Omaha High-Low Poker Strategy

Introduction to Omaha High-Low Split

Steve BadgerPeople ask me what book I'd recommend to a novice Omaha player. There are other useful books, but my normal reply is: the Bible. Omaha has the tendency to drive beginning players to prayer, but it really shouldn't.

I am also often asked about writing my own book on Omaha. This is not a book. It is not meant to deal with all the advanced and difficult skills that the strongest Omaha players master. This is an introduction to the key strategies behind the game. While it's not meant to deal with the most complex concepts, it does deal with concepts that should benefit many experienced players too, not just novices.

What I mean by "Omaha" here is: Limit Omaha High-Low (aka Omaha8, Omaha Hi-Lo Split, Omaha Eight-or-Better). Omaha is also played Limit High Only, Pot Limit High, Pot Limit High-Low and occasionally No Limit. While concepts here are sometimes applicable to the other variations, sometimes they definitely are not. Check out the several other articles linked on the Omaha Poker Tips page for strategy ideas on the other variations. Also check out Omaha Myths, which deals with common misconceptions about the game.

In general, in all forms of Omaha, players who treat the game as a party are dominated by players who treat the game as business. Optimists enjoy Omaha; realists dominate Omaha. Players exercising mathematical realism, discipline, adaptability and creativity get the money from players out to have fun and gamble to get lucky.

Two cards, always two cards... Omaha hands consist of three of the five community board cards, plus two cards from each player's hand -- always three off the board, always two out of the hand. You can use the same or different card combinations to make your high hand and your low hand (if any), but you always use two from your hand, three from the board. This is important not just from the perspective that it is a rule and you have to do it, but also in thinking about how your hand must integrate with the board. Your hand must cooperate with the board. (Cooperation is a recurrent Omaha principle.) You should never think of your hand in isolation. It needs three cards from the board for high, and needs three cards for low. (Some new players find it helpful to focus more on "three from the board" rather than "two from the hand.")

Nut low means best possible low... Reading low hands often confuses newbie players -- experienced ones too -- but there is an easy way to do it. First, you must remember the two cards from your hand, three from the board rule. A board like 87532 might make 2367 somewhat hard to read but you read your low hand simply by taking the lowest card combination to be found using three cards from the board and two from your hand.

But what is the lowest? What about when your cards are paired (counterfeited) on the board? Think of it this way: the lowest/best possible hand is a wheel, a 54321 -- or 54,321. The highest/worst possible qualifying low hand is 87654 -- or 87,654. Read your low hand as a number, starting with the highest card and working down. The player with the hand/number closest to 54,321 wins (or ties if someone else has the same hand/number). Omaha players often speak of "the nut low." This is the best possible low in this particular hand. While A2 combined with an 876KQ board creates the best low possible, 54 combined with a board of A23KQ makes the nut low in another case. And, 23 combined with a 764KA board makes the nut low (64,321), not an A2, which only can make a 76,421. If you get confused by how your cards are paired or counterfeited by the board, at the showdown, show your hand and ask the dealer to read exactly what your low hand is.

Omaha is a game of nut hands, so as hands unfold, practice reading what the nut low hand is. Then start thinking of your low hand in relation to the nut low. It's not important to know how low your low is, what matters is how low your low is in comparison to the nut low.

Why play Omaha?... While some newbies reading this Introduction will be hard pressed to do it right away, the aim is to win at Omaha -- not have fun, or even to irritate yourself. Frankly, at lower limits, winning at Omaha is easy, if you really are trying to win because most Omaha players play terribly, much worse than they play Hold'em (which is not so good to start with).

In many ways, lower limit Omaha is mathematically simplistic. If you play only good starting hands and your opponents see fit to play almost every hand, and don't care whether they play for one bet or four, soon the math of that will work in your favor. Omaha is a great game to make money if you have a small bankroll. $3/6 Omaha should require less of a bankroll for a sensible player than $3/6 Limit Hold'em, but generate a higher hourly win rate.

Bad players have virtually no chance to beat Omaha over any meaningful period of time, but they can win big pots, and have really good sessions. This is true of Hold'em too but to a much smaller degree, because Hold'em edges are generally small in loose games. Weak Hold'em players can "school" together and get pot odds on their poor draws and therefore not be playing all that bad. On the other hand, there is no parallel schooling phenomenon in Omaha where very often five players draw stone cold dead while two players have all the outs between them (for example, on the turn the nut flush and the top set are the only live hands, and five other players with two pairs and baby flushes are drawing dead).

Loose game Limit Omaha is a game of massive edges; loose game Limit Hold'em is a game of smallish edges. Low limit Omaha games are the easiest poker games to beat -- if you play properly. Most players do not have the ability, or more important, the desire to play properly in low limit Omaha games. If you are playing to win, generally Omaha games are the place to play because they are cheaper (less bankroll), more profitable (higher hourly win rates) and have weaker players playing much more poorly. It's deadly dull tho. What winning loose-game Omaha is not is a barrel of laughs.

Boring PokerSo, for less experienced players, there are some contradictions at work here. Omaha is a great game for good players... but most inexperienced players are not good... but it is very easy to teach a player to play way-above-average Omaha... but the basic advice is to play with great discipline... but having discipline is an advanced skill... and is boring as paste.

Omaha is a game of non-random accuracy... One thing to understand about Omaha is that since you get a higher percentage of your final hand sooner, your hands are generally much more defined than in Hold'em or Stud. After all, 7/9ths of your hand is known on the flop. Then, when it comes to the betting, the outcome of an Omaha hand is often precisely known. A player that can count twenty, or ten, or four outs to the nut hand often has exactly that many outs to win.

In Hold'em random outcomes are common. Facing several opponents, they can win by hitting oddball kickers or spiking an underpair. On the other hand, Omaha is far more concrete. You know often your precise outs -- how many cards make you the nut hand. In loose games there is very little mystery. In tighter games you often don't need to make nut hands to win, since you face fewer opponents, but in lower limit situations, there is usually little randomness to the game. Unlike Hold'em, before the river card is dealt, usually you should know exactly how many possible cards make you the winner, and how many don't.

Omaha is a game of information. Hold'em is a game of uncertainty. That's how they were designed! Loose game Omaha is about ending up with the nuts. Loose game Hold'em is far more shadowy and difficult.

Many players seem to draw the wrong conclusions from the greater certainty that is part of Omaha. They think because their nut flush on the turn gets beaten on the river when the board pairs that Omaha has some mystical randomness to it. The opposite is true. There are a precise number of cards that pair the board, and make you lose. There are a precise number that do not pair the board, and make you win. On the turn, if you have the nut flush, with no cards in your hand paired on the board, and your opponent has a set, with no other cards paired on the board, there are exactly forty possible river cards. Exactly ten pair the board to make you a loser. Exactly thirty do not pair the board and make you the winner. That's it -- pure, simplistic math. In the long run, you win three out of four. This is known. This is Omaha.

Do not let yourself be confused by irrelevant concepts. What matters in any form of poker, but particularly in Omaha, is the probability of winning -- not who is temporarily in the lead. Whether you flop a made hand or a draw or a backdoor draw is irrelevant, what matters are your prospects, your probabilities, of having the winning hand on the river. What counts is how many cards, in what combinations, make you the winning hand. Know how many cards make your hand, and then know that in the long run you will win pots in the mathematically appropriate percentage: if you have x% chance of making the winning hand, you better be getting at least the correspondingly appropriate pot odds.

Omaha is a game of accuracy, clarity and concrete information. Sure, sometimes you get unlucky, and since Omaha edges are so huge, when you get unlucky it can be hard to swallow, but since the edges are usually so big, if you play good starting hands in Omaha, and get unlucky, you can still win. You just have to keep your discipline.

Starting hands... Unlike Hold'em, where post-flop play is far more critical, winning Limit Omaha High-Low is fundamentally rooted in starting hands. Starting hands exist before the flop, which is where you get enormous edges in Omaha against a field. On the turn you will often have times where some players are even drawing dead, and that is clearly the juiciest money in the game, but the simplest, most direct, most necessary way to beat these games is to not play crap hands and to get more money in the pot when you have A255 and several of your opponents have hands like K965. Getting garbage hands with a low winning expectation to pay as much as you can before the flop when they are large dogs is a big part of winning Omaha.

Not counting AA and perhaps KK, in looser, multiway games, Limit Texas Hold'em hands run much closer in value than Limit Omaha High-Low hands do -- regardless of what urban myths claim. If you don't know and appreciate this basic concept, you are going to be in trouble in Omaha. In multiway pots, Omaha has a fairly large group of hands that will win at double the rate of randomish hands. Few Hold'em hands can say the same. Only playing good starting hands, and raising before the flop with most of them, is the basic path to of winning.

Omaha PokerSchooling in Omaha... "Schooling" is a common phenomenon in loose-game Hold'em. When several players play badly by calling with weak draws, like gutshot straights or backdoor flushes, these players partially protect each other by making the "price" on each of their calls better. If only one player calls with a gutshot draw, usually that is a significant mistake, but if several players make similar calls, now the pot is big enough to make the calls profitable, or at least less bad. Properly understanding the strategy involved in schooling is a key skill in loose-game Hold'em. (See Hold'em schooling.)

There is no parallel schooling phenomenon in Omaha -- quite the contrary. In Omaha, schooling benefits the favorites, not the underdogs. This reverse schooling phenomenon is what makes Omaha often mindlessly profitable. Players with four outs or less call bets from players with twenty outs, and no matter how many people call, the twenty outs player continues to have twenty outs. Despite the definite reverse profitability of "schooling" in Omaha, poor players engage in it all the time. They look at a big pot and call bets hoping to get lucky, even though they may be drawing totally dead.

Suppose you flop a top set of three kings against seven opponents. The true enemies of your KKK (or any strong Omaha hand) are the first two callers (meaning the two opponents with the most outs). On a flop of KsQd7c for example, we are afraid of AJTx wrap-straight draws. That's the first caller or two. Then we have open-end straight draws. We are the favorite over those (and all the rest of the draws). Next are backdoor flush draws. Then we worry about the lame backdoor straight draws around the seven. Naturally, many of these longshot draws overlap each other. For instance, if the Ace-high spade flush draw calls us, we certainly love the five-high spade flush draw to call, drawing dead. Yes, they may win sometimes, but we love these sixth, seventh, and eighth callers!

With the KKK, if we assume we won't win unless we fill up, and we don't fill up on the turn, we will have ten outs of the forty-four possible cards, meaning we will fill up 23% of the time. Even if we lose to quads the 3% part of that, that's still a one out of five win percentage, for a scoop, while getting six, seven or eight way action. Additionally, we'll normally have our own backdoor draws. If we have two backdoor King-high flush draws, this will further destroy what little power the sixth, seventh and eight callers have, as their backdoor baby flush draws in our suits are contributing totally dead money on that aspect of their hands.

So, building a pot with a raise before the flop in Omaha does not benefit schooling opponents, it benefits players with good hands that are more likely to make nut values. The flip side of this phenomenon exposes another key difference between Omaha and Hold'em.

In loose Hold'em games, there are a lot of hands you can profitably add to your arsenal, most obviously Ace-rag suited and suited connectors. This is not true in Omaha. Again, the difference in value of hands multiway in Omaha is much more dramatic than in Hold'em. The majority of hands simply are never playable (outside the blinds). If you are on the button and everybody limps in, 3456 is still a worthless piece of garbage. It does not matter if you have three opponents or seven, the hand stinks. You can play a small number of additional hands, but for the most part, no matter how loose your opponents are, you can't add many more hands to your playable repertoire.

The thing to "loosen up" in such a game is to want to play for a raise most every hand you play. In tight games, calling when someone limps in front of you is sometimes the right play. In a loose game, raising is usually the correct play because you are playing a hand with way the best of it. You want dead money in the pot, and you want dead hands hopelessly chasing it! And they will.

If you build it, they will come... drawing dead.

A "river" game?... Some people call Omaha a "river game" because the last card often determines the winning hand. While that is true, the thinking behind this "river game" idea is very flawed. Poor Omaha players wait to the river to bet -- when they know they are going to win (or lose). That's just not sensible or profitable. Omaha is not a "river game"; it is a game of preparation.

Before the flop: you should play hands that have a high expectation; you should manipulate the pot size; you should try to manipulate your opponents so that when you have a hand that plays well against fewer opponents you are playing against fewer opponents and when you have a hand that plays well against a full field you are playing against a full field.

After the flop: the flop is critical. Here you should roughly calculate your various probabilities and deduce how favorable your chances are to win. Again, here a player should be manipulating the pot -- get more chips in when the odds favor you, try to minimize when you have a longer shot.

The turn card is the least important aspect of Omaha but it's the end of the main math part of the game. In loose games, you can pretty much calculate precisely your chances of winning some or all of the pot.

Whether a player then makes or doesn't make their hand on the river really doesn't matter. You do everything right mathematically up to this point, and lose to a one outer, that is fine -- just do the same things again and again the next times. Omaha (and all the other games) is about having the best of it in the longrun. There is no "leader money" in poker. The "best" hand is the one with the highest winning potential (including the understanding that some hands will win more bets than others). Don't think what just happened was an aspect of a "river game". I can't emphasize this strongly enough: All the truly important actions in this hand occurred before that river card happened to bring you bad luck.

Another thing to consider is that only a tiny percentage of money action is on the river in Omaha. Poker is about money. Omaha is not about the river. That's naive. Omaha is about getting money in the pot in a mathematically advantageous way before the river. Limit Omaha High Low is an anti-river game!

Put another way, if you play a coin flip game against a guy, and he says he'll give you $5 for every time it comes up heads, but you have to give him $1 for every time it comes up tails, it would be wrong to refer to this situation as "a flip game"! The key part of the game was in the pre-negotiation, not in the flip itself.

Driving the pot... Loose game Omaha is mostly about nut hands. If there is a flush, you sure want the nut flush. If there is a low, you sure want the nut low. The obvious reason, of course, is because you have the winning hand rather than the second or third best hand. But that's not the only value to playing nut hands.

Again, winning Omaha requires pot manipulation -- get more money in when you have clearly the best of it; play for cheap when you don't. Nut hands and nut draws using quality cards can "drive the betting" where non-nut hands cannot.

For instance, let's look at the enormous difference between KK and JJ -- not in terms of how much more often KK makes the winning hand, but in terms of the difference in the pot sizes. KK is a much more valuable holding in part because KK can drive the betting in many pots that JJ can't -- like on a turn board of KQQ7 versus a board of JQQ7. The difference between those two situations is enormous. There are other reasons why KK is a major holding while JJ is a minor one, but the difference in how each can drive the betting (or not) offers an excellent illustration of what situations you want to be in when playing Omaha.

Likewise, there is a very large difference between A23x and A2xx on a 87K flop. The latter hand should win less money, not just because it will be counterfeited sometimes and not make the winning hand, but because it cannot drive the betting nearly as much (if at all) as the A23x can. A256, A247, A269, all these hands should win extra money not just because you make winners more often, but because you should be driving the betting with them far stronger than with the one-dimensional A2.

Cooperation... Greedy players make lousy Omaha players. Foolish greed often costs players bets because they simply don't recognize that the game frequently requires cooperative betting.

Suppose there are three people in a pot. On an 8s7s5c flop, Player A bets and is called. The 9h comes on the turn. Player A bets again, Player B calls, Player C raises, Player A reraises, B calls, C caps, A and B call. Now the river card pairs the board with a flush card, the 9s. What now? Often Player A will bet, with no high hand, and Player B will raise, with no low hand. This will drive Player C with a straight and a weak low out of the pot. Translation: stupid Player A and Player B.

