Drawing Multiple Cards in Lowball

Roses are red,

Violets are blue,

Don't draw to eights,

And don't draw two.

Old-timers sometimes give this simplistic advice to lowball beginners – and then proceed to break the advice themselves. Like many such sayings about poker, it contains a bit of truth, but is not the whole story. Just as in much poker advice, this one needs to be tempered with the ubiquitous "it depends."

Lowball is a very position-dependent game, perhaps more so than any other form of poker. That advice is a good general approach to lowball strategy from early positions. But, if you don't draw two cards in the right situations and you don't draw to eights, you may win a little in the small, loose games – although you won't win much because you'll get such a reputation as a rock that your good hands won't get paid off – but in higher-limit games, you probably won't win.

T

his column is not going to give you opening and calling strategy for lowball. You can get those from Mike Caro's excellent draw strategies (now, sadly, out of print) or from my Free Money: How to Win in the Cardrooms of California (look for it on Amazon.com or Half.com). I want to talk about when to draw more than two cards.

Now, first, there's a time when you can always draw three, four, and even five cards. That's in a Southern California-style double-limit game when you're in the big blind and the only players have limped. That is, you get a free draw. In that situation, you usually draw to the best hand you can. Sometimes, you start with five facecards and have no choice but to draw five cards. But why not? It's "free." That is, it costs you no more.

But this is where many old-time lowball players make a mistake. Four people are in the pot and they have 10-9-9-2-joker. They draw one in this spot after the small blind draws two cards. Here, it's better to draw three cards. Sure, you end up with the best hand more often when drawing one to the 10 than you do when drawing three to the deuce-joker.

In a typical matchup, the one-card draw wins about 14 percent of the time. This is, of course, less than the 20 percent of the time an average hand ought to win, but you started with a hand somewhat worse than average. Now, the three-card draw wins close to 13 percent of the time. This is not as good, of course, but when the three-card draw does win, that win is often accompanied by multiple bets after the draw. The best the one-card draw can do is call after the draw, usually losing a double bet in the process, so that 1.2 percent more often that it makes the best hand is misleading. Why draw to a hand that you don't even want to make? There are times in lowball to draw to tens – they're rare – but this is not one of them. And here's another factor: When you pair on the one-card draw, half the time that's a pair so large that you can never call with it. But sometimes you pair very small with the three-card draw and can pick off a bluff.

The matchup is even closer if you don't have the joker. Drawing one card to 10-9-2-A against typical opponents in a limped pot wins about 9.5 percent of the time. Drawing three to A-2 wins about 9 percent of the time. Sure, you don't win very often when you start with that miserable hand, but you didn't expect to. If the pot had been raised, you wouldn't even have been able to play, so in effect the 9 percent of the time you win is "gravy." And your gravy becomes even more piquant when you make something like a three-card 6 and beat a two-card 7.

How about if the pot is raised when it gets to you? As it turns out, against typical opponents, a three-card draw if it's to 4-joker or better has positive expectation. Yes, you still lose 87 percent of the time, but it cost you one bet to make the draw. Against four players, your call represents 11 percent of the pot (since the big blind represents one bet) – less if the small blind or button doesn't play – so winning anything more than 11 percent of the time is profitable. And you do; you win 13 percent of the time, approximately, when you're drawing to 4-joker or better. Plus, when you win, you often win several double bets. The situation is the same in Northern California's single-limit lowball games, where you should draw three cards to 4-joker or better against any combination of players that is not composed of all rocks. And if the majority of players in your game are rocks, what are you doing in it? (Yes, I know, the right one or two live ones can make even the rockiest game worth sitting in.)

Just don't use this advice as an excuse always to draw three cards when you have the big blind, paying no attention to the action. And don't be drawing to worse than 4-joker if it costs another bet to get in.

Now, what if your hand is something like 10-10-9-8-7? Here, it turns out that while drawing one card to the 10-9 is horrible, it's not nearly as horrible as drawing four cards to the 7 or five cards. And if you're going to draw to a rough hand, you're better off drawing one to the 10 than two to the 9 or three to the 8. Doing so is another mistake old-timers make. The only time to make those multiple-card draws is when you can draw smooth. This means that it's probably also preferable to draw one card when you have 10-9-7-7-6 than three cards to the 7-6. Drawing three to 6-5 versus one to a rough 10 is probably a toss-up. That you can still make a hand with which you can get multiple bets after the draw may lean in favor of the three-card draw. But if you draw to 7-6, the best you can make is a rough 7, and that's rarely a raising hand after the draw, so you're better off drawing just one to the rough 10. (Again, in all of these situations, "better off" means "less bad off.")

