The Secret of Limit Omaha High-Low Tournaments

Steve BadgerThe secret of getting ahead is getting started. -- Mark Twain

For years people asked me if I ever held anything back when writing about poker strategy. The answer is, yes, one thing. Here it is.

Prior to retiring from tournament poker in 2008, I won more Limit Omaha High-Low tournaments than any other player to that point. I don't say that as a way to suggest my opinion on Omaha strategy is therefore going to be "right". It's just a fact for any relative newbie to be aware of when considering the opinions voiced here. Doing as well as I did was due to more than one factor of course, many of which had nothing at all to do with the cards, but the key card-playing philosophy I followed can be boiled down to four simple words, which I call "the secret of Omaha".

Actually the secret is more micro-specialized than that, as in: "The secret of Limit Omaha High-Low tournaments." However, the secret also largely applies to Limit High-Low ring games, and somewhat applies to Pot Limit High-Low too.

The secret is: play five-card hands.

Or more accurately, "focus on playing five-card hands." This is a basic concept, but few players actually follow it.

I'll answer the simple question first: "how the heck can you have a five-card hand in a four-card game?"

Easy. Aces play for high and low, so hands like AK32 have two high/Broadway cards and three low cards: 2+3 = 5

Not only do aces play both high and low, they are the highest high card and lowest low card! Since aces have this flexibility that other cards don't, voluntarily entering any pot without at least one ace is illogical on its face. Sure, sometimes in tournaments if you have five chips total, and four of them are in the big blind, you normally should put your last chip in with anything, but these very rare exceptions are trivial.

A secondary strategic concept is you should normally try to play pots where you isolate yourself against one opponent when you have position. Obviously you can't force everybody else to fold, and sometimes you get great hands out of position, but your general aim should be to contrive situations where: you play a five-card hand, in position, against one opponent. (This goes hand-in-hand with the parallel concept of raising or reraising to drive the blinds out to split up their equity, or to get them to put in extra bets with inferior hands.)

Some four-card hands, like AKQJ, are also clearly playable, but their strength is limited by their limited flexibility. Five-card hands offer more "escape hatches" than four-card hands do. The phrase "emergency low" is common in Omaha, where a five-card hand like AKK8 can play out like K♠Q♠2♢7♢4♠. The AKK8 here may lose high to a flush hand like A♠T♠9♢2♡, but backs into a lousy low that wins. A half pot win is a half pot win, whether it is the nut low or a garbage one. Chips are chips. In contrast, AKKQ does not have that emergency low option, and is thus limited in the ways it can not lose. While we want to win pots in all situations, in tournaments "not losing" chips has a value unto itself, since we can't just reach into our pocket and buy more chips whenever we please.

While four-card hands like AKKQ are still be playable, "pretty" fish-bait like 23TT is muckable garbage. Let's compare a not-exceptional, not-pretty five-card hand like AJ64 versus 23TT, headup with no flush suits. The equity on the two hands is close, but equity is not a good measure of high-low hand value. The AJ64 scoops about 7.7% times more than the 23TT does (377,603 vs. 349,956). Not a massive edge, but not a trivial difference either. Next, consider that when the 23TT garbage hand wins high in a simulation, it will often be on boards like 35K8Q. How would you like to call-call-call with 23TT on a board like that? And, since to get its full value the 23TT must call 35K8Q boards, it must also call 35J8Q boards and get scooped. In contrast, when 23TT makes the nut low, the AJ64 makes a pair of aces and a weak low, making for easier calls. (Sometimes the AJ64 will also have a tough call-call-call to extract its full equity, but these instances occur far less often than fish-bait four card hands.)

When considering your low chances, what would you prefer, having three low cards or two?  Hands like 23TT and 23KK look soooo pretty and easy to play to weak players. They think the hand has "flexibility" because if it comes an Ace that kills their pair, they will now have a low. This is like thinking "if I lose my left arm, I can still use my right". Instead, how about playing a hand that makes something both ways when an ace comes?

But the 23TT garbage is revealed as even worse when faced with a more solid hand, like AJ53. Now the AJ53 scoops about 46.5% more often than the 23TT does. This is a slaughter. This is approaching a 3-2 edge... and that is without even adding the advantages in suited-ness that AJ (or AQ, AK, AT) have over (non-ace) big pair hands. As we can see, a general rule of thumb is a five-card ace hand has about a 50% edge over seemingly similar ace-less hands when they play each other head-up. (Again, split pot equity is irrelevant head-up, as both hands win when they split... if two hands split 99 times out of 100, the equity on the two hands will be very close, but the hand that scoops one-out-of-100 is literally infinitely better.)

Let's compare KKQQ versus AT53, the best ace-less four-card high hand versus a good but not great five-card hand. It's still a big edge... the five-card hand scoops 20%+ more. And again the bet-ability kills the four-card hand's actual value, since to get its full simulation percentage KKQQ is going to have to call-call-call A769K boards.

At least the KKQQ has three solid two-card holdings: KK, QQ and KQ. Take a look at the 23TT in contrast. It has a weak 23 holding (the biggest sucker hand in Limit Omaha High-Low) and a weak TT holding, against which any AK, AQ or AJ in a five-card hand is that proverbial coin-flip we see in Hold'em games. Its other holdings are the nearly worthless T3 and T2. In contrast, the five-card hand of AJ53 has AJ, A3, and 53, plus A5, and the weak J5 and J3.

The disparity between four-card hands and five-card ones becomes even larger when the quality of the five-card hand increases, like A♠K♡3♠4♡ versus Q♠2♠Q♢3♢, or more overwhelmingly versus a double-scheckie like Q♠2♠3♢5♢. (And that doesn't even get into how dominating a two-ace hand like AA67 is over any four-card hand.)

The value of five-card Omaha High-Low hands is usually not close to four-card hands... yet, a lot of players still like to play four-card hands; they still even re-raise to isolate with their four-card hands!

(Remember, we are talking mostly about tournaments here, where play is often head-up. In ring games where action commonly can be five-way+, some more speculative hands can be played.)

Here is an example of where a four-card hand value is close to a five-card one: AT75 versus KK32. When both hands are offsuit, the KK32 scoops about 13% more often. But when the AT75 has a suit, it scoops 2.5% more often. These two examples illustrate a few points. One is that even just about the best four-card hand is not greatly better than one of the crappiest five-card hands. Another is adding a suit to the AT75 guts a lot of the value of the bare KKxx hand, flipping the ace hand from a middling underdog to a small favorite just from that seemingly minor switch. (Since small changes in hands can turn favorites into underdogs, it is best to focus on playing types of hands that can defend themselves when played against a wide variety of opponent hand types.) Third, and most important, look at the link in this paragraph again. Notice AT75 wins low about 28% more than the KK32 does. Most weak players look at the 23 versus A57 fight as the 23 has the better low draw. In reality, the two lower cards make the better low much less often. This again reflects the value of playing a five-card hand with three low cards... having that extra low card matters. (Turning KK32 into 8832 or 4432 to give it a third low card just makes AT75 more profitable.) In a tournament, your first aim is not to make nut low; it is to not be eliminated. Having the counterfeit protection of three low cards allows a player both more flexibility and a better chance to survive.

A hand doesn't have to be pretty; it just has to not lose.

I could go on and on, but if a game is designed to give a large bonus value (remember, aces don't just play high and low, they are the best high and low cards) to four cards in the deck, why play any hand that doesn't include an ace? And, in a game where you are dealt four cards, why would you commonly fold any five-card hand?

It's not news to any good Limit Omaha High-Low player that playing hands with aces is smart, but most don't accept that playing hands without an ace is bad, and toxic in tournaments. (As we see from the above simulations, an exception can be sometimes calling head-up in the big blind with quality ace-less hands, where getting 3.5-1 on your money can make these speculative hands playable for a bet... and then if they flop decently, go from there.)

Steve BadgerSome extra points... 1) the five-card hand effect is even stronger multi-way (except as noted in #4 below). It is harder to definitively simulate examples multi-way, but easier to summarize: if two or three opponents are playing five-card hands, why would anyone choose to play a four-card hand? Again, some four-card hands with an ace like AKQJ are plenty strong enough to fight multi-way, but ace-less hands very, very seldom are.

2) Obviously all five-card hands are not equal. I'd rather have AK32 than AKQ2, and would rather have AKQ5 rather than AQ75, but you can't just switch your cards. In tournaments especially, you have to play with what you are dealt. You can't fold until you get A♠A♡2♠3♡. When a five card hand shows up, you should normally be ready to play... especially at the later stages in a tournament when you and/or your opponents have small stacks of chips relative to the blinds, and head-up play becomes the norm.

3) Many weaker players like four-card hands like KQ32 or K432 because they are "easy to get away from". That attitude is disastrous in a tournament structure, where you don't have a nearly infinite bankroll like in a ring game. You simply can't be a winning tournament player and have the strategy of playing big underdog hands and hope to get lucky. There should be no easily disposed of soldiers in your precious chip stack.

4) A standard disclaimer with any poker game strategy is that the worse your opponents are, the more hands you can play. And the texture of the game you are in impacts your choices. Some of the weakest, unsuited five-card hands are better folded in some family pot limpfests where the winning hand will likely need to make the nuts. The basic idea of playing a five-card hand is to make enough of a hand to not lose... if nine people are seeing the flop, you need to be playing hands that showdown the nuts. Omaha tournaments almost always switch from having multi-way pots at the beginning to being mostly head-up toward the end. The proper focus is always to want to play five card hands, but the later you go in tournaments, the looser you should be in terms of entering the pot first with any five-card hand. Again, make the nuts early; make do later.

5) Finally, once you get down to head-up or three-handed in a tournament, any ace is gold, and many ace-less four-card hands become playable, especially from the button or big blind.

It's not a minor point to say it is foolish to play four-card hands in a game where five-card hands are an option. It's absolutely essential. Poker is about edge, finding advantages where you can. Playing when you have an extra card, and not playing when you don't have an extra card, is to seize an advantage and avoid a disadvantage.

Some bloated ego players think they are so clever that they can turn crap starting hands into gold. But imagine playing five-card hands in an Omaha tournament as like a playoff basketball game... where one team plays five players while the other team only plays four. A team of five journeymen pros would beat a team of four stars (and obviously a team of five stars would wipe out a team of only four journeymen).

You shouldn't leave home without your phone, car keys and other essentials. Likewise, don't be getting involved in Limit Omaha High-Low pots without the essential of an ace, and learn to especially appreciate those hands where the ace you have gives you a five-card hand in a four-card game.

See also The Myth That Omaha Hand Values Run Close Together