|
|
#1
11-15-2007, 07:34 AM
|
|||
|
|||
Things it took me a while to learn part 9, Hand Reading
Continuing our comparison of tournaments to cash, tournament players have an easier job in front of them when it comes to hand reading. This is because the shorter stacks in tournaments (except in the first couple of levels) begin restricting the amount of hands players can profitably play, and force an enormous importance on position. However, it’s a safe bet that cash players will get better at hand reading much quicker, simply because they don’t have so much of the work done for them by a tournaments structure. If you want an easy answer at how to get good at post flop hand reading, play a ton of 6 max cash at a comfortable amount of tables where you can sit back and think things through. Save that, I’ll try to sum up what we can here on a sort of complicated subject.
Every time I hear the phrase “Well, I put him on XX” I want to punch the guy in the throat. That singular phrase is the excuse for more bad poker than anything else in the game. So many live pro’s (yes, I know, half these articles seem like a collection of cheap shots at the way live pros think) seem to have this obsession with pinpointing EXACTLY what their opponent has, so when they do manage to catch him with it they lean back, put a smug smile on their face, and announce “HA! That’s what I put you on!” I’m pretty sure Daniel Negreanu spends half an hour a day in the mirror practicing this with varying facial expressions in order to maximize its effectiveness when his moment of glory comes. Here’s the phrase to remember: You don’t your opponent on a hand, you put him on a RANGE of hands. Putting a player on a singular hand is pointless. The process by which you do hand reading is by starting with a range of hands your opponent has, then eliminating much of that range as the hand progresses and you’re given more and more information. Also important to this process of course, is having a general idea of how your opponent plays. If he’s brand new to the table online the best chance you have is to look him up on OPR. If he’s brand new to the table in a live tournament, you need to make some broad generalizations about him based on his appearance. So how do we go about the process of eliminating hands from our opponents range? You have more information to you than you think. Let’s start with an extremely obvious example: 50 FO on any given site. Blinds are 15/30 and everyone still has their starting stack of 3000. The players in the hand seem straight forward and basic. Preflop: UTG opens to 90, UTG+1 reraises to 270. Alright, now if we’re assuming both players are fairly straight forward here’s what we can deduct: UTG probably has a range like AJ+, ATs+, 77+ and KQs. This range isn’t precise as he could have smaller pairs, or perhaps wouldn’t raise ATs here, but it’s likely close. When UTG+1 reraises we can really eliminate a ton from his range. For the most part a straight forwards wide range here will be something like JJ+/AK, but giving them a range of QQ+ is by no means unreasonable in many cases. So in just two actions we’ve been able to isolate one players range down to about 3-5 hands. As I stated previously, this is an extremely obvious and simplistic example, but it demonstrates how with very little action we can get a pretty precise read. Most pre flop hand reading in tournaments comes down to watching peoples raise tendencies, and watching whether they understand stack sizes and position. Here’s an explanation of how these factor in: Stack sizes: Some players know or can feel that at 13-20 BB’s they shouldn’t be opening in a lot of spots where they can’t stand a reraise. However, players who are clueless to this will therefore have a much wider range with these stacks. Also, some players will call in the blinds with stacks that are either push/fold not knowing it’s incorrect to do so. You also need to identify which players flat call with hands even when their stack is much to short, and which players only flat call monsters on short stacks. One more key stack size understanding to watch for is resteals, plenty of players will try to resteal on a stack that’s very inappropriate to do so, and occasionally you’ll have to jam lighter back. I once had a player who was constantly trying who was constantly 3 betting me despite our 25-30 BB effective stacks, and as a result I 4 bet jammed QJo with 28 BB’s after he’d made it 8.5 BB’s and he folded. Situations like these come along where watching a players understanding of a concept takes a play from being incredibly awful (which 4 betting QJo all in with 28 BB’s is 99% of the time) to something that can be profitable. Position: The biggest pre flop situation that widens peoples ranges in regards to position is blind play. Lots of players call much looser in the than they should (especially the