Thanks for all the comments on What’s Your Play? Big Draw vs Bad LAG. I hope you’ll find the results and analysis more interesting than the typical bad beat post, which if I’m being honest was part of my motivation in sharing this hand.
The reason I find it interesting is that it reminds me of an important conclusion from Mathematics of Poker that I’d forgotten about entirely until I reread that book. In “Playing Accurately, Part I: Cards Exposed Situations”, Chen and Ankenman demonstrate that there are situations where a player with an obvious (exposed, in their hypothetical) draw actually does better by raising all-in on the flop, knowing he’ll put the rest of his money in from behind (though with sufficient pot odds) rather than calling and giving his opponent the opportunity to bet him out on a blank turn or check-fold when the draw comes in.
Of course there are any number of reasons why that may not be applicable to this hand. It presumes an opponent who will correctly bet the turn when ahead and correctly check-fold when behind. If, as Eddie argues, Villain can be expected to run a big bluff on cards that complete Hero’s draw, then there is more room to outplay him on future streets.
On Calling
I think many commenters are overestimating how easy it will be to outplay Villain on a blank turn, though. Shoving over a turn bet works only if Villain has a lot of air in his range, as he probably isn’t bet-folding a Q or a 9 or a better draw. Calling in hopes of getting there on the river probably requires decent implied odds to be better than getting it in on the flop even if Villain never folds.
I’m not at all sure what would be the best play if Villain checked a blank. I suppose I would opt for a smallish bet of perhaps 1/3 pot, but getting check-shoved is such a disaster that I don’t feel too great about that plan.
I don’t think the best play is particularly clear if the turn is a J or a T, either. Some people seem to want to call down on blank rivers, and I can see why, but Pepito also makes a good point that “villains of these sorts tend to thin hero calling rather than triple barreling with air (particularly in a 3b pot where hero has called a check raise and a second barrel).”
On Shoving
Pepito also asserts that, “shoveling 4.5K into 1.2K pot against a callbox seems terrible– particularly with position.” Having already explained why I don’t see a lot of ways to capitalize on my position, I’ll now address the “callbox” argument.
The tricky thing here is that this hand actually blurs the line between a value bet and a bluff. On the hand, it certainly feels icky to stick in a lot of money with Jack-high against a player who’s shown a willingness to call down extremely light. But what, really, can his looseness do to hurt us?
No one is check-raise-folding the hands that have us in truly bad shape, which would be nut flush draws, two pairs, and sets. I wouldn’t really expect Qx to take this line either – people might check-raise-call it, or they might not check-raise it, but I don’t see it getting check-raise-folded.
So what can Villain’s looseness add to his calling range, and how can that hurt us? All he can do is start calling with really weak hands, and Hero is a pretty big favorite against those, as Sean F. points out. Hero has 63.5% equity against 9d 8d and 76% against 8s 7s.
There’s a lot of talk about not wanting to put your stack in “on a flip”, but that’s a misleading way of framing the situation. What you really need to consider is not just the end result – how much equity do you have when the money goes in – but rather how can your actions manipulate the situation to give you more or less EV.
Regardless of your flop action, there are a lot of scenarios where you end up playing big pots against the top of his range. Against the bottom of his range, you’re a solid favorite, and raising the flop can only lead to good things. Villain can either forfeit a significant amount of equity by folding, or he can put his money in from behind and lose the ability to play well on future streets (he might continue bluffing a spade with air, but will he bet/stack off with 98?).
A lot of the value in calling comes from inducing bluffs. Maybe we get to shove over a bet on a blank turn, and maybe we induce some big bluffs when we hit.
Results
Raising less than all in still leaves open the possibility of inducing a bluff. If Villain check-raises the flop with more than just pairs and draws, then there’s a real chance he’ll four-bet-fold some of that air. He’s shown a propensity for this sort of play in wide range situations before.
