Last time, I wrote about an HSNL hand where putting your opponent on a very exact range made it mathematically possible to determine the best line. In this installment, we are going to expand on that by going back to HSNL to read this thread. Descriptively titled "fullhouse", it contains a 100/200 NL hand between Grimstarr (a well known LAG) and another good online pro, Jinksop. This particular hand has some great hand reading material leading to a very tough, but necessary decision on the river.

Let's take it street by street. Preflop, with 25K stacks, Grimstarr completes 63o and Jinksop checks. The flop comes down 7 6 3, Jinksop bets the pot (400), Grimstarr raises to 2000, and Jinksop calls. What can J have at this point? Clearly, a wide range (any seven and up), but with the aggression factor of these games, we can rule out any set or better* and any big draw (98, for example, would drool to be all in here). What does J think G has after his raise? It can really be almost anything (with Grimstarr, I doubt 86 is the bottom of his 'real hand' range, not counting all the bluffs), but let's conservatively assume something like "draw/pair + draw/decent seven or better" for this particular ugly looking limped pot board. However, J does know that G will almost never show up with a set here since G will probably be raising any small pair PF.

*This deserves an explanation. As you can see, there's some good debate in HSNL about whether people will check hands like 77-66 in the BB when the SB completes heads up. I don't have the nosebleed stakes experience to comment on this, but the combination of checking PF and then just calling the big raise on that board makes no sense. Given that people frequently 4 bet A7 in that spot, it's very very unlikely that J has a set. By the same token, we can also discount 54, although PF makes that more likely.

The turn is the T and both players check. What can J have now? The T itself (not the club) won't usually have improved him after bet/calling a big raise, but he can definitely have whiffed on a CR with a flush or a straight. Most hands that he could have bet/called with are usually going to check/something on the turn, as bet/call, bet is a comparatively rarer line and the board just got uglier. G's range, though, narrows down considerably for J; good LAGs will not check behind small flushes on boards like this, and most other 'really good' hands will also bet. The hands that will often check behind here are bluffs that have given up, one pair marginal hands that check this board because they might be good but would hate a fairly standardish semibluff CR (bottom 2 pair is pretty much in this category) and occasionally bigger hands like 76/54 that also hate a CR (although these will often bet anyway.)

The river is the 3. J leads 3000 into the 4400 pot; not counting bluffs, occasionally, he'll be in here with a big one pair (really 2 pair) hand like A7, or a T that wanted to bluff and caught. Usually, though, this is the top part of his range, because a lot of G's 'vulnerable made hand' range from the turn is autocalling when the board pairs the river. For example, J cannot fully discount AA-QQ from Grimstarr, and nobody folds those with that action.

Grimstarr raises to 8k with his bottom full house. We can instantly forget any marginal hand raising this river (I hope you see why**); it's a bluff or a straight+ (maybe a very thin raise with A3 if he raised the flop with it), much more often a flush. Keep in mind that, if G has a flush, his range is heavily weighed towards bigger flushes - 4 3 doesn't check behind on the turn very often.

**This is only the case vs. thinking opponents. If you're playing 25NL vs. a guy whose hand range by this point is any two cards, you can often go ahead and raise something like A7 for value.

Now Jinksop shoves for 25K. High stakes poker has its share of legendary hands, and a 3 bet bluff in this spot would certainly be legendary, but we can discount it down to 'just about never'. However, he knows Grimstarr's range is essentially...let's say 15% bluff, 25% "big hand < a big flush", 60% "a big flush or better" (we can play around with the percentages, but it won't make that much of a difference) - but never sevens or sixes full. It's a rare spot where villain knows not just the lower part of your range, but also the upper. He also knows that Grimstarr will never call a 3 bet with a medium flush or worse (again, this should be self explanatory), yet he shoves anyway. Therefore, J is saying he can beat a flush, and 63, much like A x, is a bluff catcher. In a very sick spot, G has a full house heads up in high stakes NL and should fold it.