How about K-K-K-8-joker? Here's where lots of pros make another mistake. You're better off drawing four to the joker than you are three to the 8-joker. Yes, the three-card draw wins more often, but you never make a hand with which you can win multiple bets. Specifically, in typical matchups, the three-card draw wins about 9 percent of the time, while the four-card draw wins about 7.15 percent of the time. But when it wins, the four-card draw often wins multiple bets, more than making up for the 1.85 percent fewer wins.

And here's a similar choice. If it comes down to drawing four cards to an 8 or drawing five cards, you might as well draw five. That's another one I see old-timers get wrong.

There is no equivalent to the four-card draw in the Northern California single-limit lowball games. Against multiple players, you almost never would take four cards, nor would you call half a bet to draw three cards to an 8. The times I'm describing here are when you don't have to put in extra money and need to decide between drawing three cards to an 8-joker or four cards to just the joker – for "free."

Here's another interesting point that the old-timers and pros often get wrong about these three- and four-card draws. What do you do when you make an 8? Usually, the player who drew that many cards to make an 8 thinks, "No one will believe I made a hand, so I have to bet." Well, surprisingly, while players may be suspicious, they usually do believe. They don't call with tens and worse, and they certainly don't call with pairs. But what happens if you check? Players completely disregard the three-card draw. They bet worse eights, they bet nines, and they bluff. You can call and win all of those pots. And often when you make your miracle three- or four-card 8, someone makes a smoother 8 or a 7 behind you and raises. Now, you're stuck for a call, because you don't know if the person is raising because he really made a good hand or just because you drew so many cards. Most of the time he'll have you beat, but often enough the raise is with a rough 8 or even a 9 – and sometimes even a pair – that you have to call. Most often, you lose two bets in a situation in which you need only have lost one. Plus, if you bet, the next player raises, and the player behind reraises, you are forced to fold when it gets back to you, thus losing one big bet in a situation in which you didn't have to lose any. That is, if you had checked and there was a bet and a raise behind you, you easily could have folded when the action returned to you.

A curious thing about these old-timers getting it wrong is that had they drawn one card and made an 8 and there were three players behind them, they would have checked. Why should a three- or four-card draw resulting in the same hand be any different? If there's a bet behind you, or a bet and a call, you can just overcall and surprise the initial bettor when he finds his one-card rough 8 beaten – or be pleased to have saved a bet when a big hand gets shown down.

From this you should take that you never draw more than three cards if it costs anything to do so. This would be a raise-open or a raised pot in a Southern California double-limit game, or just an open in a Northern California single-limit lowball game. Watch out for that never, though. If your opponent has the small blind or the button, and you know this player regularly to open for a raise in that spot with a two-card draw, or, better, in the single-limit game just to open, you can draw four cards in this spot. This is because a player who opens with most two-card draws has exactly that – a two-card draw – most of the time. Usually, your four-card draw should be only to the joker. It's nice if this is against the small blind and he likes to gamble, because then when you draw four cards, he might just bet blind. And if he doesn't do so spontaneously, you can suggest he do so. Some players might try to shame their opponent into a bet by saying something like, "What, you ain't gonna bet blind into a four-card draw? Sheesh, don't you have no gamble?" (Cardroom habitués pride themselves on their poor grammar.) Then, you just need to call with a hand that offers pot odds. That is, if there are four small bets in the pot, a bet after the draw brings the pot up to three big bets, and you just need to have a hand that beats a two-card draw 25 percent of the time or more. Against a typical draw, that's about a pair of fives or better. Game theory also has you raise with some percentage of your worse hands, but that's beyond this discussion. However, if you call with a pair of fives or better, raise with about a 9 or better, and also raise with trips or worse, you won't be far wrong.

You would draw four against a button open less often, because most of the time you're forced after the draw into the situation of either folding or trying to pick off a bluff. This being a position game, you don't want to give up most of your edge by indicating ahead of time that you're taking far the worst of it. Even a poor player would stay pat here on a Q-J or, certainly, a J-10, whereas, if he had the same hand in the small blind, and thus had to act first not knowing what you were going to do, he most likely would draw two cards. About the only time you would want to take four cards ahead of the opener would be if you knew he would bet blind when you checked. Can you know that? Sure. Lots of players pride themselves on their "gamble," and betting blind when you have checked is a point of honor for them. Of course, in such a spot, if you knew your opponent was going to bet blind, you would check any hand 8 or worse, and then use the same calling rules as in the earlier situation. The other time to draw four against the button is when this player does all of the following: regularly raise-opens with a two-card draw (or, in Northern California, regularly opens on the button with the hand), never stands pat on a J-10 (or a bluff!), and never bluffs after the draw. Are there players like that? Happily, yes.

Again, don't take this to mean that you need to call regularly with your four-card draws against the live one when he raise-opens (or opens in Northern California) from the small blind. Sometimes you can – but rarely.