CardSharp Mailbag: Split Two Pair in NL Holdem vs. Two Pair in Draw Poker

5 Card Draw, No Limit Texas Holdem, Reader Questions

October 2, 2008

One of my readers sent in the following question which I think deserves discussion:

Wayne,

I have stumbled across your website and I think your articles are excellent. I am writing to request your insight about split 2 pair. An article or an email would be much appreciated. I have heard it referred to as a trap hand and I can see why. As I understand it Mark Twain once said “That knowing how to play two pair correctly is worth a college education and costs about as much to learn how to do so”. I am paraphrasing but you get the point. The heart of my dilemma is this: I do not know if I should play all split two pair hands as drawing hands hoping to fill up or play them strongly knowing I will have to win with just two pair. Two pair (especially top two pair ) seems to be too strong a had to play as what amounts to a gun shot draw but more often than I like I seem to be toast to a set on the flop or beat by the river. I know board coordination plays a role but I REALLY hate getting my money in when I am beat to a set on an uncoordinated flop board. Is this just a cooler like set over set. What is a poker player to do? In my mind there is a big difference between 3 types of split two pair you can have and thus there should an equally big difference in how you play the hand. But I am not sure how to work this all out or if I am on the right track. Any advice you can give or articles you will write are much appreciated! I look forward to reading more of your poker insights in the future. Thanks.

This is a good question. Mark Twain’s comment about two pair certainly refers to draw poker. While draw poker is essentially a dead game in this day and age, it’s worth looking at what happens there as background for thinking about 2 pair in holdem.

Two Pair In Draw Poker

In draw poker, it’s a fairly common occurance to be dealt 2 pair before the draw. And at first glance it appears to be a pretty decent hand – it’s somewhat rare for any of your opponents to be dealt a hand better than two pair, so the chance you’re ahead before the draw is very good.

However, in draw 2 pair is often times a problem hand – especially in NL 5-draw. The problems arise when both pairs are of low rank. When this is the case, 2 pair plays poorly because it ends up on the wrong side of reverse implied odds.

Suppose for example, that two players see the draw in a fixed limit “guts” draw game with no joker. The Player A has AAxxx, opens the pot, and draws 3. Player B has 7744x, calls behind, and draws 1. Player A has roughly a 3:1 against shot at outdrawing player B. So it would appear the 2 pair is a solid favorite. But something funny happens after the draw. Player A has a pretty good idea of where he stands. First off, it’s not that hard to put player B on 2 pair. It’s unlikely he would draw1 at a straight or flush with only one opponent in the pot. He could be drawing one to conceal trips, or have four of a kind, but those are far less likely. So A is wise to assume B has 2 pair. Now, after the draw A is going to be able to play his hand with a lot of clarity – if he makes two pair or better, he knows he’s likely good. If he doesn’t make 2 pair, he knows he almost certainly has the worst hand, and can get away from it easily. Player B on the other hand has no clue if he’s good. He can probably put A on aces before the draw, but has no clue if he’s good after the draw unless he hits his full house. Suppose A mixes up his post-draw play when he hits and sometimes leads out, and sometimes check-raises. B will have little choice but to call down most of the time. Thus Player A will figure to get more than one big bet on average when he hits – say 1.5 big bets. Now, the amazing thing is that A actually has the best of it here – 1 time in 4 he wins on average 2 big bets plus the antees, and 3 times in 4 he loses 0.5 big bets. Even without worrying about the antes he’s netting out an average of 0.125 big bets every time the hand plays out this way.

Now the above example makes some big assumptions by fixing the hands and some of the lines of play and ignoring the small chance B hits a full house, but the concept should be clear here: reverse implied odds renders two small pair much worse than you might think. If you you don’t believe me on this, you can find confirmation in any of a number of draw books including the SuperSystem 1 chapter.

This feature of two pair hands, namely that they are often inferior to one big pair hands, is probably what Twain was talking about. The situation obviously gets worse in NL 5-draw since the post-draw bets will be larger compared to the pre-draw bets and thus player A’s information advantage will be worth even more.

The above information probably isn’t of much use to the average player, but it is a nice example to think about because it clearly illustrates how medium strength made hands can get into a lot of trouble.

Two Pair In NL Holdem

Let’s move away from Twain’s Mississippi riverboat poker and talk about modern holdem. Obviously no one can have two pair before the flop with only two hole cards. From the flop on, two pair hands become possible. These two pair hands need to be classified a little further. I’ll refer to the case where two unpaired cards in a player’s hole match two cards on an unpaired board as “split” two pair hands. It should be noted that all other two pair hands, which involve a paired board, are much weaker and should often be treated as if they were one pair, or even weaker than that. These aren’t what the question above is asking about, so I’ll ignore them forfor the rest of the article and just talk about “split” two pairs.

A split two pair in holdem is much stronger than two pair in draw. This is partially a function of the odds involved, and partly a function of the extra betting rounds in holdem. Consider a situation analogous to the one above where player A has a high split pair and player B has a split 2 pair – say AK vs 87 on a K87 board. Player A has about a 10% chance of drawing out on player B on the next street – much less than in the draw example above (where it was 25%). Furthermore B gets another opportunity to bet before A gets the other half of his draw. This means that B has a much easier time betting enough to make A’s draw unprofitable. So you shouldn’t feel like two pair is a garbage hand in NL the way it often is in draw.

An interesting aside is that split two pair hands are in practice fairly rare in NL holdem – they’re usually rarer than sets for example. Because many unpaired hole card combinations are unplayable, many hands that would eventually evolve into two pair should be folded before that can happen. It’s never profitable in holdem to enter the pot primarily hoping to make two pairs. so when a two pair does happen, it tends to be a “fluke” occurrence like the example above, where player B played the 87 not for two pair potential, but because of straight potential (and ideally flush potential if suited).

Oftentimes you’ll see players classifying two pair hands by which board cards they pair. For example, with the example above, K8 would be the top two pair, K7 would be top and bottom pair, and 87 would be bottom two pair. This distinction is not as useful as it might appear, however. If you hold the 87, and your opponent is even vaguely selective about starting hands, it’s highly unlikely that your opponent holds K9 or K7. That means that in effect you hold the only plausible two pair, and it doesn’t really make any difference that it’s bottom two pair. The only time this distinction matters is when the board has multiple plausible two pair combinations (in which case it’s actually more likely someone has a straight) or your opponent is very loose and could have almost any two cards. It should be mentioned that if you frequently find yourself holding two pair especially if that was the preflop plan, you’re probably doing something wrong.

Because two pair hands are very rare, and two pair vs. two pair hands far rarer, in practice it’s often a good idea to treat any two pair hand you end up with as if you held aces. Usually the two pair will be exactly one notch better than the aces in terms of practical hand rank, so this isn’t a bad approach. Previously on CardSharp I’ve posted a lot on playing aces and similar hands post-flop here, here, here and here. if you simply take the post-flop advice from those articles and replace aces or top pair/top kicker with two pair, you’ll have pretty solid advice. The techniques suggested therein to prevent your opponents from set farming you to death should work equally well when you hold two pair. Hopefully that answer’s the reader question that prompted this article. In summary, two pair hands should be played like most other made hands that are short of the nuts.

It should be noted that while it often works well to treat two pair hands as if you held aces postflop,your preflop play often wasn’t consistent with holding a big hand. First off, this is an additional argument for raising your suited connector type hands preflop as I suggested in my preflop articles (1,2) since this will make these hands play better postflop if you do hit two pair by accident. However, it’s also reasonable that you may end up limping preflop with a hand that subsequently makes two pair. These situations are very dangerous. The aces and payoff rule articles linked above illustrate that you need to scale the amount of money you would like to get into the pot postflop based on the amount of money that went in preflop. So in the case where you limped preflop, it’s quite likely that you won’t actually want to get very much money in postflop at all. This may seem counterintuitive when holding a seemingly good hand like 2 pair, but it’s an inescapable consequence of the mathematics involved.


This article is part of Project Cash Game No Limit Holdem - You can find more great strategy articles there.
Like this article? Subscribe to the CardSharp RSS Feed


Leave a Reply