Consider this blog as practice for my poker book (in the somewhat unlikely event I decide to write it). The format is going to be "Hands with Sauce", borrowed wholesale from the classic bridge book "Bridge with Reese" written by the renowned bridge pro Terrance. The idea is that I take the reader through my thought process, such as it is, in intriguing hands, stopping along the way to touch on various theoretical topics of interest and then, upon further reflection, dissect the plays which result. Without further ado then,
Hands with Sauce #1: 6bet shit-flinging vs AARookie
Stakes: 40/80nl 6-Max. Game was 5-handed
Hero (SB): $33,739.5
Pre-flop: 9c6c dealt to Hero (SB) ... UTG folds. MP folds. BTN raises to $178. I 3-bets to $720. Villain 4-bets to $1920. I 5-bet to $5127. Villain 6-bets to $9055. I called.
Flop ($18,288): 8s 2s As … I checked. Villain bets $6650. I called.
Turn ($31,588): Qc … Check check.
River ($31,588): 4c … I shoved. He called with 4d6d.
This hand seems to be fairly infamous across the internetz, and I still get pms asking about it fairly regularly. In the past I've given some one or two sentence answers which seemed to satisfy basically nobody, so here I'll try to do you guys a bit better and go through things piece by piece.
First, in order to get into character for this hand, you have to understand just how insanely aggro AARookie was playing during this session and previous sessions. As far as opening pots and 3betting, Rookie was unspectacular, playing maybe 15% 3b vs steal and then something like 40% steal himself. However, it seems as if Rookie felt he had found the holy grail in the positional 4bet, as I have never seen a sober person 4ball as much as AArookie did during the 20k hands of 40/80 he played in the couple months surrounding this hand. My HUD had him 4betting something in excess of 40%, and while I don't have a stat explicitly counting cold 4bets versus 4bets after opening, it was unmistakeable that his cold 4betting frequency was even more off-kilter than his 4bet after steal frequency. In order to combat this, of course, me and everyone else with half a brain at the table had tightened up our 3betting ranges significantly whenever AARookie had position (even if he was halfway across the table) and yet he was undeterred, and continued to 4bet seemingly everytime he hadn't checked the auto-fold button.
Judging from the opening size of $178 (although I don't remember explicitly), I'm guessing that the opener in this hand was Takechip, aka FLIPokerHer on Full Tilt Poker. At this time, Takechip was playing a curious strategy of opening nearly 100% of buttons, and then folding over 75% of the time to 3bets. He also seemed strangely unresponsive to punishment, so I was laying it on pretty thick and playing a strategy of 30% or so 3bet from the small blind, depending of course on the predilections of the particular bb. In this case I held 9c6c and so I was torn- on the one hand Takechip had offered me 2.2 bbs, but on the other AARookie was lurking behind me. Taking the hand from Takechip's perspective, I decided that he was likely to be especially loose from the button in this exact situation as Rookie was nitty versus steal in the bb, playing something like 40% of hands vs a minraise. Takechip was the type to pick up on this, but he wasn't the type to realize I would pick up on him picking up on this, so I decided to throw in $720 and see what happened.
Unfortunately, Rookie woke up behind me and made it his customary $1920. At this point you could argue that I got overly stubborn; there are certainly more ideal hands to 5bet bluff with than 9c6c. Previous to this action I had made three 3bets at the table, and on all 3 of them Rookie 4bet me, and finding myself with nothing I elected to fold. I believe 2 of 3 were cold 4bets, and I had witnessed a few other cold 4bets made by Rookie. Appraising Rookie's range in light of today's play and his overally splashiness I was confident it contained plenty of fluff and I could justify 5betting. To put a number on it, I guessed he was 4betting this spot with between 12% and 35% of hands. Quantifying my own strategy, my Rookie-avoidance plan was to go ahead and 3bet a fairly polarized 10-14% vs steal. As a rule of thumb, when the guy who puts in the 4th bet has a wider range than the guy putting in the 3rd bet, either the 3bettor plays awful or the 4bettor is putting too much money in too often. So, I knew I would be 5betting a frequently with my range in this spot for high profit, but not having tested Rookie's expansive range yet with one of my bluffs, I decided the time was going to be now since if he gave me credit he could be folding as much as 85% of his range. I made it $5127.