That last point was the deciding factor for me. I three-bet the flop $1300, Villain made it $3100, I shoved, and he did some thinking (about what, I’m not sure – whatever it was presumably should have happened before he four-bet) and then called with what turned out to be Ad Kd. That was an outcome I hadn’t anticipated, but again I’m a 65% favorite in the case where Villain decides to make a really loose call, which means that getting it in on the flop is better for me than shoving over the check-raise and having Villain fold (though I would have preferred a fold to the five-bet).
The board ran out blank, and Villain collected a massive pot with Ace-high while the table collectively gawked. I warned you this was a bad beat post!
super ugly brick out, but then definitely sip water and fist-pump auto-reload.
given this juicy new data point on villain, more interesting questions arise: what extrapolations on his logic can be made? how far to the extreme can the bars for thin value betting and hero calling be dragged? how can his newfound perception of you be exploited?
a good starting point is speculation on how villain would have played the turn had you flatted his raise.
despite his kamizake chk-r/4b/call, i’m inclined to believe there’s a strong chance villain would have actually snap checked this run out to the river. his mindset/thinking are arguably: “everyone is always bluffing, so if i feel i’m ahead, i’m going to raise to find out.” consequently, your b/3b in fact only further strengthened this original suspicion, so he doubled-down on his original gut feeling and 4bet you.
he tank-calls your 5b shove because he now (wrongly) believes you aren’t bluffing, but rather are jamming it in with Qx (as that is preceisely what he would do). the annoying extra time spent before calling was him laboring over the calculations of whether he has the pot odds to hit an A, a K, or his gutshot.
revision: should say “backdoor” gut shot.
I hope you practiced your best Tommy Angelo afterwards. Show the table you’re playing a different game from them.
Thanks for sharing and allowing us to learn from your pain.
I like this hand analysis.
Oh wow, yep that’s pretty amusing except for the fact you lost such a big pot. To me it illustrates how hard it can be deducing correct lines and pot equity against the range of bad LAGs/maniacs, but of course we totally want them in the game.
Tell him with a wink that the money is a loan, and re-load with your great position!
Agree with your analysis! In that the theoretically correct play must be to 3bet less than all in – as Andrew says to allow him to bluff while you have greater equity. However, even if a little results-orientated, the let’s call it less sophisticated “losing value” 3bet shove must have some merit because the fold/fear equity is greater – even for this maniac. (Perhaps this supports Bill Chen’s conclusion too?) And if villain’s never folding, the flat call evaluate turn must still have some merit – so I’m a little conflicted on the best line, which is I suppose what makes it such an interesting hand to analyse.
I read something from Snowie where it will shove in a spot like this with J high but just call with the nut flush draw due to having some showdown value.
Carlos, unless the results have a typo, Villain spazzed out with overcards and a backdoor nut draw: the flop was two-tone with spades, and V. bluffed with AdKd. I don’t think Snowie would call a 3-bet OOP on that flop. Which leads to my question for Andrew: Sklansky/Miller point out (in NLH, Theory & Practice) that our semi-bluffing frequency with nut draws, or with draws that are completed by an overcard to the board, should be inversely proportional to the effective stack depth: deeper stacks make semi-bluffing these draws more -EV; we don’t want to fold out worse, because the completed draw is likely to elicit a bet from villain (with either a made second best hand or a bluff). However, as Carlos has pointed out, villains in today’s wild small games (1/2 and 1/3) will often spazz out with overpairs and big slick (Andrew notes that he did not consider AdKd in villain’s 4-bet range, but at smaller stakes he probably would have). So my question for Andrew is against these type of players, do we need to reconsider the EV of these semi-bluffs? How would you adjust your flop betting range in this spot on future hands against this type of player?
Andrew, here’s that PokerSnowie post I mentioned. I know it’s not exactly the same, but maybe we can glean something from this.
http://www.pokersnowie.com/blog/2014/12/28/old-vs-new-ai-bluff-raise-flop
Interesting, thanks. Seems consistent with the MoP conclusions.