In the actual hand, Grimstarr timed down, called and was shown 73 for the slightly bigger boat. In my opinion, in addition to that money, he lost a few thousand almost as precious Sklansky bucks.

EdmondDantes says

Can you lay out the distinction between Sklansky and Galfond bucks here? I was under the impression that Sklansky bucks are the EV$ of the call. G's dead at the river so all the dollars lost at the river are EV-, yes? He's also losing Galfond $ but a smaller amount of them because we aggregate [(probability of hands we beat here x the money won) + (probability of hands we lose to x the money lost)] which in this case is less than the Sklansky bucks. Am I not making the right distinction?

04/14/07

Adanthar says

Sklansky bucks are simply "money that Grimstarr makes/loses by calling vs. J's range [which is, say, superbluffs/weird flush shoves/boats]", so, yeah, it's the EV of the call. The G-buck thing doesn't really apply here, because it's a range-wide concept - to clarify, if Grimstarr calls a 3 bet with the top X% of the hands he raises with, he would make or lose Y G-bucks vs. J's range. In other words, it would apply if you went back one step and asked "I'm deciding whether to raise the river in the first place. What range should I raise with (if I call a 3 bet with the top X% of it) in order to maximize my EV vs. J's entire range?"

04/14/07

EdmondDantes says

I read the thread on 2+2 and like your analysis here, but as someone with more grey hair than most posters, I'd suggest that anyone analyzing this hand be careful about discounting the tendency smart people/players to make and compound mistakes. If I run out the math in the EV calculator, the breakeven of the call is somewhere around 35% as seen here...

TwoRags.com EValuator Tool Scenario (change hand ranges)

Grim's Hand: 6d 3h
Board: 7c 6c 3d Tc 3s
Pot: $27400
To Call: $15000
Pot Odds: 1.8 to 1

Villain Hand Probability Win%
1: 7h 3c 70%
2: Ac 4c 30% 100%

SUMMARY RESULTS

Win Probability for this Scenario: 30%
Less: Break-even win% needed: (35%)
Negative Variance: (5%)

Total Expectation: $12,720
Less: Amount of Call: ($15,000)
Negative Expectation: ($2,280)

So I have to know villain has the goods over 65% of the time or it's a bad call. Ok, he's probably got it at least that often given the line.

That said, in the 2+2 thread, Grim says they're playing four other high stakes tables. Yet still the majority of posters in that thread assume that the line the villain takes is logical throughout. Once the hand moved past the flop, given the stakes, chances are both guys focused pretty closely, but to blanket assume that villain played logically from the start is a stretch. It was an awkward hand for both and one that's generated a few days of discussion. What was it like in real time?

If Grim had called with his dinghy (a boat, nonetheless) and snapped off a misplayed A4c flush or worse, the post would likely have ended up in some BBV "I can see into your soul" thread. Instead, he got shellacked and most responded, "Of course." I'm not saying your analysis here is playing results by any stretch--it's first-rate--I'm just saying my experience is poker and life is to be leary of rushing to assume optimal or logical play by even a very smart villain.

Bottom line, I'd want to have watched villain closely throughout the session, be convinced he's playing solid, focused poker (even though he's on 4 other tables) and taking a straightforward, logical line under time pressure. Phil Mickelson's a great golfer with a couple of majors on his trophy shelf, but he sometimes makes decisions under pressure that leave even the experts asking, "What was that?". We all do.

Edmond

04/14/07

lakong says

Great hand and really nice analysis. I'm assuming that you're saying it's not a great call (or a bad call?), but I wasn't clear on your thinking -- would you have had no trouble folding in that spot?n One theme I noticed in the thread on 2plus2 was the line that goes like 'good players would raise PF from the button with a medium pair' or 'good players would raise from the blinds with a medium pair', or 'good players would never make the last all-in raise if they couldn't beat the bottom set, so therefore a good player must fold to the all-in with 63'. Well if this were true, and a good player would never raise at the river with a hand that could not beat the bottom set, then perhaps a really good player would raise with a worse hand. Why? Because a good player would think that a good player would not raise in that spot, so therefore a good player could not call with bottom set. Am I making myself clear here?

