The scoop on Omaha

Now, what I can't figure out, about much, of the Omaha analyse's is this. When we talk of a scoop in Omaha, we usually talk obout a 12345and/or6 low straight scoop; and we tend to not give enough to the other scopp, the high scoop -- when the low doesn't come - and it doesn't come a lot.


In 8 or better, I look at my high hands much harder than my low hands. The low hands are pretty much automatic, but the high hands have a much larger standard deviation, I think - -lol.


But the high hands SCOOP, don't usually split or tie, and you don;t get mislead, like a heartbreaking woman - lol - i.e. you can stay out of trouble-----and you can aften have much more confidence in them much earlier which means you can sell them earlier s----------- - I would greatly appreciate a little more deference given to the high hand in Omaha 8 or better - lol.

There are two major problems with high only hands in O8.


First of all, it is a lot harder to scoop with a high only hand than you think. There are 20C3*32C2+20C4*32C1+20C5=735984 boards with no low possible, of the 52C5=2598960 total boards, meaning a board without a low occurs about 28.3% of the time. Now assuming you play for high with an average of 3 high only cards in your hand (i.e. an ace and 3 faces), your chances of a board without a low get worse since you are holding quite a few of the high cards -- 17C3*31C2+17C4*31C1+17C5/48C5=0.231 or 23.1%. Since most of these high boards will contain a flush or straight you have no part of, your chance of making a strong high hand you can confidently bet are probably something less than 10% of the time you are dealt a decent high only hand containing an ace and 3 faces. Combine this with how rare it is to be dealt a decent high only hand (20C4/52C4=0.018, or a 1.8% chance of being dealt a hand with no card lower than ten, and some of these are dogs like QQQJ or AJTT), and you will not be making many high only scoops.


The second problem with scooping high only is that many of the flops you like contain only 1 low card, which kills your action from low hands and means the pots you do manage to scoop are smaller.

Great analysis. The best scopper hands are low hands with a pair or an ace high flush draw. They can take heat (when they flop well) and apply it. Plus they win the biggest pots. Hands such as KKQJ just don't get much WELCOME postflop action.


IMO, flawed high only hands (e.g., KQTT) should be played in back for one bet. And any high only hand that contains a nine usually should not be played for a full bet. But then again, I may be playing too tight ;-).


Regards,


Rick

Just to be a picky bastard, I don't think you've included a few other boards with no low possible -- the boards where there are three or more low cards, but low pair(s) make the low impossible.


So add in the boards with a low pair, another low card, and two high cards:


8*C(4,2) * 28 * C(20,2) = 255,360


and the boards with trips and two high cards:


8*C(4,3) * C(20,2) = 6,080


and the boards with two pair and a high card:


C(8,2) * C(4,2) * C(4,2) * 20 = 20,160


and the boards with trips, a low card, and a high card:


8 * C(4,3) * 28 * 20 = 12,800


and the boards with a small full house:


8 * C(4,3) * 7 * C(4,2) = 1,344


and the boards with small quads and any other card:


13 * 48 = 624


All this adds up to 296,368 more high-only boards, so (if your previous computation is correct, which it looks like it should be), there are now


735984 + 296368 = 1,032,352


total boards where no low is possible, or 39.7%.


But now to be an even pickier bastard, we should really consider that, when you're looking at your hand that is presumably chock-full of high cards, that those 20s up there should really be 16s, 17s, or 18s, depending on whether your hand contains zero,

one, or two aces.


The math is going to get messy so I'll do it in a separate note either now or soon, depending on whether this meeting for which I'm on call happens soon or not /images/smile.gif


--JMike

If you have zero aces, the numbers are:


Five high cards:


C(16,5) = 4,368


Four high cards and one low card:


C(16,4) * 32 = 58,240


Three high cards and two low cards:


C(16,3) * C(32,2) = 277,760


A low pair, another low card, and two high cards:


8*C(4,2) * 28 * C(16,2) = 161,280


and the boards with low trips and two high cards:


8*C(4,3) * C(16,2) = 3,840


and the boards with two low pair and a high card:


C(8,2) * C(4,2) * C(4,2) * 16 = 16,128


and the boards with low trips, a low card, and a high card:


8 * C(4,3) * 28 * 16 = 14,336


and the boards with a full house, low over low:


8 * C(4,3) * 7 * C(4,2) = 1,344


and the boards with low quads and any other card:


8 * 44 = 352


48 cards = 1,712,304 total flops


So the probability of a board with no low possible is 31.4%


--------


If you have one ace, the numbers are:


Five high cards:


C(17,5) = 6,188


Four high cards and one low card:


C(17,4) * 31 = 73,780


Three high cards and two low cards:


C(17,3) * C(31,2) = 316,200


A low non-ace pair, another low card, and two high cards:


7*C(4,2) * 27 * C(17,2) = 154,224


A pair of aces, another low card, and two high cards:


C(3,2) * 28 * C(17,2) = 11,424


Low (non-ace) trips and two high cards:


7*C(4,3) * C(17,2) = 3,808


Two low non-ace pairs and a high card:


C(7,2) * C(4,2) * C(4,2) * 17 = 12,858


A pair of aces, a low pair, and a high card:


C(4,2) * 7 * C(4,2) * 17 = 4,284


Low non-ace trips, a low card, and a high card:


7 * C(4,3) * 27 * 17 = 12,852


Three aces, a low card, and a high card:


28 * 17 = 476


A full house, low over low:


7 * C(4,3) * 6 * C(4,2) = 1,008


A full house, low over aces:


7 * C(4,3) * C(3,2) = 84


A full house, aces over low:


7 * C(4,2) = 42


Low quads and any other card:


7 * 44 = 308


48 cards = 1,712,304 total flops


So the probability of a board with no low possible is 34.9%


--------


If you have two aces, the numbers are:


Five high cards:


C(18,5) = 8,568


Four high cards and one low card:


C(18,4) * 30 = 91,800


Three high cards and two low cards:


C(18,3) * C(30,2) = 354,960


A low non-ace pair, another low card, and two high cards:


7*C(4,2) * 26 * C(18,2) = 167,076


A pair of aces, another low card, and two high cards:


28 * C(18,2) = 4,284


Low (non-ace) trips and two high cards:


7*C(4,3) * C(18,2) = 6,426


Two low non-ace pairs and a high card:


C(7,2) * C(4,2) * C(4,2) * 18 = 13,608


A pair of aces, a low pair, and a high card:


7 * C(4,2) * 18 = 756


Low non-ace trips, a low card, and a high card:


7 * C(4,3) * 26 * 18 = 19,656


A full house, low over low:


7 * C(4,3) * 6 * C(4,2) = 1,008


A full house, low over aces:


7 * C(4,3) = 28


Low quads and any other card:


7 * 44 = 308


48 cards = 1,712,304 total flops


So the probability of a board with no low possible is 39%


--JMike

Nothing wrong with being picky when you are right! /images/wink.gif


With the other boards you mention included, it's still devilishly hard to both get dealt a good high only hand and complete it with no low possible, but you already knew that which is why you called it being "picky" I suppose.

Yeah, I definitely agree that high-only hands should be evaluated with caution, because of the smaller-pots argument and the low-board probabilities.


--JMike

There is a much simpler way to calculate the number of times there is no low. Just flop a bunch of hands and it will be clear very quickly the number of times there is no low. It does not take a complicated statistical analysis. This statistic is right in the same ball park with flipping a coin.


This is a real easy statistic to determine, so someone must have it exactly. I once did about 90 hands and got a percentage of about 40%. If my sample is right then I am sure low does not scoop the pot 40% of the time.


Additionally, often when you are playing in a tight game where almost everyone is predominantly playing low hands you sort of have the high arena to yourself and you can make it real hard on the low players by making it very expensive for them to draw.


When I am in this situation I raise constantly, so the lows cannot draw cheaply which causes great consternation and confusion (many players hate this and almost act like I am being unfair - lol). Sort of like we should all be ladies and gentlemen about these raises - lol. I am trying to force them to play my game which makes it difficult for rocks who play by a sort of formula i.e. look at the flop cheaply and then fold or continue. But if they have to make two or three bets to look at the flop then they are in trouble. This approach has always worked well for me?


I play the lows just like everyone else, except that I raise those more often before the flop as well, just to create more confusion in the game and so they don't know if I am high or low which allows me to get more calls with my high hands and to make it hard on the rocks who like to look at the flop, as mentioned, as cheaply as possible. I am talking about limit omaha here.


P.S. otherwise I play solid (no danglers). I just raise more often than conventional wisdom might dictate.

Pretty close.


I get 1,037,232/2,598,960 = 0.3991 if you are not taking your own cards into account.


"High" has a double meaning here. There are high cards and there are hands that win for high, and they are not the same thing.