BB of course) due to the discount. Few factor in that this will force them to play OOP for the rest of the hand and put them at a huge advantage. I’ve already written about what a normal players range should be like out of the blinds, pay attention to guys who call much wider, as well as guys who play very tight from the blinds. Another thing to watch is for guys who call reraises OOP much looser than normal. For many good players at the mid to late stages of a tournament, they are basically never calling your reraise OOP without a monster unless stacks are quite deep. Alright, let’s get into some more examples of hand reading and go into the thought process. First I want to start with an example of what happens when you really know a persons play. I spend 4-8 hours a day on Skype with Luckychewy discussing tournaments and poker hands. At this point I can follow the logic behind his plays with enormous accuracy. For example I was watching him deep in the stars $300 tonight on AIM and witnessed the following hand: Poker Stars No Limit Holdem Tournament Blinds: t2000/t4000 (Ante: t200) 7 players Converter Stack sizes: UTG: t57888 UTG+1: t60291 MP1: t148859 Hero: t140802 Button: t218154 SB: t65545 BB: t66932 Pre-flop: (7 players) Hero is CO with Q[img]/images/graemlins/diamond.gif[/img] Q[img]/images/graemlins/spade.gif[/img] 3 folds, <font color="#cc0000">Hero raises to t10000</font>, 2 folds, BB calls t6000 <font color="aaaaaa">(pot was t17400)</font>. Flop: 8[img]/images/graemlins/spade.gif[/img] A[img]/images/graemlins/heart.gif[/img] 4[img]/images/graemlins/diamond.gif[/img] (t23400, 2 players) BB checks, Hero checks. Turn: K[img]/images/graemlins/heart.gif[/img] (t23400, 2 players) BB checks, Hero checks. River: 7[img]/images/graemlins/club.gif[/img] (t23400, 2 players) <font color="#cc0000">BB bets t8000</font>, Hero folds. Uncalled bets: t8000 returned to BB. After witnessing this hand I immediately IM’d Chewy with “QQ?” “Yep” he responded. “QQ.” Now, why does that seem so clear to me? Because I understand Chewys thought process. Let’s go street by street. Preflop: Obviously I know Chewy can have a pretty wide range here, although given what stacks are in the SB and BB I feel like he should have a fairly tighter range than normal since he’s expecting to be shoved on with a higher frequency. Flop: I expect Chewy to be betting a great C bet board like this with 100% of his range that has no showdown value. It’s not totally unreasonable to think Chewy would check back with an A here, since the board is rainbow without any real straight draws and the player will often give him credit for an A. I also expect him to check back hands like KK/QQ/ and sometimes JJ but I know he’ll often bet TT down more turn cards ruin his hand and it’s fairly easy to play with a bet. Turn: When the K hits that now means KK is much more unlikely. When the player checks again I expect Chewy to always bet with an A, since he needs to start getting value and the K will often hit the BB, meaning he’ll probably call at least one street of value. It also puts a flush draw out, even more reason to bet. When Chewy checks behind again, it’s at this point I know for sure he must have a hand like QQ, since betting there would basically be a bluff and his hand still has a considerable amount of SD value. I also know that he knows with this line he probably can’t get a fold from a K just betting one street, since villain will expect Chewy to have bet an A on the flop. River: When the guy bets 8000 Chewy thought it over and folded. He realizes there are very few hands in the villain’s range he beats and that if the villain is going to bluff the bet will normally come on the turn. He also realizes that betting 1/3rd pot is almost always a blocking value bet. This is all nice and easy when you know someone’s play that well. What about when you don’t? Again, you have to resort to generalizations and trying to figure out which level of thinking your opponent is on. Paying attention to who’s who in online poker ends up being kind of important as you go higher and higher in stakes. This is where using Pokertracker with PAHUD comes into play, as well as looking up a players results on something like OPR. Any piece of information you can get has value. Now let’s look at a more complicated hand. Here’s a recent post by Psyduck about a hand he played at the 10k ME in Foxwoods: Background: villain is a mostly straightforward TAG leaning on the tight side. He is one of the better players at the table and seems to be able to hand-read. He opens fairly tightly but also c-bets with a good frequency, almost always heads-up but he has given up some pots 3-way when he was the preflop raiser. He also doesn't value-bet thin at all, opting to showdown in position with middle pair-TK type hands, and some weak top pair hands as well. He also doesn't seem to