 10
I'm working through all 60 concepts at the end of Sklansky and Miller's No Limit Hold 'em: Theory and Practice.
Concept No. 10: Sometimes you should go for a check-raise bluff on the river when a bluff bet would be unprofitable.
Sklansky and Miller are technically correct that you should sometimes check-raise as a bluff, but the advice is still bad. It's confusing for them to state this advice without noting the unusual circumstances under which it holds. They say, "sometimes you should..." Well, yes, but rarely. Readers who take Concept 10 at face value are in danger of trying check-raise bluffs indiscriminately, which would probably be a big mistake. My problem with this advice is actually very similar to Concept No. 4: "Sometimes you should bluff to stop a bluff." In both cases, Sklansky and Miller are missing the main point, an idea that I've attributed to Mike Caro (though I can't find it anywhere - maybe I thought of it myself!): If your opponent bluffs either way too much or way too little, be more willing to check to him. Or, more generally: If you've identified any weakness in your opponent's play (such as bluffing too much), try to put him in situations that will allow you to exploit this (for example, by checking and letting him bluff).
In Concept 4, Sklansky and Miller refer to an opponent who folds way too often to a small bet. You should bluff often against this opponent. Here in Concept 10, they use an example where your opponent bets 5/9 of the pot and is bluffing "the majority of the time." This is too often. (When betting 5/9 of the pot, game theory suggests you should be bluffing only 5/14 = 35.7% of the time.) Against an opponent who bluffs too much, checking is commonly the best play. This is the real lesson of the example they give: If you know your opponent bluffs too much, exploit it!
I know Sklansky and Miller are not talking about game theory here, but I just want to acknowledge that theoretically, it is indeed sometimes profitable to check-raise bluff even if your opponent is playing perfectly. For an optimal strategy, these bluffs are needed to balance the times you check-raise with a strong hand. In this theoretical case, the hands that are best to check-raise bluff with are actually not the very worst hands in your range (which are the best hands to try regular bluff bets with). Instead, the best hands to check-raise bluff with are the very best of the hands that you would otherwise fold. For example, if 66 is the worst hand you would call with in a given situation, 55 is actually the best hand to try a check-raise bluff with. In practice, if you want to try a check-raise bluff, the decision should depend a lot more on your read of your opponent than on your own hand, but I always think it's worth considering what game theory has to say about a situation.
Concept No. 10: Sometimes you should go for a check-raise bluff on the river when a bluff bet would be unprofitable.
Sklansky and Miller are technically correct that you should sometimes check-raise as a bluff, but the advice is still bad. It's confusing for them to state this advice without noting the unusual circumstances under which it holds. They say, "sometimes you should..." Well, yes, but rarely. Readers who take Concept 10 at face value are in danger of trying check-raise bluffs indiscriminately, which would probably be a big mistake. My problem with this advice is actually very similar to Concept No. 4: "Sometimes you should bluff to stop a bluff." In both cases, Sklansky and Miller are missing the main point, an idea that I've attributed to Mike Caro (though I can't find it anywhere - maybe I thought of it myself!): If your opponent bluffs either way too much or way too little, be more willing to check to him. Or, more generally: If you've identified any weakness in your opponent's play (such as bluffing too much), try to put him in situations that will allow you to exploit this (for example, by checking and letting him bluff).
In Concept 4, Sklansky and Miller refer to an opponent who folds way too often to a small bet. You should bluff often against this opponent. Here in Concept 10, they use an example where your opponent bets 5/9 of the pot and is bluffing "the majority of the time." This is too often. (When betting 5/9 of the pot, game theory suggests you should be bluffing only 5/14 = 35.7% of the time.) Against an opponent who bluffs too much, checking is commonly the best play. This is the real lesson of the example they give: If you know your opponent bluffs too much, exploit it!
I know Sklansky and Miller are not talking about game theory here, but I just want to acknowledge that theoretically, it is indeed sometimes profitable to check-raise bluff even if your opponent is playing perfectly. For an optimal strategy, these bluffs are needed to balance the times you check-raise with a strong hand. In this theoretical case, the hands that are best to check-raise bluff with are actually not the very worst hands in your range (which are the best hands to try regular bluff bets with). Instead, the best hands to check-raise bluff with are the very best of the hands that you would otherwise fold. For example, if 66 is the worst hand you would call with in a given situation, 55 is actually the best hand to try a check-raise bluff with. In practice, if you want to try a check-raise bluff, the decision should depend a lot more on your read of your opponent than on your own hand, but I always think it's worth considering what game theory has to say about a situation.
