An Unbelievably Long Guide to Hand-Reading.
#8629256 - 01/02/07 12:57 PM
So you've got bottom set on a three-flush board and your opponent just check-raised you all-in. What do you do? Well, the first thing you do is you put your opponent on a hand.
Hand reading is simultaneously the most important and the most difficult thing that a poker player does. Some of the best poker players in the world ignore many (or all!) of those other "rules" in poker, have deadly-accurate hand reading skills, and make mountains of cash for their troubles. Other players are exquisitely good at all the fundamentals but stink at hand reading and therefore struggle to consistently beat $50NL.
Unfortunately, there is no "magic formula" for hand reading; after all, your opponents work very hard to AVOID letting you know the cards in their hands. However, while there is a great deal of artistry in hand reading, the basics are describable. After some thought I came up with these Five Commandments of Hand Reading (I had ten, but one of the tablets broke when EMC dropped it -- sorry, blame the mod) to get you started.
1. Know thy numbers.
So you know that your opponent who just limped has a VPIP of 30%; what does 30% really MEAN? We know it's loose, but are you really aware of what "30% of all hands dealt" actually looks like? Here are some probabilities of getting a hand in a particular preflop range:
Super-premiums: AA, KK, QQ, AK. Total probability: 2.6%.
Premiums: AA-TT, AK, AQ, KQ. Total probability: 5.9%.
Any pocket pair: AA-22. Total probability: 5.9%.
Any two broadway: Two cards, both T+, including pairs. Total probability: 14.3%.
Any suited ace: A2s-AKs. Total probability: 3.6%.
Unsuited ace: A2o-AKo. Total probability: 10.9%.
"Maximum suited connectors": JTs-54s. Total probability: 2.1%.
Any ace: A2o+, A2s+, AA. Total probability: 14.9%.
Any two suited: literally. Total probability: 23.5%.
Any two connectors: 32s-AKs, 32o-AKo. Total probability: 14.5%.
To give you a VERY broad feel for what different percentages translate into, here are some potential example ranges:
5% = "pairs 77+, AK, AQs" or "pairs 99+, AK, AQ."
10% = "pairs 66+, AK, AQ, suited aces, KQs, QJs"
15% = "any pair, AK, AQ, KQ, suited connectors 54+, any suited ace"
20% = "any pair, any two broadway, any suited ace"
25% = "any pair, any suited broadway, any ace, any suited connectors 54s+, KQo"
30% = "any pair, any ace, any suited king, any suited broadway, any suited connectors 54s+, KTo+, QJo" or "any pair, any ace, any suited king, any broadway"
40% = "any pair, any ace, any king, any two broadway, any suited connectors 32s+"
50% = "any pair, any two suited cards, any ace, any two broadway, K5o+"
Again, remember to use the relevant range: a player who is 65/10 is looser than the 50% range when he limps but has a rather tight range if he actually raises. Also realize that some players who are loose and passive will raise with SECOND-best hands like 88-JJ, AJ, KQ, and 54s-JTs, but NOT the ultra-premiums like QQ+, AK, and AQ -- those they will slowplay to disguise their hands. Watch these folks at showdown to try and figure out how they play their really big hands.
Postflop aggression numbers also reveal much about a player. When a player with an average aggression of 8 raises preflop and then checks to you, suspect a trap: this is very often a strong hand that's going for a check-raise. When you get that same preflop bet and flop check from a player whose average aggression is 1.3, that's more likely to be a player who missed and is giving up. When that same play comes from someone with a postflop aggression of 0.4, you gain no information from the check -- checking is just what this player does. Mentally split players into three groups -- high aggression, medium aggression, and low aggression -- and then adjust accordingly. Be VERY afraid when a player is playing out of character: the passive postflop player who leads and raises almost always has a monster, and you can fold without a worry. The ultramaniac who check/calls two streets also has a monster, and is waiting to beat your brains out. Don't fall for it. All of this leads us to our second commandment:
2. Know thy enemy.
A leopard never changes its spots, especially at uNL stakes. Together, the HUD trinity (VPIP, PFR, and average aggression) tell us much about a player. Loose players play loose; tight players play tight. Aggressive players play aggressively and passive players play passively. Categorize your opponents on three separate measures:
Preflop looseness: a loose preflop player has a VPIP over 40% (I made up the number, but you get the general idea). These guys have crap preflop, and any hand you're willing to play is beating their range. Beware of these folks postflop, however -- there's no flop that definitely missed your enemy. Be prepared to play with caution when you don't have a monster. That's not to say you should be check-calling; rather, expect to be ahead and bet consistently with your made hands, but keep the bets small and try for pot control. Alternatively, raise light and raise strong preflop while your hand dominates your opponent's range. His mistake is playing too many hands -- exploit this mistake by hammering him preflop while you're way ahead.
On the other hand, a tight preflop player has a VPIP under 20%. These folks won't enter into a hand unless they've got something worth pursuing. They aren't really making a major mistake preflop, and the only way you can really take advantage of this characteristic is by stealing their blinds remorselessly (though you'll have to instafold if they catch you stealing and you don't have an honest hand).
Preflop aggresssion: a passive preflop player has a PFR under a quarter of their VPIP. That means that this is a floating scale: while 10% PFR is passive for a player who has a VPIP of 55%, it is aggressive for a player who has a VPIP of 15%. Alternatively, an aggressive preflop player will have a PFR over half of their VPIP. When we're trying to decide a player's preflop holding, we should use their VPIP and their PFR to come to a conclusion. Say a player has a VPIP of 40% and a PFR of 20% and they limp in front of you. What sort of holdings do you expect them to have? Well, we know the player is willing to play with 40% of his hands; our sample range for this looked something like "any pair, any ace, any king, any two broadway, any suited connectors 32s+." But we can already refine this range some more: we know that with 20% of those hands, villain would have raised, and here he didn't. If we assume that villain raises with his top 20%, our sample range for that looks something like "any pair, any two broadway, any suited ace." So take THOSE hands out of his current range: just like you discount 72o when a nit is in the hand, you discount QQ when a maniac limps. A likely range here is going to be the difference between the two ranges, or "any unsuited ace, K9s or worse, K9o or worse, T9s or worse." That's a much easier range to play against. Be sure to watch this opponent's showdown hands, however -- you'd really like to know if your assumption about him raising the top 20% of his range is correct or not. If you have a tricky opponent who raises his middle 20% and limps the top 10% and bottom 10%, your range will be significantly wrong, and you'll be in a good deal of trouble with your hand reading. One huge word of warning: people often look at a player's VPIP and conclude that the player is a loose idiot. Then they call his raise and are shocked when they find out at showdown that he had rockets. Remember, if you are raised preflop by a 65/5, he has roughly the same hand range as when you're raised by a 12/5. Don't mistake his typical preflop looseness with a wide range when he raises....
Postflop aggression: after the flop, a new game begins. People often make the mistake of assuming that a tight preflop player is tight postflop, or that an aggressive preflop player is aggressive postflop. This can only be discovered by observation. I will tell you that every combination of preflop and postflop playstyles is possible, and none are terribly uncommon. The true TA/TA (tight-aggressive preflop, tight-aggressive postflop) and LP/LP (calling station pre and post) is only one form of opponent. Another very common player is the TA/LA: always aggressive and solidly tight preflop, this opponent plays hands so infrequently that when he finds something -- ANYTHING -- that he's willing to play preflop he cannot bring himself to let it go. He turns into a maniac postflop, relying on folding equity and a better starting hand range to win money. Another common player is the TP/TA, who goes from nit preflop to aggro-monkey postflop. If they miss, they're out of the hand, but if they hit the hand they're going to hammer every street and try to get all-in by the showdown. Less common at uNL, but increasingly common at higher levels (and deadly-dangerous when they do it right) is the LP/TA: this player is a total calling station idiot preflop. You'll often see preflop numbers of 75/11 or 68/6 for these folks. Don't fall for it! It's a ruse. These folks are splashing around in as many pots as they can as cheaply as they can, and then postflop they play POKER. They'll be folding a tremendous fraction of the time on the flop, escaping for 1 BB, but when they hit it will be with something sneaky-as-hell and solid-solid-solid. They then go into aggro mode, betting incredibly hard and relying on unobservant TAGs to call them down lighter because "hey, this idiot is 72/7, my TP3K dominates his range." They stack more TAGs than you and I could ever hope to, because their image gets them mad phat postflop action. So the rule is: postflop is a new game; expect people to play a different style postflop from preflop, and try to quickly figure out BOTH of these styles.
Now, once we're on the flop and beyond we need to use average aggression and postflop tightness to decide what a player's holdings are, refining our original range based on their preflop play. This leads us to the third commandment:
3. Know thy board.
Flops have different "textures," and those textures can be much more or much less scary, depending on your holdings and your opponent's range. More importantly, different people respond in different ways to different board textures. On a draw-heavy board, if a loose and aggressive player is check-calling you can expect him to have the near-nuts, but if a loose and passive player is check-calling you can expect him to have ... well, any damn thing. What affects the texture of a board? Well, let's start with the flop.
Suitedness: flops can come "rainbow" (three different suits), "two-suited" (two of one suit and one of another), or "monochrome" (all three cards in the same suit). The more "suited" a flop is, the bigger the hand most opponents will need to call. However, note that many hyperaggressive opponents will be more likely to bet out, check-raise, or "float" (smooth-call on the flop with the intent of taking the pot on the turn) with either a pure bluff or a semi-bluff (draw) on these types of boards. If you are first to act, you can often steal these pots for a reasonably small (2/3rds-pot) bet; if you are called, beware of the flush draw! One small bit of math: let's say that the flop comes with three spades and you have none in your hand. The odds that your single opponent flopped a made flush are 3.3% and the odds that he flopped a flush DRAW are 15.8%. If you've hit a solid hand (say, TPTK) DO NOT PANIC AND START CALLING! Bet out and protect against the draw that is 4.75 times more likely than the made flush that has you killed. Besides, if your opponent actually has a second-best hand, he'll be more likely to pay off a bet on a monochrome flop than he will to pay off a bet when the turn has FOUR spades (assuming he doesn't have one). Bet while your hand is best and charge him well to try and outdraw you. Incidentally, if your board has three spades and you have one in your hand, the odds that your opponent has two spades drops to 2.6% and the odds that he has one spade drops to 14.4%, so the odds that your opponent is drawing to a flush are now 5.6 times higher than the odds that he flopped the monster. Bet and protect!
Connectedness: here we're talking about how many cards to a straight the board has. A monochrome flop of J-T-9 is MUCH more dangerous than a monochrome flop of J-7-2. Always be aware of straight draws -- they're a gold-mine to the savvy poker player because so many people miss them. When the flop comes A-K-Q, the player holding JTs just stacked the preflop raiser holding AK. When the board is connected, you need to beware of two separate possibilities: your opponent might have two pair and your opponent might have an open-ended straight draw. Often, two pair is the scarier event, because your weak-but-made hand is often drawing very thin against it. A straight draw can become an almost unbeatable monster, but it has to GET there first. Two pair is already there. At small stakes games, many players will play very passively with a draw, check/calling in the hopes of improving, but they will play aggressively with two pair. Your more aggressive opponents will bet BOTH hands strongly. When someone plays back at you on a fairly connected board, you need to decide if they're likely drawing or if instead they've flopped some powerhouse hand. Then you'll proceed based on the strength of your hand in relation to the range you think is likely for your opponent. Much like a suited board, a connected board can often be used as a powerful bluffing or semibluffing tool. Say your 30/11/3 opponent raises preflop from MP and you call in position with 33. Heads-up, the flop comes 7-6-5. This is a REALLY good flop for attack aggressively: considering your opponent's stats, the raise makes overcards much more likely than usual, so the odds that this flop has completely missed your opponent are higher than usual. A flop raise or a flop float can prove extremely valuable to you. In this analysis I'm completely ignoring your inside straight draw -- that's virtually worthless since it's highly unlikely to happen and also highly unlikely to get paid in any significant way by your opponent while still being the best hand. No, I'm saying that this flop is a good one for you because it is unlikely to have improved your opponent in any meaningful way. Pressing back hard should win you this pot quite often WITHOUT getting to a showdown.
High Card Value: your opponents love playing high cards. Sure, you've outgrown calling raises with KJo and A9o (you HAVE outgrown that, right?) but they haven't. Playing these easily-dominated hands will prove very expensive for your inattentive opponents, but realize this general rule: a flop that is high-card heavy is much more likely to have connected with your opposition than a flop that is high-card light. If an ace hits the board in a multi-way pot and I don't have AT LEAST AQ, I'm usually done with the hand. Opponents love nothing more than playing aces, and when those aces hit the board your opponents will hang on to their aces like they were made out of solid gold. Worse yet, a pair of aces with a J-or-worse kicker is going to be in oodles of trouble unless that kicker connects, too. Think about this:
say you have AJ on an ace-high board. The next highest board card is a T. If one other player has an ace, what are the odds that his hand beats yours? Well, AK and AQ obviously have you outkicked, and the unlikely AA has you decimated.
However, there are four OTHER aces that beat you -- the ones that have made two pair. That means you're behind about as often as you're ahead in this situation, and that's even assuming that your opponent "only" has an ace! You throw in the other random two-pair and set hands and your hand will win at showdown less than half the time. Worse yet, most opponents will get the message and fold their aces with weak kickers, but they're unlikely to fold any hand that beats you. The odds are that if you somehow create a big pot, you're even MORE likely to be behind. In short, proceed with great caution on ace-high boards, even if you have an ace. King high boards are pretty dangerous, too, because the looser opponents will play many kings, especially suited ones. Q's and J's are less scary as a player's high card, but VERY dangerous as a player's LOW card. Someone willing to play KJo preflop is virtually never going to fold that hand on a J-high flop.
Recognize that the odds that your opponent has missed the board are highest on low-card boards, and much lower when the board has high cards. This is especially true if the board has more than ONE high card. One major exception to this rule: if you RAISED preflop, don't give up when the flop comes with a high card, especially if that high card is an ace.
This is a fantastic chance to steal the pot. Statistically speaking, virtually any opponent you could face has a less-than-50% chance of having an ace in this situation, but if you bet the flop they will assume you DO have one. A standard continuation bet will win the pot a surprisingly large fraction of the time. If they play back, fold and move on to the next hand.
Paired Boards: usually, a paired board is a cause for celebration. Why? Because with an unpaired board there are nine separate cards in the unseen deck that could give an opponent a pair. However, with a PAIRED board, that number falls to only FIVE cards. In other words, it's now almost 50% less likely that an opponent has made a hand good enough to want to continue. You should use this against them if it is reasonable for you to do so. Mind you, if you limped preflop and the board is AAK, you can usually check-fold, because your opponent is not going to believe that you have the goods.
However, if you raised preflop and the board comes 884 a bet in a heads-up pot is virtually MANDATORY: your opponent will realize he's missed, assume you have a pocket pair, and fold even more often than he would fold to a typical continuation bet. Paired boards are perfect for continuing preflop aggression. Also, realize that most aggressive players know this, so if you happen to be in a pot that someone else raised, the flop comes paired and you've got a sneaky monster, consider a slowplay like a flop check-raise or even a "check/call flop, check/raise turn." Your aggressive targets will fire off a continuation bet quite often, and you can then "snap off a bluff" and win a bigger pot than you otherwise would. Obviously, this will be opponent-specific, but keep your eyes peeled for such opportunities.
On the turn and river, similar issues with connectedness, suitedness, high card value, and board pairs will continue to pertain, and will define the "texture" of the board. As a general rule, a tight opponent will continue on "wet" (highly coordinated) boards when he has a strong hand or a strong draw, but a loose opponent may continue with as little as top pair. An aggressive opponent can bet "wet" boards with a draw, a "combination draw" (straight and flush), or a pair+draw, and may even bet these boards on a pure bluff. A passive opponent betting into a "wet" board usually has the goods -- these folks rarely bet their draws. Now, to start to put this all together, let's move on to the next rule:
4. Know thy hand history.
Here we're discussing how this particular hand has played out: who bet when and how much? Start looking for betting patterns, as different people will have different patterns. A few general issues:
Check-raises: when an opponent check-raises, he is sending the message that his hand is unusually powerful. He knows you are betting and he doesn't care. What's more, he was confident enough to risk your checking behind in his quest to get more money in the pot. These types of bets will usually mean one of three things: either your opponent was monsterously strong and slowplayed an earlier street, or the last card just helped your opponent in some way, or he is bluffing in a situation where he thinks he can scare you off your hand. As a general rule, trust check-raises from passive players. Completely. If you don't have a sneaky MONSTER (and I mean MONSTER with a capital *everything*) you should be folding to this raise. People often ask "can I ever escape from pocket aces?" This would be one situation where escaping would be easy. Another general rule is that the more aggressive a player is, the more likely a check-raise is a bluff. I would say that until an opponent has an aggression factor of at least 2 you shouldn't worry much about a check-raise semi-bluff, and until he has an aggression factor of at least 4 you shouldn't worry much about a check-raise bluff. People are very quick to put a player on a bluff when he check-raises; I believe this event is much rarer than most people think. One caveat to this: check-raises on the flop are far more likely to be bluffs or weak hands than check-raises on ANY other street. On the flop, people will often take a check-raise line against a frequent c-bettor, even with hands like "bottom pair, no kicker" because they know that their opponent will frequently have pure air. So: a check-raise on the flop usually means "I can beat ace-high," but a check-raise on a later street usually means "I can beat YOU."
Check-calls: this play is highly player-specific. Against a passive calling station this means "I have two cards. Look! Spades are pretty. I like pie." Against a tight and moderately aggressive player this often means "I'm on a draw." Against a highly aggressive player this often means "I have a monster and I'm going to let you bet yourself to death." Check-calls are precursors to check-raises on later streets from very aggressive players; from very passive players, they just precede more check-calls.
Donkbets: a "donkbet" is when someone who does NOT have the betting lead makes an unexpected bet. For instance, if a player called your preflop raise but then leads into you on the flop, that would be a donkbet. Similarly, if a player calls your flop bet but then leads on the turn, that would also be a donkbet. At these stages, a donkbet should be interpreted as saying "that card helped me." The more passive your opponent, the more straightforward this interpretation is. When a passive player comes alive on a third straight card, or a fourth flush card, or a pairing of the board, or some odd-looking random card, you should expect that the card has helped his hand out and he is now value-betting. Of course, how strong his hand is remains to be seen, but the card has improved him. Don't expect that this means he has the nuts: I've seen passive players wake up and bet the fourth heart...because it gave them two pair. On the other hand, very aggressive players love to donkbet on scare cards. This is a cheaper bluff than a check-raise but it works just as often at these levels, and many aggressive players will take advantage of a turn ace or a third/fourth club, or a board pair to try and steal the pot. Be aware of this.
Unexpected checks: a player who has been betting steadily in the hand suddenly starts to check. What does this mean? Well, one obvious interpretation is that he has been bluffing and has now given up on the hand. Most players are straightforward enough that this will be the case. Against a medium-aggressive opponent, this will often be a good opportunity to bet with any two cards, since your folding equity will be through-the-roof. Another common situation is that a player flops a decent hand -- say top pair on a T-high board -- and then slows way down when the turn brings a K. They are worried about the overpair, and so have stopped betting. That does not necessarily mean they are ready to fold; some players will go from betting/raising to check/calling all the way to showdown. However, when an aggressive player stops being aggressive, that's usually a sign that your situation is not as dire as you thought it was. Beware of the hyperaggressive opponent who unexpectedly checks, ESPECIALLY if he check/calls a scare card. For some reason, these players have taken the "strong = weak, weak = strong" philosophy to heart, and will often bet with little or nothing but immediately slowplay the moment they get a hand. Watch your opponents carefully to see if the turn check usually means they are giving up or if it usually means they are trapping you.