After a bit of thought, Rookie made it $9055. At this point your guess is as good as mine as to his range. I have my particular way of strategy-reading, and one thing I am uncomfortable doing is making intuitive leaps with regard to a range I have never encountered before, and I had never encountered a Rookie 6bet before, certaintly not one with over 300bb stacks. Interestingly, many people who I have talked to about this hand thought Rookie's 6bet was obviously a bluff, but I have to admit that I don't think I could ever make that read, or would want to. But taking his incredibly wide 4betting range as an assumption, I could envision various ways in which he might decide to play, and which were consistent with his overall approach. So I ruled out a strategy where he would fold everything but QQ+ AQs+ to 5bet; he was probably an aware enough player to realize an opponent like me would be coming for him with bluffs at least occasionally and that if he didn't play back his strategy would be easily exploitable. This statement is not in contradiction with my previous paragraph where I conjectured he might fold 85% (all his non premiums), as once I saw the min 6bet against my first 5bet, the Bayesian probability of him playing fit or fold against 5bet goes down a little bit, and perhaps even more than the bare probability sugggests since it was the very first time I had ever 5bet and Rookie might be tempted to draw a line in the sand now in order to continue his putative exploitation of me in the future. I decided to attribute to him a strategy of playing back with between 20% and 80% of his bluffs (quite the wide confidence interval!), roughly 15% of his ATs-AJs, 99-JJ, AQo combos, and then roughly 80% of his QQ+ AQs+ combos. However, since his postflop strategy was also weak, and since occasionally Rookie's strategy might be very tight for 6betting, and since I held a hand with poor equity all in preflop without blockers, I decided not to jam. This was probably not the right choice incidentally, since I risk $28,612 to win $14360 if I jam preflop, and if I have 27% vs a calling range, I must succeed at least 58%. Taking a weighted average of my assumptions, he should fold at least 58%, making a jam +ev. Still, a call may very well be even better. Getting $3928 to $14360, or 3.65 to 1, and being only a 2.57 to 1 dog against a nightmare scenario of TT+ AK+, I felt I couldn't fold, needing to realize only around 2/3 of my equity against that range to break even. So we saw a flop of As 8s 2s and I checked, as I would do with my whole range.
Rookie bet $6650 into a pot of $18288. By all accounts I should fold here given that I have the absolute bottom of my range. I was reluctant though, as Rookie had a very persistent betsizing tell which made an appearance. Generally, on dry boards, or boards which looked favorable to Rookie's range, and especially in position, Rookie liked to bet a polarized range on the flop for 1/3 pot. He had exhibited this tell many times across various sub-textures and in various potsizes. I also had some stirrings of memory of him making a larger betsize with some strong but vulnerable hands, at least sometimes, and so this betsize made me think that he was some unknowable amount more likely than normal to have a range of AxKs+ type hands along with pure bluffs (as opposed to semibluffs). He might also have a range of AQ+ QsQx+ and then pure bluffs, electing to checkback with Ax type hands, mediocre made hands, and mediocre spades, and then some percentage of his nuttiest hands to trap. Whichever of these strategies he chose, everytime he had made this betsize his turn strategy was very transparent- he would barrel on the turn with the strong parts of his range as well as his strong semibluffs and the occasional pure bluff, and then check the remainder. He might also checkback some very strong hands on the turn in order to induce a river bluff, but as I had never seen him do so previously, I thought he was significantly more likely to play these hands as a bet/bet or as a check back flop to call/call. This texture presented an especially good opportunity for me to take an OOP float against many of the strategies he was likely to play- in all previous cases Rookie had not shown down a non-suited hand in a 4bet pot except for an offsuit ace, and since he cold 4bet with very deep stacks I expected him to have a range with more suited connector hands than his usual for deception and