Tuesday, September 01, 2009
Analyzing NLHE:TAP Concepts 8-9
Discussing the concepts at the end of Sklansky and Miller's No Limit Hold 'em: Theory and Practice.
Concept No. 8: Other things being equal, when you're in one of the blinds your preflop raises should generally be a little larger than normal.
I often follow this advice in my own play, but I'm not convinced that it is generally correct. Maybe it is, but I'm not sure. This is rather difficult to analyze objectively, and I think it will be a subject of discussion among poker theorists and enthusiasts for a long time. Anyway, Sklansky and Malmuth list three reasons why they think this advice is correct. Let's look briefly at each.
1. S+M say:
You'll be out of position, and raising bigger increases the chances you'll win right away.
Me: Yes, but you could also argue that you'd rather play a small pot if you're out of position, and so you should either raise smaller or only with very strong hands like JJ+.
2. S+M say: Cuts down your opponents' implied odds.
Me: Well, yes, but it also costs you more. Is it worth it?
3. S+M say: In early position, it's more important that you narrow down your opponents' possible hands, which you can do by raising more.
Me: Good point. Again, I wonder: is it worth the extra investment?
Finding the right strategy for raising in the blinds is something I'm still working on. It probably deserves a lot more consideration than I'm giving it in this post, but I don't feel that I have anything insightful to say about it beyond what I'm writing here. My current strategy is this: In the blinds, I like to raise only with a small number of strong hands. The actual range will depend on who has limped into the pot so far, but sometimes I will check with a hand as strong as pocket tens (although I will usually raise this hand). Occasionally, I will bluff with weaker hands (sometimes even extremely weak hands such as 72o if I think my image is very tight and I will likely make everyone fold). Since I seldom bluff from the blinds, I don't usually feel the need to make my raises particularly large. If my opponents know me well, they should fold to a normal-sized raise of about 2 BB larger than the current pot size, which is about what I consider normal-sized from other positions, as well. However, I will often raise larger than this if my opponents don't know me or if they seem to play too loosely. Since this is often the case, I actually do raise larger than normal from the blinds pretty often.
A related question is how much to raise in early position versus late position. I've seen all three alternatives argued: larger raises in early position; larger raises in late position; keep raises about the same size. The argument that you should raise bigger from early position is the most popular: you don't want to get called and then have to play the pot out of position, which is the same as S+M's 1st point above. I've also seen it argued that you should raise the same amount from each position. One argument for this is that it will always give the blinds the same pot odds if nobody else joins the pot. I've rarely seen it argued that you should raise bigger from late position than from early position, but this is what Chen and Ankenman argue in The Mathematics of Poker. They say that since your raising range is much smaller from early position, it's not necessary to raise as much. Also, by raising small, you save money if one of your opponents reraises, which is much more likely if you are raising from early position with 7-9 players yet to act than if you raise from the button with just 2 players left after you. Personally, I've decided to take the middle road and raise about the same regardless of my position. For me, this philosophy extends to how I play from the blinds, where I raise about the same as from anywhere else.
I think this topic deserves a much more objective, EV-based analysis, but this seems like a big project. I'm not up to the challenge just yet. I hope to return to this question at some point after I'm done going through the rest of these concepts, but that will not be for quite a while.
Concept No. 9: Bets are usually more important than pots.
I'm tacking this one on the end here because I don't have much to say about it. The concept's claim is not literally true because even in no-limit holdem, bets are usually smaller than the pot, but S+M's point is well-taken: in contrast to limit poker, winning or saving extra bets is extremely important in no-limit poker.
No-limit players should be much more willing to give up a pot in order to save a bet. In limit poker, on the other hand, the pots tend to be big compared to the bets, and so calling is much more commonly the right decision.
For limit players, it can be difficult to adjust to this difference. I had the opposite problem when I started playing more limit holdem last year: I was folding way too often for a limit game. My experience with no-limit poker had gotten me used to the idea that saving bets was important, and giving up pots was okay. In limit, the exact opposite is true!
Concept No. 8: Other things being equal, when you're in one of the blinds your preflop raises should generally be a little larger than normal.
I often follow this advice in my own play, but I'm not convinced that it is generally correct. Maybe it is, but I'm not sure. This is rather difficult to analyze objectively, and I think it will be a subject of discussion among poker theorists and enthusiasts for a long time. Anyway, Sklansky and Malmuth list three reasons why they think this advice is correct. Let's look briefly at each.
1. S+M say:
You'll be out of position, and raising bigger increases the chances you'll win right away.
Me: Yes, but you could also argue that you'd rather play a small pot if you're out of position, and so you should either raise smaller or only with very strong hands like JJ+.
2. S+M say: Cuts down your opponents' implied odds.
Me: Well, yes, but it also costs you more. Is it worth it?
3. S+M say: In early position, it's more important that you narrow down your opponents' possible hands, which you can do by raising more.
Me: Good point. Again, I wonder: is it worth the extra investment?
Finding the right strategy for raising in the blinds is something I'm still working on. It probably deserves a lot more consideration than I'm giving it in this post, but I don't feel that I have anything insightful to say about it beyond what I'm writing here. My current strategy is this: In the blinds, I like to raise only with a small number of strong hands. The actual range will depend on who has limped into the pot so far, but sometimes I will check with a hand as strong as pocket tens (although I will usually raise this hand). Occasionally, I will bluff with weaker hands (sometimes even extremely weak hands such as 72o if I think my image is very tight and I will likely make everyone fold). Since I seldom bluff from the blinds, I don't usually feel the need to make my raises particularly large. If my opponents know me well, they should fold to a normal-sized raise of about 2 BB larger than the current pot size, which is about what I consider normal-sized from other positions, as well. However, I will often raise larger than this if my opponents don't know me or if they seem to play too loosely. Since this is often the case, I actually do raise larger than normal from the blinds pretty often.
A related question is how much to raise in early position versus late position. I've seen all three alternatives argued: larger raises in early position; larger raises in late position; keep raises about the same size. The argument that you should raise bigger from early position is the most popular: you don't want to get called and then have to play the pot out of position, which is the same as S+M's 1st point above. I've also seen it argued that you should raise the same amount from each position. One argument for this is that it will always give the blinds the same pot odds if nobody else joins the pot. I've rarely seen it argued that you should raise bigger from late position than from early position, but this is what Chen and Ankenman argue in The Mathematics of Poker. They say that since your raising range is much smaller from early position, it's not necessary to raise as much. Also, by raising small, you save money if one of your opponents reraises, which is much more likely if you are raising from early position with 7-9 players yet to act than if you raise from the button with just 2 players left after you. Personally, I've decided to take the middle road and raise about the same regardless of my position. For me, this philosophy extends to how I play from the blinds, where I raise about the same as from anywhere else.
I think this topic deserves a much more objective, EV-based analysis, but this seems like a big project. I'm not up to the challenge just yet. I hope to return to this question at some point after I'm done going through the rest of these concepts, but that will not be for quite a while.
Concept No. 9: Bets are usually more important than pots.
I'm tacking this one on the end here because I don't have much to say about it. The concept's claim is not literally true because even in no-limit holdem, bets are usually smaller than the pot, but S+M's point is well-taken: in contrast to limit poker, winning or saving extra bets is extremely important in no-limit poker.
No-limit players should be much more willing to give up a pot in order to save a bet. In limit poker, on the other hand, the pots tend to be big compared to the bets, and so calling is much more commonly the right decision.
For limit players, it can be difficult to adjust to this difference. I had the opposite problem when I started playing more limit holdem last year: I was folding way too often for a limit game. My experience with no-limit poker had gotten me used to the idea that saving bets was important, and giving up pots was okay. In limit, the exact opposite is true!
Sunday, August 30, 2009
Analyzing NLHE:TAP Concept 7
It's my weekend again, so it's time to get back to analyzing Sklansky and Miller's No Limit Hold 'em: Theory and Practice concepts. I'll try to do one or two during the week, as well.
Concept No. 7: Don't telegraph that you have one pair unless you can profitably call big bets.