04/14/07

Anonymous says

I watched the hand and couldn't believe the action on the river. Ugly! What many people are missing is the reality of playing online. The analysis makes a great deal of sense, but can you truly apply it in the few seconds that you're given online? Probably not so it comes down to instinct. That's why discussions like this are great because the more of it we participate in, the more likely we are to burn the scenarios into our soul and we can then just react correctly, or as correctly as possible if given a similar situation.

04/14/07

tateissy says

I liked your analysis of why a set on the flop makes no sense for either player once we get to the river. However, I'm not convinced that the villian can't have a set. Let's say he has 77. So he calls PF from the blind. It's probably not his normal move, but it's certainly possible some % of the time. Then he hits his set and leads out, but gets raised. The main hand he's worried about is a flush draw, but does he have to re-raise here every time? 100% of the time? Can't he just call and let a non-club come off and then go for the check-raise? I guess that wouldn't be great, would it given that he might find himself giving two free cards which is very dangerous. He also isn't going to call on the flop and then lead out a non-club turn since he would have then been better off just raising in the first place. Anyway, I guess I'm almost in agreement with you here, but just not 100% convinced it couldn't happen.

04/14/07

Adanthar says

I should make something clear - in nosebleed NL, when I say 'never' I probably don't really mean it, solely because of the pure aggressiveness of the typical player in that game. It's theoretically possible for either of them (not just Grim) to show up with T3o on the flop solely because they planned to bluff the turn. Similarly, either of them can have 77. It's just wildly unlikely, because the way that nosebleed HU matches work means that on a scary board like this one you can sometimes 4 bet something like A7 on the flop for value. There's no reason not to bet/3 bet a set, because neither of them is folding much anyway.

I would definitely fold in Grimstarr's spot, BTW. That said, if you know you can't make a big fold like that one, it's a factor in whether you raise (in any given spot) and what you raise to. I might write about that in the future.

04/14/07

Anonymous says

Adanthar, you say that you would definitely fold in Grimstarr's spot. If this is the case -- that a good thinking player would fold in Grimstarr's spot -- then wouldn't you have to agree that the villain should almost always push with any two cards on the river to force a thinking player to fold... assuming of course that the villain knows that his opponent is a thinking player. The villain would know that Grimstarr's range at the river is a flush or baby set. So he should be confident that his push will not be called, right? This seems to be what Brad Booth did on high-stakes poker when he pushed Ivey off of his KK on the flop with his massive all-in re-raise. He could not do this against a mediocre player who would place to much value to KK in that spot. But Ivey was good enough to lay down his KK and Booth knew that.

04/16/07

Adanthar says

I would definitely fold in Grimstarr's spot *as a LAG who's not supposed to fold anything*.

The trick is to know what your opponent wants/expects you to do. People who raise the river with hands that are worth raising the river with (not bluffs) veeeery rarely fold to 3 bets, and this goes triple for superLAGs playing nosebleed. So a 3 bet is a megabluff, or something that J wants to get called with. On this board, a megabluff that was planned on the flop is really unrealistic, so we're really talking about J 3 bet shoving a made hand as a bluff, which also never happens. (Read some of Jason Strasser's posts in HSNL for thoughts on why it should happen more often.) The bottom line is that J probably wants G to call this bet and definitely doesn't expect any of the top part of Grimstarr's range to fold. Therefore, you should probably disappoint him.

04/16/07

Chooose an identity

Login/Register Enter Name Anonymous

Log in with your TwoRags.com account. Click here to register. Email: Password: Remember log-in information Name: