There are two universally applicable techniques for reading hands in all poker games and one more for open-handed games like seven-card stud, razz, and hold 'em. Most commonly you analyze the meaning of an opponent's check, bet, or raise, and in open-handed games you look at his exposed cards and try to judge from them what his entire hand might be. You then combine the plays he has made throughout the hand with his exposed cards and come to a determination about his most likely hand.

Here is a simple problem in reading hands that should make this point clear. The games is seven-card stud, and your opponents are decent players:

Player A with the pair of aces showing bets; Player B with the pair of kings showing calls; and Player C with the pair of queens showing calls. There are no raises. You are last to act. How should you play your three 7s?

If you combine what you see on board with what your opponents have done, there should be no doubt in your mind that you must fold; your three 7s have no chance whatsoever. The crucial factor is that the pair of queens overcalled. Player A may be betting with aces alone. But when Player B calls him, Player B must have at least kings up. Being a decent player, Player C knows this. Therefore, C could not call without having kings up beat. What are C's possible hands? Well, C cannot have aces and queens or kings and queens because there's a third ace and a third king out, making it impossible for C to have two of either. So, he must have three queens or better, and while your three 7s might beat the first two hands, they cannot beat C's three queens or better. Therefore, you fold.

Here is a good example of this kind of hand reading, which to my chagrin cost me half a pot. I was playing five-card stud high-low split with a replace on the end. With an ace and an 8 showing, I called the maximum raises on third street even though two other players each had a 6 and a 5 showing. There was another player in the pot with an obvious pair of kings. When it got down to the last card, I had A,8,6,3 showing. One 6,5 had folded, but despite the strength of my board, the other stayed with a ragged 6,5,10,Q showing. And of course, the pair of kings stayed. Now I was betting and raising, hoping the Q, l O low would get out. But that player read me too well. He didn't even take the opportunity to replace one of his cards.

What I was trying to do was win the whole pot, the high and the low, from the two kings, but the Q, l O low was clever enough to figure out my hand. He said to himself, "Sklansky is representing an 8 low, but could he have an 8 low? No, he couldn't. Why? Because he would never have called all those raises on third street with three cards to an 81ow when there were two other players in the pot who looked as if they had three cards to a 6 low. Therefore, he must have another ace in the hole." He was, of course, absolutely right. I won the high with my two aces, beating the two kings, but the Q,10 low was rewarded for his accurate reading with the low half of the pot (which I would have won against the two kings with my two aces counting also as a low pair). The player with the Q,10 low considered the way I played the hand not just at the end, but from the beginning, and he combined my play with the cards showing to arrive at the correct conclusion about what I was holding. He also analyzed the order in which I received my up cards. He knew I started with A,8 and then caught the 6 and the 3. If he had not known that - if, for example, he had not been sure whether I started with A,8 or A,6 - it would have been impossible for him to conclude with such certainty that I had a pair of aces.

It is in this way that you use logic to read hands. You interpret your opponents' plays on each round, and in open-handed games you note the cards they catch on each round, paying close attention to the order in which they catch them. You then put these two pieces of evidence together - the plays and the up cards - to draw a conclusion about an opponent's most likely hand.

In that high-low split hand, the Q, l O low was able to put me on a specific hand quite early. However, it is generally a mistake to put someone on a specific hand early and then stick to your initial conclusion no matter how things develop. A player who raises on third street in seven-card stud with a king showing may have two kings, but he may also have a small pair in the hole with the king kicker or a three-flush or a J,Q,K or a number of other hands as well. Drawing a narrow, irreversible conclusion early can lead to costly mistakes later, either because you fold with the best hand or because you stay in as a big underdog.

What you do in a game like seven-card stud or hold 'em or razz is to put an opponent on a variety of hands at the start of play, and as the hand continues, you eliminate some of those hands based on his later play and on the cards he catches. Through this process of elimination, you should have a good idea of what that opponent has (or is drawing to.) when the last card is dealt.

Suppose, for instance, in seven-card stud a player starts with a queen of spades, then catches the deuce of spades, then the 7 of spades, then the 5 of hearts, and he's betting all the way. You have a pair of l Os which does not improve. Your opponent bets on the end, and clearly you can beat only a bluff. The question is - might your opponent be bluffing? With something like a four-flush and a small pair, he would probably have played the hand exactly the same way - semi-bluffing right to the end, assuming you didn't catch any dangerous-looking cards. Therefore, while your opponent may, in fact, have a pair of queens or queens up, there's also a chance he has a busted hand. Very possibly you should call his final bet, given the pot odds you're getting - but realizing at the same time that he may indeed have been semi-bluffing yet still caught his hand on the last card.

Suppose, on the other hand, your seven-stud opponent started with that same queen of spades and you with that same pair of lOs. Once again your opponent is betting all the way. But this time he catches the 7 of diamonds, then the 4 of clubs, then the jack of hearts. Now when he bets on the end, you should almost certainly fold your two unimproved l Os because when he caught the 7• and 44 but continued betting, you had to eliminate the flush draw as one of his possible hands. Therefore, he is almost certainly betting on the end for value with at least a pair of queens - more likely two pair. Ironically, it can sometimes occur that because your opponent's hand looks less dangerous on board it is more of a threat to have you beat when your opponent bets on the end, because nothing showing suggests he might have been semi-bluffing as the hand progressed.

At the end of a hand it becomes especially crucial to have a good idea of what your opponent has. The more accurately you can read hands on the end, the better you can decide whether you have, for example, a 20 percent chance of having your opponent beat or a 60 percent chance or whatever. You use your ability to read hands to come up with these percentages and then decide how to play your own hand.

In practice, most players don't arrive at exact figures like 20 percent or 60 percent, but at the very least they try to decide whether their opponent has a bad hand, a mediocre hand, a good hand, or a great hand. Let's say your opponent bets on the end. Usually when a person bets, it represents either a bluff, a good hand, or a great hand, but not a mediocre hand. If your opponent had a mediocre hand, he would probably check. If you have only a mediocre hand yourself, you have to decide what the chances are that your opponent is bluffing and whether those chances warrant a call in relation to the pot odds. If you have a very good hand, you must decide whether your opponent has a good hand or a great hand. If you think the chances are high he has only a good hand, you would raise. But if you think he may very well have a great hand, you would just call. If you are virtually certain he has a great hand, you might even fold your very good hand, depending upon the size of the pot. You ask yourself two questions: What does it look like my opponent is representing? Could he have the hand he's representing and have played it the way he did? Once you draw your conclusions about your opponent's hand on the basis of his play and his up cards, you decide on the basis of your own holding and the size of the pot whether to bet, check, call, raise, or whatever.

We have seen that in open-handed games one way to read hands is to start by considering a variety of possible hands an opponent might have and then eliminate some of those possibilities as the hand develops. A second or, more accurately, a complementary way to read hands is to work backward. It is that sort of thing my high-low split opponent did. If, for instance, the last card in hold 'em is a deuce and an opponent who'd been quiet from the start suddenly bets, you think back on his play in earlier rounds. If there was betting on the flop or on fourth street, that player would not have called with nothing but two 2s in the hole. So he is betting now either as a bluff or has something other than three 2s. If, on the other hand, everyone checked on the flop and on fourth street, it's very possible the player caught three 2s on the end. Every step of the way you must work forward and backward to zero in on your opponent's most likely hand.

• Using Mathematics to Read Hands
• Reading Hands in Multi-Way Pots
• Things To Remember