HU rake considerations in split pot games

[jbrennen]

There's a game theory equilibrium point here, I think it's something like always betting your potential scoop hands, and betting your low-only hands around 30% of the time -- but the exact fraction depends on pot size and how often you actually have a scooping hand.

The idea is to bet just enough of your one way hands that your opponent can't profitably fold when he has unimproved Aces -- but betting more of your one way hands than called for by the game theory is just going to entice your opponent to call you down every time, which just ends up increasing the rake on those split pots.


I think that many players will just jam the pot here essentially 100% of the time with their low lock, not realizing that by doing so, they are making their opponent's decisions easy. (And due to the rake considerations, they're actually giving up EV.)

[Tha Stunna]

I think everyone is trying a little to hard to answer this question; the main point is to bet when your opponent has even a small chance of folding, mainly when you have a made low but your opponents unpaired board beats you for high. Occasionally you can pull this off when they have an open pair; some people are stupid. This works a lot better if you make a low when you have a high door card.

Don't feel obligated to bet when your opponent showed lots of strength earlier, although if they may have been stealing or bluffing at a third player betting becomes useful again. A few stupid bets don't really hurt anyways, particularly if you need to vent or something.

The same logic applies when you have something like a made flush and your opponent has a made low.

[AlanBostick]

[ QUOTE ]
There's a game theory equilibrium point here, I think it's something like always betting your potential scoop hands, and betting your low-only hands around 30% of the time -- but the exact fraction depends on pot size and how often you actually have a scooping hand.

The idea is to bet just enough of your one way hands that your opponent can't profitably fold when he has unimproved Aces

[/ QUOTE ]

Unless your board is superscary, your opponent with unimproved aces has to call essentially all of the time anyway.

So, unless your one-way hand is showing three connected cards or a three-flush, you are guaranteed a call if you bet a scoop draw.

How often should you bet a low-lock with no high draw in order that your opponent is indifferent to calling or folding? More than 100% of the time.

Jam your low locks.

[jbrennen]

[ QUOTE ]
[ QUOTE ]
There's a game theory equilibrium point here, I think it's something like always betting your potential scoop hands, and betting your low-only hands around 30% of the time -- but the exact fraction depends on pot size and how often you actually have a scooping hand.

The idea is to bet just enough of your one way hands that your opponent can't profitably fold when he has unimproved Aces

[/ QUOTE ]

Unless your board is superscary, your opponent with unimproved aces has to call essentially all of the time anyway.

So, unless your one-way hand is showing three connected cards or a three-flush, you are guaranteed a call if you bet a scoop draw.

How often should you bet a low-lock with no high draw in order that your opponent is indifferent to calling or folding? More than 100% of the time.

Jam your low locks.

[/ QUOTE ]

In the absence of rake considerations, I'd agree. And I certainly agree that you can (and probably should) jam if you have a scoop draw.

But let's assume that you have no draw to win (but a lock for half the pot), and rake is an issue.

If you jam 100% of the time, you are essentially making it correct for your opponent to look you up with only Aces, while maximizing the rake paid.

If you bet 80% of the time, and don't jam, it's still correct for your opponent to look you up with only Aces, but you'll lose less to the rake.

So 80% is better than 100%...


The equilibrium point, which maximizes your profit, is to bet just enough of your non-scoop-capable hands to induce your opponent to look you up with Aces. Bet more than that, and your opponent still looks you up, and you just pay more rake. Bet less than that, and your opponent would figure out that you're primarily betting your scoops on the end, and he'd be able to profitably fold Aces. My guess is that the equilibrium is around 30% of your non-scoop-capable hands -- but obviously it will vary greatly depending on how scary your board is and how likely your opponent is to have a hand better than just Aces.

[RustyBrooks]

I think your estimate is off.

Every time you bet on the end with a lock on one side, where your opponent has a marginal hand on the other side, you lose .05bb more than if you check.

Every time he calls you when you get both sides, you gain 1 extra bb

You'd have to only get one side 20x time as as often as you get both in order for it to be bad to bet. I think you scoop more than 1/20 times on the end when you have a lock on one side, no? If he never folds when you have a lock on (at least) half the pot, I think that's a very good thing.

[RustyBrooks]

Actually I think I see your point. Him calling every time is good, because sometimes you scoop, but him calling every time you scoop, and ALSO you not losing an extra .05 some of the time, is a little better.

[jbrennen]

[ QUOTE ]
Actually I think I see your point. Him calling every time is good, because sometimes you scoop, but him calling every time you scoop, and ALSO you not losing an extra .05 some of the time, is a little better.

[/ QUOTE ]

Exactly. He'll come along and pay you off on your scoop hands if there's even as little as about a 25% chance that you're betting your low-only hand. Betting your low-only hands more often than that shouldn't bring him along any more often, but it will cost both of you in extra rake.

But this is really only true in this particular situation where you have no draw to win high, a non-scary board, and rake is an issue. If any of those three conditions is not true, bet/raising is usually better than check/calling.

[Andy B]

No one is folding a pair of Aces in the scenario you describe, especially in a game this small. If the rake is not maxed out, check behind. If the rake is maxed out and the other guy can't have a low, go ahead and bet.

There are times where you might want to bet a made low with a small chance at a scoop, say with a gut-shot. There, you might be laying $.50 to win $6 or something, and that might be worth it. Your opponent might not see it that way.

[THEjDonk]

[ QUOTE ]
I'd probably start with reading as much as you can on these games... Sklansky on Razz for razz, there's also a tiny section is SS/1 on razz. There's a good section on omaha/8 in SS/2. "Seven Card Stud for Advanced Players" is good for stud and I think it covers stud/8 also? I haven't read it. There are a few books on omaha and omaha/8 out there, I think Tom McEvoy has an omaha book. I haven't read that either. Anyway, these books should offer good launching points, particularly for hand selection and play on early streets. I think every player should understand starting hand selection well before they even really venture into the game. You need more than that to win, but I don't think you can win without it.


[/ QUOTE ]

I've read most of Sklansky's books and the Super System books too. I guess it wouldn't hurt to reread some of course.

[ QUOTE ]
Not to sound harsh, but if you can't arrive at the conclusions that I did above, you might want to spend some time learning basic gambling mathematics.

[/ QUOTE ]

That seems like good advice, but I still think it sounds unnecessarily harsh [img]/images/graemlins/smirk.gif[/img]

I'm not taking it personally, though. I totally agree that it should be easy to reach those conclusions if you know a little basic mathematics.

[RustyBrooks]

[ QUOTE ]

[ QUOTE ]
Not to sound harsh, but if you can't arrive at the conclusions that I did above, you might want to spend some time learning basic gambling mathematics.

[/ QUOTE ]

That seems like good advice, but I still think it sounds unnecessarily harsh [img]/images/graemlins/smirk.gif[/img]

I'm not taking it personally, though. I totally agree that it should be easy to reach those conclusions if you know a little basic mathematics.

[/ QUOTE ]

I don't intend for it to be personal, it just sort of... is... I think it's important to be able to frame problems this way, and if you can't already do it, you should find some way to learn it. There's nothing wrong with not knowing something already.

The game-theoretical discussion in this thread is great and that's more math than most people know, including me, but an awful lot of the time you can make correct conclusions knowing nothing else but super-basic sub-algebra and a little combinatorics and probability.

Anyway, regarding books, I've read every (respected) book I can get my hands on, even some that are not really for games that I play. 7csfap is on my list and so is ray zee's book but for the moment there are some other books I *have* which I haven't finished yet, so those first. Although I don't play NLHE much any more I am part way through NLHE: Theory and Practice and Harrington's 3rd tournament book (the 2nd one improved my game so very much. I don't think it's an accident that I made 2nd in the first MTT I played after reading it when I'd mostly been muddling around going out in the middle of the field most of the previous times)