Heads-Up Lowball Strategy

If you play lowball for long enough in Southern California, or if you play online, you will find yourself heads up and you will want to keep playing. Maybe it will be to hold a game together until the live ones come in or come back, or maybe it will be because your opponent is a live one. Maybe everyone left for dinner except you and the live one. Maybe a particularly juicy producer always comes in Thursday night at 7 p.m. It's 6 p.m. and just you and a semipro are at the table. Ten names are on the list. If you leave, you go to the bottom of the list. You want to make sure you're seated when the live one comes in. Whatever. The opportunity will come up when it will be better for you to play heads up than to quit.

The strategy for head-up play differs greatly from that of a ring game. Don't feel like you have to quit when the live one offers to keep playing head up, or you see someone waiting online to play. Use this strategy to beat that person! You can play at least even with the best pros, or better yet, beat them because they don't know proper twohanded strategy.

I'm talking here about double-limit lowball, that is, a Southern California-style or online double-limit game. The discussion does not apply to Northern California-style (single-limit) lowball, where you generally don't want to play heads up against a tough player.

Players generally employ two strategies when playing heads up. Each has problems.

Heads-up lowball involves two blinds. The small blind (SBB) is on the button, while the big blind is held by the other player. The reason for this is not to give a double advantage to the button. The button has a greater advantage in lowball than most other games because not only does the button get to see how the other bets, the button gets to see how many cards the other draws. Allowing the button to act last before the draw would be just too much of an advantage, so before the draw, the button acts first. After the draw, the betting starts with the other player.

The worst of the two strategies is the straightforward approach. The player limps when he has a poor hand and raise-opens with a good hand. More specifically, such a player limps with a two-card draw or a one-card draw to a rough hand and raise-opens with a one-card draw to a good hand or a good pat hand. Since he's playing shorthanded, he probably considers a one-card draw to a good hand to be a draw to an 8 or better, and a good pat hand a pat 9 or better.

It's easy to defend against this approach, and you should easily beat someone who plays this way. When he limps, raise with any better hand plus a few bluffs. A better hand would probably be any one-card draw to about a 9 or better, or any pat jack or better. You probably would bluff also with your worst hands that contain several low cards, such as two pair, aces and deuces, or a small full house. You're probably even safe playing a fistful of paints, except that having four or five small cards leaves fewer cards in the deck for your opponent to catch. And remember, you don't want to bluff too often, because your opponent might catch on and start modifying his play. Most of the time it's best to be raising with good hands, and, of course, good hands here just means better than what he likely has. Many of the hands with which you raise in this situation you would not even play in a game containing more players. Much of the time when your opponent limps, you can just take the free draw. After all, you're getting almost infinite odds on your call. If your opponent limps and you have, for example, ace-joker, you're getting great odds to draw three cards. Or, in an online game without the joker, you're still getting a good shot with any three-card draw.

This player telegraphs his holdings when he raise-opens. You know he has a strong hand, and can play accordingly. When you have a hopeless hand, just give up and surrender your blind. If you have a reasonable draw, and that probably includes most good two-card draws, call (because you're getting 3-to-1 on your draw). And when you have an excellent hand, like a pat 9 or better or a good one-card draw, raise back. If he's not observant, he won't even notice that you don't play the absurd rough hands when he raise-opens.

What might make such a player a great opponent is if he never bluffs after the draw. You can't have everything, of course, so rarely bluffs should do. Even better, what might make him the perfect opponent would be if he doesn't adjust for short play against you. That is, if you raise-open, he just calls with his best draws and doesn't reraise with his rougher pat hands. Such a player doesn't have a chance.

A better strategy is the one adopted by many aggressive players. When they open, they come in for a raise with any hand they're going to play. They throw away the hopeless hands, but with any good two-card draw or better and any pat jack or better, they open for a raise. This strategy works against a timid opponent, but actually does pretty poorly against a smart player.

It's also relatively easy to defend against this approach, and you should also beat someone who plays this way. Your fluctuations will be wilder and your variance will go way up, but you still should win.

Since this player never limps, and since he plays any two-card draw or better, most of the time that he opens will be with exactly a two-card draw. Since he plays any pat jack or better, another large portion of the time that he comes in will be with a very rough hand. And since he also plays all of the rough one-card draws, a further significant portion of his opening hands will be those. His good one-card draws and his good pat hands ("good" being, as before, a one-card draw to an 8 or better or a pat 9 or better) should make up a relatively small portion of his starting hands. Knowing that, you can play him - and punish him - accordingly.

That is, if you have a good two-card draw, call and draw your two cards. You will often be drawing as good as he, and, even when you're not, you will be getting odds of 3-to-1 for your call, and most of the time will not be a 3-to-1 underdog.

If you have a good one-card draw - and that probably means any reasonable draw to a 9 or better - raise. If you have a pat 10 or better, reraise. Yes, you'll get trapped sometimes when he actually has a hand, and will lose four bets or more on a hand that might not even have cost you one bet in a full game, but you'll win enough hands to stay ahead.

He's playing much too aggressively, and you can easily take advantage of this play. If such a player also has an ego problem, he won't back off his losing play, and will continue to try to overpower you. All the while, you will continue to take his money.

The aggressive strategy works against a timid player who does not adjust. But it does not work against someone smart enough to modify his play based on what he observes.

But a strategy exists that simply cannot be beaten in twohanded play, because it involves game theory. Use it against a relatively sophisticated player, and the worst you can do is break even in the long run. Such a player, one against whom you should only break even, is rare, indeed, however. Most players who are capable of modifying their play based on your play won't adjust properly, and you'll hurt them. Use this strategy only as a starting point, of course. As soon as you see your opponent modifying his play too much in one direction, move in the other direction.

Ideally, you would like to play your poor hands cheaply and your good hands for multiple bets. But, you can't play that way against a sharp opponent, because he will use the counterstrategy I outlined earlier. You might think of playing only good hands, and coming in only for a raise. That's not good, either, though. That's a fine strategy for a full game, but not shorthanded, because the playable hands are hard to get. Twohanded, you have a blind every hand, so you need to be playing more than half the hands or the blinds will cost more than you win on your good hands. But if you play that many hands, always for a raise, your observant opponent will quickly realize that most of your hands are substandard, and will use the appropriate counterstrategy.

You need a compromise, something in the middle, that will keep your opponent from profiting by using either of the counterstrategies mentioned thus far.

Sometimes when you limp, you have a good hand. Then, when your opponent raises, you can reraise, and suddenly he has lost three or more bets on a hand that he wouldn't have lost even one had he known what you had. That is, sometimes when you limp, you're drawing to a bicycle, and you can put in several bets before the draw and perhaps several after. That will slow him down.

Sometimes when you raise-open, you don't have a premium hand: You're drawing two cards, drawing to a rough hand, or playing a very rough pat hand. That way, your opponent loses by following the strategy of playing only with very strong hands when you raise-open.

But you have to randomize, and do it in such a fashion that usually when you raise-open, it's with a hand that you'd like lots of action on, and usually when you limp, it's with a hand that you'd prefer to play for the minimum.

How do you do that?

Notice that when you raise-open, your opponent is getting pot odds of 3-to-1. For example, with blinds of $10-$20, when you raise-open, you put in another $30, bringing the pot to $60. His call of $20 offers 3-to-1. His call will become one-fourth of the pot. So, one-fourth of the time, your open should be counter to what your cards are.

That is, three out of four times that you have a hand you would like to be raising with, do so; one time out of four, limp with that hand. Contrariwise, three out of four times that you have a hand you would like to be playing for the minimum, do so; one time out of four, raise-open with that hand.

Watch what happens. If your opponent thinks that you limp only with garbage, he will raise you on substandard hands. One of those times, he will run into your pat 8 or so, you will put in the third bet, and he will find himself calling a raise either to draw to a 9 or stand pat on a 10, and then will lose another bet after the draw. Suddenly, he won't be so eager to raise when you limp. On the three out of four times that you do limp with your own substandard hands, do so with some backup. That is, if you limp with a pat jack, make it a pat J-10-8 or better, so that if your opponent raises and stands pat, you can draw two cards.

Also, if your opponent thinks that you raise-open only with good hands, he will throw away sometimes when you raise-open with poor hands, which is what you want. Or, he will just call with his good drawing hands, only to discover on one of those times that you either draw two cards, draw to a 9 or 10, or stand pat on a 10 or jack. He will then start giving you more action when you raise-open, which is what you want most of the time.

You'll have him guessing, but you'll be playing a calculated strategy. And as long as he's guessing, he won't be using the proper counterstrategy. The proper counterstrategy to this strategy would still be to play you as if you were always limping with your poor hands and raise-opening with the good ones. He'd be wrong one-fourth of the time, but still would profit overall. The thing is, though, that he won't figure it out. He will just notice the times that he has cost himself money, and will act based on that knowledge. And the times he will have cost himself money will be the times that you limped with a good hand and he raised with worse, and the times that you raise-opened with worse hands than his and he either folded or did not push back. But, again, he will be wary of those responses, afraid that you are slow-playing a good hand or wondering if your raise-open is an apparent steal attempt.

Lowball players, like most other poker players, are always full of hope, the wrong hope. They want to think that when you are playing strong, you are actually weak, and they have a better shot at beating you than they really do. They also fear that when you are playing weak, you are trying to trap them, and consequently back off any hand but the true monsters.

How do you accomplish this random one-fourth of the time such that it is truly random? Decide ahead of time on some one card in your hand - say, the first one you look at. If it is anything but a club (for example), play your hand straightforwardly. If that card is a club, though, play it the opposite way.

For example, on your button, your first five cards, in order, are 7 6 4 3 K. Raise-open. Depending on how he plays, if he raises back, you might want to go one more bet, or just call. If you have a better draw than that draw to a rough 7, you would probably automatically play back. Now, if your first five cards, in order, are 6 6 3 2 A, limp. If your opponent raises, of course you reraise.

Again, on your button, your first five cards, in order, are 3 2 5 K K. Limp. If your opponent raises, just call. (You might actually reraise one time out of four. You could do that if, say, your second card is a club. You might be putting in the third bet, then, on a two-card draw one time out of six. But that's a refinement to this strategy that is probably unnecessary.) Now, if your first five cards, in order, are 4 K 3 2 4, raise-open.

If your first five cards, in order, are 8 4 3 6 2, raise-open. If your opponent raises back, naturally, you reraise. But if your first five cards, in order, are 8 A 3 2 4, limp. If your opponent raises, reraise. (I would say that pat sevens or better are so hard to come by that you should always raise-open with them, no matter your first card. You can make that exception to the general rule.)

Again, on your button, your first five cards, in order, are J 2 10 4 A. Limp. If your opponent raises, just call. If your opponent just calls and draws, stand pat. But if your opponent raises, when it's time to draw, whether your opponent stands pat or draws one card, you should probably just draw two cards. What you do here depends on what you know of your opponent. If your opponent will check any hand worse than an 8 after the draw and bet any 8 or better, but will rarely bluff, you should probably stand pat, because while standing pat against a good one-card draw has a negative expectation, drawing two cards has a worse expectation, and you can act based on whether he bets. That is, fold if he bets; show your hand if he checks. But if your opponent will bet nines and tens for value here and will sometimes bluff, you're better off drawing. Now, if your first five cards, in order, are 10 8 3 2 J, raise-open.

Similarly, your first five cards, in order, are 9 8 K 4 A. Limp. But if the first cards are 2 7 9 A A, raise-open.

Very important: Observe how your opponent reacts to your strategy. This strategy is only a departure point. You have to modify it based on your opponent's reaction. If he remains tricky and impossible to read, continue with the strategy. But if you observe that the only time he raises is when you limp, start limping with a much larger percentage of hands. If you notice that he always raises when you limp, obviously limp with every hand you play, and push back with your stronger hands. If he frequently calls and rarely reraises when you raise-open, raise-open with your playable hands. In other words, use that aggressive strategy described earlier, or lean toward it.

You can use the strategy for more than just your opening requirements. You also can react to your opponent's opening plays. Remember that you can usually do better than game theory by playing based on your opponent's plays, but against a tricky player or one you can't figure out, you can use this strategy.

If your opponent opens in what seems to be random fashion, or at least has an opening strategy that you can't figure out, react straightforwardly approximately three-fourths of the time, based as before on that first card as your "random number generator," and go against what your cards dictate approximately one-fourth of the time. Of course, since you are out of position, you should hardly ever raise or reraise with a two-card draw. You put the pressure on your opponent by your betting when he has to draw first; when you have to draw first, you make it easy on him. If he was first to act and you had raise-opened, he would be hard-pressed to stand pat on a J-10. But if you were first and took two cards, standing pat would be relatively easy, even if you had raised his opening bet or raise.

So, if your hand is a good drawing hand and your opponent limps on the button, reraise approximately three-fourths of the time and just call about one-fourth. Respond similarly if he raise-opens. That is, with 6 4 K 3 7, raise or reraise; with 5 4 2 2 A, call. Make a similar three-fourths/one-fourth distinction with your poor one-card draws, like 9 8 4 6 K and K 9 7 4 2, except the other way. That is, call if your first card is not a club; raise if it is. As before, slow-play one-fourth of your pat hands and aggressively play the remaining three-fourths. Also, as before, probably always play your pat sevens and better aggressively. But if you do, over a long period of time, the observant opponent will figure out that you never slow-play the monsters, so if you do slow-play a hand, he will figure out that it must be an 8 or worse. However, in a typical session, you won't have enough pat monsters to constitute a statistically significant sample.

Further note: If you are concerned that your observant opponent may notice that your having a club in the door affects how you play your hand, make the determining card other than the one that ends up in the door. Of course, if you are playing online, it does not matter.

The strategy documented here is a rough approach to the full strategy. Fine-tuning and refinements would make it too long to fit here. Notice, for example, that I mentioned the joker in only one place, nor did I mention the sevens rule, and the possibility that exists only online of check-raising after the draw. Those omissions are deliberate, to permit presenting this strategy in a form that works equally well in a cardroom and online.

I also said nothing about calling bets after the draw and bluffing when you miss your draw. Those decisions can be made with the same game theory approach used in ring games.