double-barrel that much, but his flop c-bet frequency in HU pots is fairly high. He probably has some tricks up his sleeve but seems like 95% straightforward. I have a losing image. I am down from 30k to 20k mostly because of two big hands, one where I double-barrelled on a J T Q board and folded on the river (lost T5500 there), and once when I 3-bet AA preflop, and bet/folded on a K 8 5 rainbow board to a massive insta-check-raise-shove (lost T4000 there, and villain saw the AA fold). I am winning all my small pots and losing all my big pots. I have donked into the preflop-raiser a couple times on the flop and have either won or shutdown, and villain is paying attention to this very closely. Eff stacks 20k, blinds 100/200. Villain opens to T600 exactly in MP at a 9 handed game, folds to me and I call in BB with 7 5 . Multiple reasons I called, but let's focus on postflop. Flop Q 7 6 rainbow (pot 1300). Check, he insta-bets 600, I call fairly quickly. Turn 7 offsuit (pot 2500). I lead 1200, he makes it 2800 fairly quickly, I call. River 9 (pot 8100). I count my stack, and see I have like 15500 left. I lead 5200. He asks me to move my hands so he can see my stack, I do so and he sees I have very close to 10k left, and he shoves out all his 5k and 1k chips putting me in after about 1 minute. Let’s talk about the hand reading that needs to go on in this hand. Important information to start off with is that villain often C bets, and that he doesn’t value bet well. This means he has some passive tendencies and prefers to get to a showdown, likely meaning he’s less probable to run big bluffs. Preflop Psyduck has his reasons for calling, which is by no means standard but that’s irrelevant here. Flop: On the flop, Psyduck knows that since villain will be betting close to 100% of his raising range that his pair is often good and a call is appropriate. This portion seems standard when we know that very obviously there are many unpaired hands in the villains range. Turn: Now here’s where knowing your player, his tendencies, and hand reading start really taking place. Let’s start breaking down villains range and seeing how they react: AK/AJ/AT other overcards/air combinations that have not hit: Will check back and only put more chips in the pot if they hit a river. If they do hit it will often be for only one bet, as few of these hands will make something strong enough to raise Psyduck expecting to get called by worse hands if he checks the turn and leads the river. Pairs 22-55, 88-JJ: These pairs will almost always check back and try to get to showdown as cheap as possible. Due to Psyducks aggressive image they will certainly call one street from him, and just maybe two. Medium strength Q’s: From the description of this player he might check back with his mid Q’s like QT/QJ and if Psyduck checks the river fire a bet out. If Psyduck doesn’t lead it’s possible he only gets one street of value out of these hands. Strong Q’s and over pairs: These hands will just about always bet the turn, and often the river if it’s a safe card, but they can’t stand a check raise. There is a slight possibility they will check back for pot control when the board pairs, but most of the time we can expect them to bet but not call a check raise. Monsters like 66/QQ/A7: With hands like these if Psyduck check raises the turn he’s committed himself to the pot and they will obviously never fold. Psyduck knows that if he check raises the turn it is very likely that only these hands (except perhaps AA/KK and maaaybe AQ) are the ones calling him. As a result from understanding what happens against the various hands in villains range, Psyduck elects to lead out instead. Despite this leads main purpose being for value (as it gets more money out of villains range than a check) when the villain raises him Psyduck also gains information. He can now eliminate all pairs 22-55/88-JJ, medium strength Q’s, and overcards/air (unless villain is getting SUPER creative with him.) River: When the river 9 hit’s it changes absolutely nothing. 99 is no longer in the villains range. Psyduck knows that the upper end (the monsters) of villains range will always bet for value, and the bottom end (which at this point is strong Q’s and over pairs) will sometimes bet for value and sometimes check behind since Psyduck called his turn raise after leading out on a 7. Psyduck decides he gains the same amount of value from the bottom of villains range by leading, without being put to a decision for a considerably larger portion of his stack, with the added benefit that if villain shoves he knows he’s always behind and can safely fold. He leads a bit over half pot and what is 1/3rd of his stack, 5200, and folds to villains river shove. Nice hand IMO and a good example of how a thinking player used his understanding of hand ranges to his advantage post flop to maximize both value and