Buzz

Good: I only used 90 hands (lazy - lol), but I knew 90 hands had a good chance of being pretty close i.e if you flipped a coin 90 times. And I do think there is room for further thought regarding how the high should be played in relation to the overall game.


I think we have sort of been mesmerized by the low end of the game because it is pretty obvious. But what seems to be sort of poetic is that however/whomever that game got started i think they got it pretty much correct, which means they had to take into account the proper relationship between high and low.

Charles - The two cards from your hand that win for high are not necessarily high cards, although I'll agree higher cards have a better chance of winning for high than lower cards. Thus A-2-3-4 can win for high about as easily as K-Q-J-T or 9-8-7-6. If you win the high side with either A-2-3-4 or K-Q-J-T, you don't have to split the pot with someone as often as you do with 9-8-7-6.


Thus, with either A-2-3-4 or K-Q-J-T you have a better chance of scooping, than with 9-8-7-6. You should want to scoop in high/low split games, rather than play for either high or low.


A-2-3-4 is better, IMHO, than K-Q-J-T, simply because there are more ways to fit with the flop. On about three out of five deals, when you miss hitting the flop for high, you have some backup chance of salvaging something by winning for low.


If you're thinking of A-2-3-4 just as a low hand, you're not thinking in the best way, IMHO.


I'll agree that looking at starting hands and ascertaining which of them has a good chance of winning the low half of the pot is easier than looking at starting hands and determining which of them has a good chance of winning the high half of the pot. But to do either is missing the point; you want to be playing to scoop, not playing to win either high or low.


Try this: Take a deck of cards and deal out eight hands. You'll have twenty cards left over, enough for four boards. Now deal out the four boards. You'll probably see that a different hand (of the eight) will make the best high with each different board. Better for you if you don't misunderstand here. I don't mean you should see the flop with every hand you are dealt. Far from it. But deal out those eight hands and four boards a few times and you will clearly see the difficulty of deciding ahead of time which of the eight hands will win for high (although you often can pick some likely losers in the eight hands).


I agree with you that Omaha-8 is a beautifully designed game.


Just my opinion.


Buzz

when you say that 9876 has to split the pot more often than A234 or KQJT, I am not sure what you are saying?? are you referring to a tie between high hands, or that ther will not be a low, or what???


since getting quartered seems to happen rather often, seems to me that a high hand of A234 would likewise be in a tie more often than 9876, or KQJT....???

I understand the power of 1234 and am as happy as the next guy to get it. I would never play 9876 unless I could see the flop free from the big blind.


Of the 40% of the time low does not get there one has to also add to the equation the NUMBER of times the players don't get there and the number of times the low is split.


Learning to play the high hands more correctly, in my opinion does a few other things. They can disguise your good low hands and they can be used to put pressure on a game that is tight and passive (player's looking for the nuts on the flop cheaply) by allowing you to raise every time you get either a premium high or low hand without giving up information, confusing the game and putting big time pressure on tight players by making them put in two bets to see the flop which also can cause players to go on tilt from frustration.


A few years back I beat the Omaha game at the Mirage every night for about a week using this tactic; and they were players that played every day. The constant raising caused many of them to come unglued - lol. In fact by the end of the week many of the players were pretty po'd with me for disrupting their meal ticket. Can't say as I blame them - lol. Away from the poker table I am a mild mannered person - lol.


I would not use this tactic in pot limit as I would also want to see that flop cheaply.

Hi Jellow -


"when you say that 9876 has to split the pot more often than A234 or KQJT, I am not sure what you are saying?? are you referring to a tie between high hands, or that ther will not be a low, or what???"


9876 has a poor chance of winning for low. but if the board has a few middle cards and/or certain low cards, (eg. 866, T76, or 765) then 9876 has a good chance of winning for high. Usually when 9876 wins for high, it has to split with low.


A234 with corresponding matching cards on the board (eg. 3AA or 52A) has a shot at high and also low. When it wins for high it usually also wins for low.


KQJT with corresponding matching cards on the board (eg. QTT or JT9) has a good shot at high. When it wins for high there usually is no low with which to contend.


Thus A234 and KQJT both have a better chance of scooping than 9876.


O.K.?


"since getting quartered seems to happen rather often, seems to me that a high hand of A234 would likewise be in a tie more often than 9876, or KQJT....??"


Actually A234 and KQJT both are tied by a similar hand about one third of the time in a full, loose game when they win for high with a straight.


A234 is also tied by a similar hand about one third of the time in a full, loose game when it wins for low. You notice the tie more often with A234 than with KQJT simply because A234 wins much more often (for low, high, or both) than KQJT wins for high.


When A234 is tied for low it only gets half of the low half of the pot. (Half of half is a quarter).


Hope that makes it clear to you.


Buzz

Charles - Omaha-8 is a very dull game with nobody providing the action. I don't like dull games much. Sounds like you provide lots of action.


"Learning to play the high hands more correctly, in my opinion does a few other things. They can disguise your good low hands and they can be used to put pressure on a game that is tight and passive (player's looking for the nuts on the flop cheaply) by allowing you to raise every time you get either a premium high or low hand without giving up information, confusing the game and putting big time pressure on tight players by making them put in two bets to see the flop which also can cause players to go on tilt from frustration."


You have my attention. What high hands?


Buzz

Hmmmmm, being primarily a holdem, stud and 8 or better stud player who only infrequently dips into omaha 8, I probably don't have the experience many of you have. But I feel that some key elements to this argument are being missed.


If the math you're using is correct, and it seems acurate for these purposes, boards containing no possible low hand coming up 40% of the time indicate a huge area to be exploited. That's almost half the time! True, the pots are smaller. But 100% of a 60% as large pot s better than 1/2 or 1/4 of a average sized pot.


The key here, as in all forms of poker is to play well AFTER the flop, since most weak players don't. Hands like 10,J,Q,K/2 or 3 suited, or K,K,Q,J 2 suited can be great PROFITABLE hands if played correctly. And that's what poker is about...profit.


Getting in cheap, in good position and then playing extremely well post flop, ie: only staying with multiple out, nut draws to straights and flushes and carefully playing only 2 pair (having top pair is criticle) and sets (carefully), these hands can be played for profit.


I've been dealing for a long time as well as playing and the players who beat omaha 8 on a regular basis are playing high only and hi/lo hands. What they don't do is play low only hands. That's the area where many omaha 8 players get into trouble. Hands like 2,3,7,8 /3 suited I see played frequently because of the low potential, but this has got to be a losing proposition in the long run as you're playing primarily only for low. 1/4's and 1/2's and 2nd nut lows kill these hands profit potential.


It all comes down to not waisting bets. A well schooled un-diciplined player will get creamed by a average experienced player who plays well after the flop. It's not how much you win in this game, we all know how to play the nut hands...it's how much you don't lose that really matters.


IMO, for a different view. Keep playing hard!

Glen - Please, if you have an inkling, do tell us what to exploit. I agree K-K-Q-J-double-suited is a good hand, but not so good, at least in a loose game, when the board is (2-2-Q)-8-8. (Notice no low possible).


Buzz

Good post Glen and my point exactly, I think - lol. And we all know different games have different textures that require different play. For example: Malmuth and Sklansky outline one of the most important counterintuitive realities when they talk about short play. In short play the loose raiser will do better than the tight player and there are certain statistical realities that show how this works.


Even better are the examples that show that bluffing with a drawing hand that is an arithmetical loser to a pat hand: e.g. standing pat on a 97 for low against a 1 card draw to the wheel can actually be turned around with a certain amount of bluffs and the opposition can't even do anything about it - they cannot win - this is the sort of spooky stuff quantum physics is made out of - lol.


Now to the point: If you are in a game of hi-low omaha and the table is sort of playing together (not raising before the flop) to bleed the unwary with pat low hands) you can blow this game right out of the water by playing aggressively

with good high hands and good low hands - because they just are not really concentrating on the high hands which also allows you to lower your standards a bit for the high hands.


In poker play is relative to the make up of the game. Last, some games are best to just leave - lol - because there is nothing you can do with them, so why even play in them. Well Doyle and chan probably can, but I am going to leave them to them - lol.

Buzz,


Re-read my post and you'll see that the key is to play well post flop. When the flop comes 2-2-Q and I hold K,K,Q,J... simple. you muck. It is not only likely someone started with a duece, it would be exteremly unlikely that there wasn't one out there somewhere.


So playing this hand for 1 bet preflop, hopefully in later position can be profitable when you get a favorable flop, and there ARE many flops that would work for this hand.


Don't forget that one variation of this game is omaha hi only. If we consider the 40% of the time that no low is possible and play the hands with omaha hi only in mind for those hands suited for that, we can find opportunities to find extra chips.


Finally, I'm saying that we, as good players, should exploit the opportunity to play hands that our opponents might not play because they are so "low" minded. And with almost half the hands offering no low, there are obviously some hands that we can find that they don't. 'Nuff said.