Bet-check-bet: a strange-looking but rather common three-street line is "bet the flop, check behind on the turn, bet the river," where the opponent has check/called the whole way. Against an aggressive opponent, this river bet is often what we call a "desperation bluff." The hand has no showdown value so the villain bets in the hopes that you will fold the best hand. Since you've shown little or no strength the whole hand through, they feel they have strong folding equity (which is true), and they are now attacking in a last-ditch effort to scoop up the pot. However, you need to be aware of what the board looked like. Another common reasoning behind this line is that the flop bet was with nothing, the turn gave the opponent a draw, and the river either made the draw or missed. If an obvious draw arrived on the end, you would really need to know more about your opponent before you knew if this was a bluff or not. Calling against some opponents will be extremely +EV even with bottom pair; against other opponents it will be -EV with anything short of a powerhouse. Again, watch your opponents and take notes on what their lines mean.
Our fifth and final commandment ties in to everything we've already discussed:
5. Know thy image.
"Image" is how the other players at the table perceive YOU. Against some opponents, this will be the very most important commandment. Against others, it will not matter at all. It all depends on how attentive your particular opponent is. How do we start to figure out our image?
Only worry about image with second-level thinkers. For some of you, this is your first segue into third-level thinking. First-level thinking is asking, "what is my hand?" Second-level thinking is asking, "what is my opponent's hand?" Third-level thinking is asking, "what does my opponent think my hand is?" Obviously third-level thinking is irrelevant against a first-level thinker. However, most opponents will at least make some token effort to guess your hand, so against your better opponents understanding your image will be important. When you have a complete idiot who never looks past his own hand while playing the game, don't worry about image -- you'll just be wasting your time and effort.
Your cards only affect your image when you SHOW them. In the last five straight hands you've had AA, KQ (flopped trips), QQ, JT (flopped straight), and 55 (flopped set). You won all five hands before showdown, and you never show your hands without being forced to do so. Recognize that your table image is now absolute CRAP. Yes, you had the goods. Sure, your hands were actually powerhouses. Of course, your starting hand selection is tight. None of that matters. All your opponents have seen is you betting and raising every hand. They doubt you, and they're very quickly going to get sick of your crap and start looking you up. This is NOT the time to get cute with QJs or 33 -- this is the time to play squeaky-clean poker as tight as you know how. Alternatively, if each of those five went to showdown and you displayed to the table your powerhouse winners, your folding equity will be HIGHER than usual, because people will begin to believe that you don't get involved with a hand unless you've got the goods. Be aware of this.
Recent history matters more than ancient history. Very few of your opponents actually have Poker Tracker. As a result, their image of you will be determined by their own personal observations. Most people have a relatively short memory, so concentrate most on your actions in the last two orbits. If your table image has fallen apart and your bluffs have gotten picked off several times in a row, tighten up and fold for the next two orbits -- in that short period of time you will rebuild most or all of your table image and you can then go back to doing your nasty deeds. Conversely, if you've been at the table for three hours playing a 12/8 game but you've gotten involved in 10 of the last 12 hands, people will think of you as a maniac, and play back at you with all sorts of crap. It's the recent history that counts, so remember how your last two orbits looked at the table.
Personal history matters more than table history. A player may not remember that you bluffed Seat 3 off his hand five times in a row, but he WILL remember that you bluffed HIM off his hand once an hour and a half ago. People have much longer memories for hands that involved them personally. If you stole a player's big blind three times in the last four orbits, he's going to know that, and he's going to play back at you with any reasonably strong hand. Your folding equity on a steal will be particularly low against HIM, but won't be low at all against the other players at the table.
Opponents will usually assume that your lines always mean the same thing. If you bet 3/4ths pot on a river scare card with the nut flush, then the next time you are in a pot against that opponent and the river comes with a three-flush (not yours), bet 3/4ths pot! Your folding equity will be tremendous. Alternatively, if the flush card DID help you, bet a DIFFERENT amount -- push all your chips in, or bet 1/2 pot, or do something else. When you want a call, don't do what you did the last time he saw you with the nuts. When you want a fold, do exactly what you did the last time he saw you with the nuts. People will remember these things.
----------
I realize this post is longer than even MY usual posts, but hand reading is a tricky, subtle, complicated, and ever-changing concept. Hopefully, this material will help you do a better job of reading hands at the table. As long and detailed as this is, know that I am just giving you the tip of the iceberg, here. Experience and study
(holdem, stud, draw poker) .a jako niepokerową grę gin rummy (w nia pierwszy duze pieniądze przed texas holdem zarabiał Stu Ungar)