information (please don’t kill me Gobbo.) For my last example I’d like to quote an entire article written by Nath at Tworags. I think this article is an excellent example of how to use hand reading into taking actions which under normal circumstances would be ridiculous: From yesterday's PokerStars Sunday Million. Villain is unknown. PokerStars No-Limit Hold'em Tourney, Big Blind is t600 (9 handed) Hand History Converter Tool from FlopTurnRiver.com (Format: 2+2 Forums) saw flop|saw showdown SB (t31052) BB (t35813) UTG (t29700) UTG+1 (t19400) MP1 (t28360) Hero (t27246) MP3 (t12775) CO (t7727) Button (t11100) Preflop: Hero is MP2 with 9 , 9 . UTG raises to t1200, 2 folds, Hero calls t1200, 4 folds, BB calls t600. Flop: (t3900) 5 , 4 , 2 (3 players) BB checks, UTG bets t3600, Hero calls t3600, BB folds. Turn: (t11100) 2 (2 players) UTG bets t4800, Hero calls t4800. River: (t20700) A (2 players) UTG checks, Hero bets t17646 (All-In), UTG folds. Final Pot: t20700 OK... now you're probably wondering why I took a line that seems really weird and determined to get as much of my money in the pot as possible while behind. And I'm going to show you why it works here. Let's look street by street. Preflop: Hero is MP2 with 9 , 9 . UTG raises to t1200, 2 folds, Hero calls t1200, 4 folds, BB calls t600. So this is the most straightforward street in the whole hand. A minraise is indicative of everything and nothing; mostly, it tells me my opponent is probably a clown. It doesn't do a whole lot to define his hand, though. Some players love to minraise their big hands to try to induce action on them. Some like to minraise hands they want to see cheap flops with, in hopes the minraise discourages a reraise. FWIW, I think both are pretty terrible, and if you consistently do one or the other you have a huge hole in your game. I try to blend my raise sizes so as to make my hand difficult to read. Anyway, having said all that, I elect to just call with 99 because I have no idea what my opponent has; if I reraise and he folds, I win a relatively small pot, but if he 4-bets me, I have to fold, and I've wasted a chance to win a big pot. So I decide to call and proceed postflop. The BB comes along because he's getting 5.5:1 and closing the action, not because he necessarily has much. Flop: (t3900) 5 , 4 , 2 (3 players) BB checks, UTG bets t3600, Hero calls t3600, BB folds. Here's where it starts to get interesting. Making a pot size bet here is often indicative of an overpair. Here's the catch: I still beat a few of the overpairs. In addition, some people panic with their AK/AQ hands when they miss here and just start firing big bets in hopes of scaring away an opponent. On this board, AK/AQ has an additional four outs against underpairs. This increased equity makes betting and getting it in with AK here really not that bad a play. It's also the big reason I don't make a move at the pot now. Some people see "overpair" and think "I have to protect my hand". Having seen him make a big bet at the pot now, I know my opponent likes his hand, but I don't know exactly what he has. It's too likely my hand is good to fold here, but he also has an overpair far too often to make raising and getting the money in profitable. I feel as if I get it in, it's going to be as a 60-40 favorite or a 90-10 dog. I don't mind getting it in as a 60-40 favorite, especially in a tournament like the Million, whose field size I feel dictates a faster style of play, a more "race to the finish line" approach-- but I get it in drawing to the two nines way too often to want to push now. So I call and decide to reevaluate based on the turn. The BB folds, and I never considered him to be much of a factor anyway. Turn: (t11100) 2 (2 players) UTG bets t4800, Hero calls t4800. The deuce doesn't change anything. Neither of us has a deuce and we both know that. Now, his turn bet is interesting-- he bets just under 1/2 pot, which seems weak, but which also sets him up for a pot-sized river shove if I call. (By the way, if you aren't thinking about manipulating pot and stack sizes like this when you size your bets, you are making a mistake.) For my part, the price is too good to fold an overpair-- but still, my hand is not good enough to raise. I suspect some time he has an overpair to mine, some time he still has AK/AQ (the bet size is actually an effective size to block-bet a draw and see if he hits it), and rarely, he has a worse overpair than mine. So considering most of his range as TT-KK and the other major part of it as AK/AQ*, I call again. * - I don't include AA here, not because it's impossible, but because it's a special case. He has the best of both worlds, and I'm screwed; the river is basically irrelevant because he's shoving all of them. I also considered 66-88 unlikely, though not impossible. River: (t20700) A (2 players) UTG checks, Hero bets t17646 (All-In), UTG