Sklansky and Miller say, "One pair can be a very tricky hand to play in deep stack no limit." I agree with this statement, but I do not agree with Concept 7 as stated. It suggests that except under certain circumstances, you need to play your pairs abnormally in order to throw off your opponents. This is terrible advice. You want your "normal" game to maximize EV. Thus, any advice to deviate from this is, by definition, going to lose some expected value. Losing EV=bad.
Certainly, we do not generally want our opponents to know what we are holding. So, the advice that we not telegraph that we have one pair seems sound. The problem with this advice is the implication that by playing a hand in our normal way, we "telegraph" our hand to our opponent. If this were the case, our entire poker strategy would need a complete overhaul. If we are playing a well-rounded game, we will never be telegraphing our hands (or if we do, it's not in a way that our opponents can exploit). Even if our strategy is not so well-rounded, it takes an extremely perceptive opponent to discern our hand precisely.
Let's consider for a moment S+M's suggestion that it's okay to let your opponents know what you have under certain circumstances. Specifically, they say, "If you welcome a big bet with one pair because you expect that bet to be a bluff, then it's ok to telegraph your hand by playing in a way that makes it obvious what you hold."
I don't really know what to make of this. If it's okay to "telegraph" my hand, should I just reveal my hand to my opponent in such a case? Surely not, although by the logic included in this concept it would seem that S+M would expect that this would accomplish the desired goal of inducing a bluff. This is plausible, but it still doesn't seem like the best way to play the hand.
Obviously, S+M are not advising that we telegraph our hands by simply revealing them to our opponent. What they mean by "telegraphing" is playing the hand in a normal, straightforward manner. However, as I said before, your normal strategy should not be telegraphing your hand, anyway.
I think Sklansky and Miller's advice is designed to protect the reader from developing a strategy that is too predictable. It's true that a NL player needs to disguise his paired hands, but it should not be done by playing them abnormally. Instead, just make sure that you play some drawing hands in the same way, as well as some stronger hands. I think conventional Sklansky-wisdom is overly concerned with playing hands in a way that avoids difficult decisions. It's good to avoid difficult decisions, but not if it means playing abnormally and thus losing EV.
Concept No. 7: Don't telegraph that you have one pair unless you can profitably call big bets.
Sklansky and Miller say, "One pair can be a very tricky hand to play in deep stack no limit." I agree with this statement, but I do not agree with Concept 7 as stated. It suggests that except under certain circumstances, you need to play your pairs abnormally in order to throw off your opponents. This is terrible advice. You want your "normal" game to maximize EV. Thus, any advice to deviate from this is, by definition, going to lose some expected value. Losing EV=bad.
Certainly, we do not generally want our opponents to know what we are holding. So, the advice that we not telegraph that we have one pair seems sound. The problem with this advice is the implication that by playing a hand in our normal way, we "telegraph" our hand to our opponent. If this were the case, our entire poker strategy would need a complete overhaul. If we are playing a well-rounded game, we will never be telegraphing our hands (or if we do, it's not in a way that our opponents can exploit). Even if our strategy is not so well-rounded, it takes an extremely perceptive opponent to discern our hand precisely.
Let's consider for a moment S+M's suggestion that it's okay to let your opponents know what you have under certain circumstances. Specifically, they say, "If you welcome a big bet with one pair because you expect that bet to be a bluff, then it's ok to telegraph your hand by playing in a way that makes it obvious what you hold."
I don't really know what to make of this. If it's okay to "telegraph" my hand, should I just reveal my hand to my opponent in such a case? Surely not, although by the logic included in this concept it would seem that S+M would expect that this would accomplish the desired goal of inducing a bluff. This is plausible, but it still doesn't seem like the best way to play the hand.
Obviously, S+M are not advising that we telegraph our hands by simply revealing them to our opponent. What they mean by "telegraphing" is playing the hand in a normal, straightforward manner. However, as I said before, your normal strategy should not be telegraphing your hand, anyway.
I think Sklansky and Miller's advice is designed to protect the reader from developing a strategy that is too predictable. It's true that a NL player needs to disguise his paired hands, but it should not be done by playing them abnormally. Instead, just make sure that you play some drawing hands in the same way, as well as some stronger hands. I think conventional Sklansky-wisdom is overly concerned with playing hands in a way that avoids difficult decisions. It's good to avoid difficult decisions, but not if it means playing abnormally and thus losing EV.
Subscribe to:
Posts (Atom)