implied odds. So, on this board texture, his probable 4betting range would either flop a strong pair of aces which would bet flop, a flush which would usually bet flop, a few big pairs with spades which bet flop, and a few Ax w/ spades hands which would bet flop. This left him with a high ratio of essentially zero equity bluffs, the precise amount of which would depend on just how polarized a flop betting range he decided to play in this case. So, I could pay $6650 for the option to gain very accurate information about his range on the turn, and then if he checked, I could exercise my option to make a highly profitable river bluff for 2/3 pot. To sum up, if his range was possibly bluff heavy preflop, and by assumption from previous play I knew he was almost 100% likely to checkback AJ or worse without a spade, and if he was bluffing flop this was his single most likely betsize, I can update the Bayesian probability of his range containing bluffs to some amount higher than I had attributed it based on preflop play alone. If I had to quantify the muddle of my quantitative and psychological reads at this point, such as they are, I would estimate he was 6betting around 40% of his 4betting range, and cbetting a range of well over half bluffs. I chose to exercise my option and see if I could get my turn checkback.
The turn was the Qc and the action was check/check. The river was the 4c, and after a suitable 2-3 second tank I shoved as planned. Rookie went very deep in his substantial timebank and ended up calling with 4d6d.
The two attached Cardrunners EV trees should do a pretty good job of modeling the profitability of my ambitious float. I tried to set fairly pessimistic assumptions relative to my various reads: the preflop conditions give Rookie an initial 4betting range of 18% (low relative to the mean of my assumptions of 12-35%), and I gave him a 6bet range of 80% of his premiums (I expect him to slowplay them occasionally) which I defined as QQ+ AQs+, 15% of his stronger hands 99-JJ AQ ATs+ (I expect him to mostly call them), and 40% of his possible bluffing combos (fairly low relative to my assumptions of 20-80%). I also gave Rookie credit for a pretty unfavorable postflop range for my float, betting flop with 100% of his AQ+ combos, and then having the presence of mind to checkback turn with 60% of his flushes and 40% of his big combo hands to induce my bluff, and I also stipulated he would call the river shove with 40% of his middle pair combos and 10% of his bottom pair type combos. It is also important to note that a turn or river spade is an especially good card for my bluff, (proceeding on the assumption he will only call a river jam with a flush) as his flop betting range is spade-light. Given all of this, on the actual runout, my flop call gets a turn check/check 54% of the time, and a river fold 62% of the time. Given his small flop betsize this makes my flop call worth $334 relative to fold, or +4bb.
Keeping postflop assumptions static and playing with his 6bet bluff frequency, my line starts to be bad at a 6bet bluffing frequency of around 35%. Although this depends a fair amount on just how good various other runouts are for my play (assuming I see the 1/3 pot cbet sizing from Rookie on all runouts when he does decide to cbet) and I haven't done further analysis of these contingencies. There is also the question as to how often he elects to hero call any pair on the river- set this number to 100% and my play becomes quite -EV, although I do expect that the EV for my range might increase. All in all, I think my float is winning me a handful of big blinds across the weighted set of his possible strategies, although its precise value is unknowable.
There are quite a few interesting points to this hand in my opinion. Mostly, Rookie's play shows us what not to do. Cold 4betting too often with only one bb invested in the pot relies upon your opponent having large pre or postflop leaks- typically of the folding variety. Since I don't fold too much, pre or postflop, Rookie's 4bet mania was bound to cost him eventually. The other lesson here concerns postflop play- by continuation betting a polarized range on a fairly static board texture, not bluffing enough turns, and not slowplaying enough turns by checking behind, Rookie allowed me to get very good information on his range by just calling a 1/3 pot flop bet. This information proved valuable enough to turn my entire range into +ev floats.