folds. Wow. This is such an interesting card. All his ace hands got there, and all his overpair hands just got scared shitless. This is the beauty of position-- I can use his action to judge what he has. Since he has a pot-sized bet left, he's going to have to shove for value if he hit his ace (or his 33, or if he was already full, or what have you). He checks. Now, many, many players' immediate reaction would be to check, thinking that "Oh, either he had me already, or he hit that ace. And he's not calling with a worse hand, and we have a pair, so let's check and hope we win the showdown." They're wrong. When he checks, he's completely vulnerable. We each have less than the size of the pot left (he has me barely covered), and if we have anything reasonable, we're pot committed. The only reason not to put in the rest of the money here is if we think, for some specific reason, that a scare card has helped so much of our opponents' range, that we should give up. Trapping would be absurd at this point for villain, given that I should be calling his push with anything reasonable, having gotten this far. So when he checks, it's not to trap, it's because that ace scares the crap out of him and he will fold to a shove. And-- this is important-- we should realize that this swings all the Ax hands and 33/44/55 out of his range, and makes his range overwhelmingly overpairs. So his range consists primarily of hands that will beat ours at showdown, but cannot call all-in. Knowing this, we should move in as a bluff. The fact that he didn't move in already gives us all the reason in the world to. He's announced to us that he is scared of the ace, and he is hoping we will let him show down his hand. So we have to disappoint him. On a side note, the stack sizes are really excellent for this move. We each have slightly less than the pot left (I have 17.5k and he has 20k in a 20k pot). Which means that an all-in bet can be interpreted as a "normal" bet size. (Moving all-in here for, say, four times the pot would be considered "abnormal".) Because of that, it makes our opponent less likely to suspect that we are bluffing; we could simply be trying to get every dollar possible out of our hand. It puts him in a pretty terrible spot, since from his perspective we could easily have the AK/AQ/33. We certainly wouldn't check those behind on the river. So it's very unlikely that villain is good one time in three, and he folds. Now, at the table, this process is much quicker, and is occasionally guided by intuition-- you don't always have the time to think out, in words, why a move will work; you "know" it. While it's good to have sound, logical reasons for your moves, it's more important to trust your intuition-- it's a part of you, and it works on what you have learned, too. In the heat of the moment, it's taking all your experience and skill and training and leading you to the right decision. If you have prepared it for the moments you need it, it will not let you down. Work on your logic away from the tables, and study and review and prepare your theory, so that your instincts have the background they need to make the right decision. Trust yourself to learn the game. I will say that bluff-shoving the river hadn't entered my thought process until the river hit, and the villain thought for a bit and checked. Then it occurred to me that he couldn't call a push, so I pushed-- it wouldn't have really mattered what I had, but I was definitely swayed by the relative weakness of my hand. (If I had, say, KK, I might have checked behind-- or pushed for value.) But that's another point of this hand-- you have to be able to adapt your decision-making process each time new information comes to light. Even though you have a plan for a hand, something may change which will cause you to abandon that, because you realize an alternate line is more profitable. Online, the pace is fast, so you need to be quick mentally. Live, you always have an opportunity to think through a hand. Online, you have much shorter time limits and may be multitabling as well, so being quick on your feet is just as important as being sound on your feet. So to recap today's lesson: a)Be alert to your hand strength relative to your opponent's range, not just to the board, or in the absolute sense b)Don't be afraid to turn a made hand into a bluff c)Don't be afraid to change plans in the middle of the hand, as you gather information d)Be a quick thinker e)Trust yourself I really think Nath covered it pretty well there and I’ll leave it at that. As always, if there’s anything more you’d like to know or have questions about, just ask. One last note, I’ve placed Betgo in charge of putting up the official review threads since I’ve been very irresponsible with them. I’ve given him access to my 2+2 email and he’ll start putting things up much more consistently than I ever did. 