Instead of cooperating to get at least one bet from Player C, they got none. If Player A stupidly bets, Player B should call, and hope to get one bet from Player C, or perhaps an idiotic raise. The better play though would be for Player A to check, have Player B bet, get Player C to call, then Player A checkraise, and have Player B now call. This way you get at least one bet from Player C, maybe two. Think about how you can use cooperative betting between high and low hands to extract bets from players in the middle. Don't be greedy and cost yourself money.

Luck... While the emphasis on the non-random mathematical nature of the game above makes the point, I'll mention a few things about luck as it applies to Omaha. All poker has luck involved. Omaha is the most mathematically straightforward poker game -- very little randomness, very much known information. So, when someone makes a miracle one-outer on the river, some people will mistakenly think of Omaha as having a high degree of luck, when the opposite is plainly true.

Omaha is a bit like a roulette wheel. If you have bets on all the numbers except one, when it happens to come up that other number that is really bad luck. But, now suppose the person who bet on that one number also put up as much money as you did. You had thirty-six chances to win, he had one, playing for the same prize. The longrun outcome of this game is surely not going to be determined by luck! You will crush your opponent, either very soon, or a little while later. When he gets lucky, he gets super-lucky, but that's just fine, as long as he is willing to keep making the same bet over and over.

Hold'em has far more random luck than Omaha (or Stud). That's why it's the most popular game. Poor players can do better, longer. Winning Hold'em is a game of exploiting tiny edges often. Winning Omaha is a game of exploiting huge edges less often.

In most ways, Omaha is a far simpler game. When played by good players, Omaha games are horrible -- unless the blinds are huge, forcing players to gamble. This is why Omaha is often played with a kill, to generate action in a game that should have very little.

This is also why Omaha will never be "the game of the future." Poor players have no chance. Good players eat them alive. In many localities, Omaha games burn brightly for a while, and then burn out as the bad players go back to Hold'em games where random luck gives them a fighting chance.

Quartered... In loose games you should hardly ever think about being quartered (when you have the same low hand as another player). It's almost never very costly to be quartered in limit Omaha. In loose games, one of the principal plays you should always have on your mind is how you can get three-quarters of a pot with hands like nut low and one pair. Too many weaker players obsessively fixate on being quartered with this sort of hand instead of focusing on getting three-quarters of the pot occasionally. The quickest way to get over a pathological fear of being quartered is to just do the math on various situations where you get one-quarter. It's hardly ever much of a loss. Now compare that to similar hands where you manage to get three-quarters of different size pots. You'll quickly see that many tiny losses getting quartered are more than compensated for by a few occasions where you can snatch three-quarters.

Scooping... High-Low Split poker is about scooping the pot -- winning it all, not splitting. Many weak and beginning players think they are playing decently because they focus on hands with A2 or A3 that make the nut low. These hands are playable obviously, and getting half a loaf is better than none, but this is most definitely not why you should be showing up to play Omaha (or Stud HiLo for that matter).

Once again, just doing some simple math is very illuminating. Scooping a pot is not merely twice as good as splitting. Suppose you play a five-way pot. Everyone puts in $80. If you split the $400 pot, you get back $200, a profit of $120. But if you scoop, you get $400, for a profit of $320. That's not twice as good, it is 2.67 times as good. In a three-way pot where you all invest $80, if you split you get $120 for a profit of $40. If you scoop, you get $240 for a profit of $160 -- four times as good as splitting.

The real reason to play A2 hands is not for the benefit of making the nut low and splitting a pot. The reason to play this hand is because while it is splitting the pot some of the time, it allows other parts of your hand to be aiming to scoop the pot. When you play A2, you actually want to be using some other aspect of your hand, something that will scoop. A2 just makes it safe for you to play, including often giving you the chance to make backdoor straights and flushes that you otherwise would not have stayed in the pot to make. This again goes back to "driving the pot". A2 allows you to drive the pot in situations like where you have A2JT with the nut flush draw and the board is 4678. Your A2 allows you to stick around for the gutshot straight draw, and allows you to aggressively bet your nut flush draw. That is where the money is, not in splitting the pot with the nut low.

Hands as units... The above illustration also should help make the point that Omaha hands are complete units. Despite the "must play two" aspect of the game, Omaha hands should not be looked at as six two-card holdings. Doing so is to fundamentally misunderstand the game.

It should be easy enough to see though that while 3d3h is a basically useless Omaha holding on its own, when combined with an As2s it now becomes a powerful aspect of a coordinated hand! Viewing the 33 out of the context of the A2 is a serious error.

Beyond the simplistic thinking about starting hands, it is critical to think of Omaha hands as complete units after the flop. You may play As2s3dQd, but end up with a flop of Qs9c2c. Before the flop no point-count system would assign the Qd2s aspect of your hand any value, but now here on the flop it is part of your whole hand, and you must think in terms of how you have two pair, a backdoor flush draw, a back door nut low draw, a backdoor wheel draw, etc. Omaha hands are multifaceted and multi-dimensional. They should be viewed and analyzed as integrated wholes, not separate parts. An Omaha hand can be greater than the sum of its parts, sometimes even less, but Omaha hands are always units of all your cards.

Situational analysis & starting hands... All winning poker requires situational judgments. Some folks just hate that. They want easy, cookie-cutter answers. Sometimes difficult problems do have easy answers, but more often they don't. Hold'em is a more situational game than Omaha, but because of that, when situational judgments are needed in Omaha, they are usually very critical -- inspirational even. For example, bluffing is not something you should do much of in loose game Omaha, but there still is a lot of profit to be made from bluffing, precisely because nobody thinks it is a big part of the game!

Omaha HandMost players play a lot of hands in Omaha, more hands than they play in Hold'em. Generally, the proper play is the reverse. However many hands you play in Hold'em, you should play less in Omaha. (Again, Hold'em is a post-flop game where playing junk before the flop can often be situationally correct.) If you are in an Omaha game with people violating this concept, as most Omaha players do, then you should only be focusing on playing strong hands and, in the correct situations, a few highly speculative hands that make for big scoops. The latter group boils down to KKxx, and QQ with two decent other cards. All other hands should contain an ace or be highly coordinated (KQJT, QJJT, 2345). The weakest of these are also more speculative (like the three examples). They aren't very good, and don't hit that often, so you want to try and play for only one bet, but when they do hit, they pay off nicely, so in weak, loose games they should be played. In tougher games, all these speculative hands without an ace should normally be mucked without a second thought.

A very good (but not spectacular) hand like AK32 with a suit on the King will scoop somewhere between 20 and 50% more than a random hand, depending on number of players and positional factors (and will split far more than random hands). If you are on the button and don't raise with this hand when everybody limps in, you are playing lousy poker. On the other hand you normally don't want to raise under the gun with low-only hands like A234 because you want players. You want to play your very good hands for a raise, you want to try to put in an extra bet when you can, but sometimes you can't... and you have to go to plan B.

A very general starting point for full-table, loose-ish ring games is: always have an Ace. KK with two decent cards and four Broadway cards (double-suited ideally) are ace-less, speculative, limp-if-you-can hands.

And then there is this: no hand loses more in Omaha High-Low than 23xx. Maybe you will find limp situations on the button or small blind to speculate, but overall "2-3 players" are a major source of a winning players income.

The end of the beginning... Advanced Omaha strategy goes quite a bit beyond the above, but most Omaha players go nowhere near as far as we go here. Once you think correctly about your approach to the game, like correctly viewing how much better scooping is than splitting for instance, advanced strategy concepts become more readily apparent, and your play will evolve and adapt.

One reason good players beat bad players at Omaha is because good players are thinking about the right game. Don't be concerned about losing pots. That's defeatist tunnel vision. Instead, be concerned with getting your money in with the best of it time and time and time again, and then letting the math take care of things in the longrun. That is Omaha. The introduction to it anyway...