Mistakes in Online Draw: Playing Come Hands

Back in California's old Gardena days, maybe 25 years ago, when cardrooms spread only draw and lowball, a player new to the cardroom sat down in the empty seat of a draw game. "Do they draw to straights and flushes here?" The grizzled old-timer was the only one to respond: "The boys does."

I find I don't make nearly as much in online draw from the brilliance of my own play as from the generally poor play of my opponents. Here is another common mistake: Playing come hands in online draw is almost always wrong. A come hand is a drawing hand – that is, a one-card draw to a straight, flush, or straight flush.

It is always right to draw to a straight flush, and it is often right to raise with this draw – even if it is an inside straight-flush draw, such as 3 5 6 7, plus one card that you will throw away. The other hands – the straight and flush draws – are the problematical ones.

In a fivehanded game, it is rarely profitable to play a come hand. In an eighthanded game, it sometimes is, but not nearly as often as your opponents play them. That, of course, is one way you profit, from the times that your opponents incorrectly play come hands.

Here are examples of typical play, each showing how the player costs himself money.

WeeQuan2 limps in from under the gun in a fivehanded $2-$4 draw game. The next two players fold. JamminJim raises from the small blind. The big blind folds. WeeQuan2 calls. JamminJim draws three cards. WeeQuan2 draws one. JamminJim checks. WeeQuan2 bets. JamminJim calls. WeeQuan2 shows a busted flush and JamminJim wins. WeeQuan2 put in $8 on a hand he should not have played. He was never getting proper odds to play his hand. Neglecting the bet after the draw, he was putting in $4 to win $4.50 (50 cents goes to the drop). But he would make his flush only one time in five.

We don't have to neglect the bet after the draw to show that his play is wrong. Let's say that whenever WeeQuan2 opens with a flush draw, JamminJim raises. Let's furthermore say that WeeQuan2 bets only if he makes a flush and yet JamminJim guarantees at least to call every time WeeQuan2 bets. In five plays, WeeQuan2 makes one flush. He wins $8. (In a $16 pot, $1 goes to the drop.) Four times, he loses $4, a loss of $16. One time, he wins $8. His net loss for the play is $8, or approximately $1.60 per hand. It's actually worse than that, though. Some of the times that WeeQuan2 makes his flush, JamminJim makes better than a flush. Let's say that three-fourths of the time that JamminJim raises, he has one pair. Approximately one time in 72, JamminJim fills up, and that costs WeeQuan2 at least three bets. One-eighth of the time, JamminJim starts with two pair. Of those times, he fills up one time in 12, again costing WeeQuan2 three bets. One-sixteenth of the time, JamminJim starts with trips and fills up one time in 10. One-sixteenth of the time, JamminJim starts with a pat hand. That's a wash, because maybe half of the pat hands he has are worse than the hand WeeQuan2 makes and half are better. So, maybe overall, one time in 20, WeeQuan2 makes his flush and loses. But he loses $16 when he loses and usually wins only $8 when he wins. That increases his overall loss per hand to maybe $2. These figures are approximate, and depend on what hands opponents raise with, but, still, the bottom line is that WeeQuan2 loses in the neighborhood of $2 per hand whenever he opens with a come hand. His average loss is more if he is drawing to a straight.

Now, we stipulated that WeeQuan2 bets only if he makes a flush and that JamminJim guarantees always to call. That is not how it works in reality, however. Either WeeQuan2 bluffs too little, in which case he doesn't get paid off when he makes the hand, or he bluffs too much, in which case his bet after the draw costs him even more, thus raising his average loss per hand. These two situations cost him money. The third possibility – and it's one that by itself has positive expectation – is that he bluffs at the correct frequency, in which case he manages to reduce his average loss to about $1 to $1.50 per hand. But anyone who doesn't understand why he is making a mistake playing the hand in the first place is not clever enough to employ game theory correctly on the betting after the draw. (Nor does this play offset the negative expectation of having played the come hand to begin with.) Also, anyone smart enough to raise WeeQuan2 when he limps in is also smart enough to adjust his strategy after the draw to accommodate WeeQuan2's betting habits.