|
|
#6
11-15-2007, 09:01 AM
|
|||
|
|||
|
Re: Things it took me a while to learn part 9, Hand Reading
that nath hand is v interesting.. if i had the 9's i'd be letting out a sigh of relief on the river that i got to showdown my hand and protect some of my stack... oops!
one of my fav in the series. thks bond (and nath)!
|
|
#7
11-15-2007, 10:02 AM
|
|||
|
|||
|
Re: Things it took me a while to learn part 9, Hand Reading
Bond,
Another great one, obv. It's excellent to get insights in the SSMTT forum that go beyond "you have AK - push!" However, I believe the examples you give -- while really interesting -- aren't very applicable to an audience that plays mostly $20 and below tourneys. For example, you have instances where you say something like, "he knows that I know this" or "villain would have done X if he had Y." But in the games I play (for now, hopefully, until I can move up), 98% of opponents aren't thinking about anything other than what they have and how it hits the board. You would think this would make hand reading easier... but that's not what I've found in my experience (while also admitting that I may simply just suck at it). The widespread unpredictability of the actions in lower-buyin tourneys makes for more guesswork than analysis. They limp AA from the SB with two limpers in already... they checkraise bottom pair on the flop AND turn... they call PSBs with gutshot draws... they insta-minraise flop bets with an underpair to the board (or is it top set?). They do so many things in the name of pseudo-deception and/or bravado that assigning any kind of a reasonable range is a real challenge. (This is a generalization, of course, as there are times when I can really narrow a range down based on action/history/observation... but that's way more the exception than the rule.) Now, I know these players' actions are generally bad and will lose money over time. But that's not the issue here -- we're talking about hand-reading... and this type of play, while clearly bad, does put up a pretty effective shield for anyone attempting to assess their range. So, I guess what I'm getting to is this: what adjustments to general hand-reading strategy do you make when playing vs weak competition? Thanks and keep up the incredible work, Matt
|
|
#8
11-15-2007, 10:56 AM
|
|||
|
|||
|
Re: Things it took me a while to learn part 9, Hand Reading
The 9s hand is really interesting. If you do the hand reading from the other perspective it's really hard to put nath on a hand that shoves that river for value. Would he call flop/turn with AQ/AJ or AK if he didn't reraise pre? Against another strong thinking player I think the river shove isn't very good and folding the turn would seem good as not too many unknowns will double barrel AK/AQ on that board.
Overall a great read though and I would love to get to the stage where this thought process is automatic.
|
|
#9
11-15-2007, 11:04 AM
|
|||
|
|||
|
Re: Things it took me a while to learn part 9, Hand Reading
Bond18, thank you for the post. I have a question regarding game preparation.
[ QUOTE ] Bond18: If you have prepared it for the moments you need it, it will not let you down. Work on your logic away from the tables, and study and review and prepare your theory, so that your instincts have the background they need to make the right decision. [/ QUOTE ] I'm always looking for ways to work on my logic away from the tables. However, once at the tables, I have trouble making decisions in a quick, decisive way. I always back down on a plan at the table if I find myself taking too long to make a decision. I have a tournament chess background, which still allows a comparatively large amount of time to analyze moves prior to making the play. What is the best way to improve my quick poker decision-making?
|
|
#10
11-15-2007, 11:59 AM
|
|||
|
|||
|
Re: Things it took me a while to learn part 9, Hand Reading
Another very good post in your series - for me the best one up to now - thank you.
With hands like the once you have given as examples I can see how much I still have to learn and how important hand analysis on the green table is, to sharpen our instinct-skills at the poker table.
|
 |
|
|
Linear Mode