You may think that WeeQuan2 can protect himself by sometimes limping in with two pair so that someone like JamminJim won't raise him every time he limps. But that, too, is a mistake. Good hands like two pair – or better – are so rare that you want to get more money in the pot when you likely have the best hand and also protect your vulnerable hands. And two pair is a vulnerable hand. If you limp in with two pair, you are essentially giving the big blind infinite odds to draw out on your hand. But if you raise-open, one of three things happens: The big blind folds, and you are guaranteed to profit by the amount in the two blinds. He calls with a worse hand, and you are playing for two bets with the best of it. And third, he has a better hand – but that is rare. Oftentimes I see someone limp in with a hand as good as aces up. The big blind draws three cards to his pair of deuces, makes a third deuce, bets, and the opener calls. He loses the pot, sometimes showing his big two pair and typing in the chat box, "Unreal," as if he has suffered a terrible bad beat. But if he had raise-opened with his two pair, the big blind would not have played. A sure $3 profit is better than taking the chance of maybe winning $6.50 at the risk of sometimes losing $6. So, sometimes slow-playing big hands to protect the times he's playing a weak hand is a poor strategy. The best strategy is to hardly ever play that weak hand. And in draw poker, a come hand is a weak hand because it wins one time in five, approximately, but rarely has 4-to-1 odds. It's not like hold'em, where you can flop four to a straight or flush and have two chances to make the hand, often with extra outs.

So, what if WeeQuan2 instead raise-opens with his drawing hand? That's a bad idea, because observant players will soon realize that his raise-open does not necessarily mean a strong hand, and will reraise with their good hands, thus costing him more money. Plus, the weak players will still play the hands they were going to play anyway, and beat WeeQuan2 four times out of five.

Of course, some of the times that WeeQuan2 limps in with his come hand, the pot is not raised. But if someone else has a pair, WeeQuan2 still loses nearly four-fifths of the time. Once in a while, he pairs one of his big cards – if he has any – and beats the smaller pair that someone else might have played. I say "if he has one" because I frequently see players either limp in or call two bets with something like 2-3-4-5 of mixed suits. Pairing the top card is probably not an out with that hand. If the hand is 9-10-J-Q and no one has raised, pairing one of those cards might add extra outs to the hand. That doesn't make the hand any more playable in most situations, however, particularly if there has been a raise.

Here's an even worse play that you will see frequently that will also make you money: JamminJim raise-opens from the button. WeeQuan2, to his left, calls from the small blind. WeeQuan2 again has a come hand, and draws one card. JamminJim takes three. The after-the-draw scenario is the same as before. That is, unless he bluffs exactly perfectly, he loses somewhere between $1 and $2 per hand by making this play; plus, his position is terrible. What usually happens is that the big blind calls also, and the draw is something like one card, three cards, three cards. WeeQuan2 checks, the big blind checks, and JamminJim checks. WeeQuan2 shows his busted flush, the big blind shows maybe a pair of nines, and JamminJim shows, say, a pair of jacks. Or, maybe the big blind has a pair of kings, and wins. That happens most of the time. Why is this play even worse? Because WeeQuan2 came in cold for $3 (one and a half bets) when the pot was offering odds of only 7-to-3. Even if you count the implied odds of getting called after the draw, the odds are still only 11-to-3. And you can't count the implied odds for two reasons. One, he might make the hand and not get called. Two, he might make the hand and get beat.

Even worse than either of the two preceding plays is this: JamminJim raise-opens from the cutoff. WeeQuan2 calls from the button, with that same come hand. Now, he's put in a full $4 with the pot currently offering only $7. Maybe one of the blinds will call, increasing the potential return. But maybe one of the blinds will reraise, and maybe the pot will be capped. Whatever he does is a mistake.

So, when do you play a come hand? In a fivehanded game, rarely. In an eighthanded game, you can play a come hand more often. The wisdom among old-time draw players was that you did not play a come hand unless the pot already offered odds of 5-to-1. If you count the blinds, you can come in if four players are already in the pot. You can see that this happens in a fivehanded game only when you are one of the blinds. In an eighthanded game, it does happen sometimes. In a fivehanded game, you can come in from the small blind for half a bet if only one player is already in, because then you are getting odds of exactly 5-to-1. But be careful if the big blind is an aggressive player and a few players have limped in, because a reasonable chance exists that he will raise. If he does, you must call, because for one more bet, you are getting better than 5-to-1. If you have the big blind and two players are in for two bets each, your call gives you current pot odds of 5-to-1, and now implied odds are relevant. If you make your hand, you will almost always get called.

You can shade this last requirement slightly if you know for sure that one of your opponents will bet if checked to, so that you can check-raise. That is, if the player on the button always bets when checked to when he raise-opens and almost always calls a check-raise, you should call to draw to a come hand and check-raise if you make it. Do not make the mistake, though, of betting if you miss – and miss you will 80 percent or so of the time.

If you end up playing fewer than 10 percent of the come hands you get dealt in online draw, you're not making a mistake. Let "the boys" draw to those hands – and put chips in your stack.

Michael Wiesenberg's The Official Dictionary of Poker is the ultimate authority on the language of cardrooms. Order it online at CardPlayer.com.