Count Winners, Count Losers

A common question broached by neophytes is "Should one count winners or losers?" And the traditional answer is that you should count losers in a trump contract, winners in no trump. I can't argue with the latter half of this presumed wisdom, for counting losers in no trump would make no sense. But I don't agree with the former. For, indeed I am partial to counting winners in trump contracts also. And I have a very good reason for being so, which is that in counting outright winners, likely winners and potential winners, you'll almost necessarily be giving yourself some idea of how you want to go about playing the hand.

Oh I won't deny that there are trump hands where it's simply easier to count losers. Particular with strong hands, say, at the five, six or seven level, you'll probably have at least two suits that are solid, needing no development, maybe even three, and you only hafta peruse the hand for any losers you might avoid in one way or another. It's a much shorter count, of course, and might well serve your interests. But below the four level -- I'm letting the four level straddle the two ways to go -- I'd would recommend a count of winners -- those you can cash outright, those you can develop and those you just might develop. For in such a count, you'll almost willy-nilly point to what you hafta do to realize them.

Best of all is to count both! Well, gosh, given the inviolate number of 13 tricks in a hand, wouldn't a count of one give us a count of the other simply by subtracting from 13? Well, I dunno, and that's the point.

Yes, of course if your count is sound, you merely need subtract from 13 to give you the number of winners if your count was of losers or losers if your count was of winners. But that's just the point. You count both to see if your count has been sound. I invite the reader to look at these 26 cards in a four-heart contract:

A Q 7 5

J 8 6 4

8 5

A Q 5

6

A K Q 9 3

9 6 4 3

K 7 3

This is perhaps my favorite hand, not particularly because a declarer kicked a positive score my way, though he indeed did that, but because his line made explicit a not too rare practice of counting a shortage in one hand as the limit of losers there, and the shortage in another suit in the other hand as the limit of losers there. Which works okay if you're loaded for bear in trump and don't get a bad split. But it doesn't work too well if you've got an exiguous trump holding or a bad trump break or both, of course.

Here an incautious declarer took the opening spade lead in dummy with the Ace, mulled over his remaining 24 cards for about five seconds, then with a flourish spread his hand and grandiloquently announced: Drawing trump, conceding two diamonds. Which, though incautious, couldn't have been argued with on a 2-2 break, which declarer didn't test for and didn't have. Indeed, there wasn't even a 3-1, which would've allowed him his contract, though not with an overtrick. So drawing trump in four rounds, on that 4-0 split, would leave him with no trump in dummy, which is to say he has to lose four diamonds, not two.

Incidentally, had declarer taken a spade hook on the opening lead, his troubles would've been over. The finesse is wrong only if it is off AND trump split 2-2. If the finesse wins, obviously it does declarer no harm. If it loses, he still gets use of that Ace to sluff a diamond unless, yes, as just mentioned, unless trump split 2-2, in which case he can't sluff a loser. And even at that, he'd make his contract -- which he didn't do.

A full discussion of the hand is given here for anyone interested but here I only want to pursue my recommendation of counting winners, yea, even in trump contracts. For let us assume that this declarer, obviously not the most adept of players, was not a complete idiot, and that he had proceeded to count his winners before displaying his hand. He would've found that he had a spade already in the kitty, five hearts and three clubs cash for . . . for . . . nine winners! Would that not have been a bit of a wake-up call?

You will note that this declarer, in conceding two diamonds, was in effect claiming 11 winners. Hmm-m-m. But only 9 are immediately cashable. The discrepancy, of course, lies in the ruffing of two diamonds, the third & fourth rounds, which he clearly has to work for and made no effort to do. And so, presuming that he's not a complete idiot, as mentioned above, checking his count of losers with a count of winners, he should've been able to see that one of those counts was unsound. Whether he then would've been able to see the reason for the discrepancy and how to rectify the count to reach 13 is something we'll never know, I guess.

In any event, I say: Don't tell me that if you know the number of losers you need only subtract that number from 13 for the number of winners, for if your count of the former is unsound, then clearly your presumptive count of the latter will be also and counting both serves to double-check your first count.

So best, I am suggesting, is to count both. Is that too much during the moderate haste of play? Well, I'll leave that up to you and only make a few suggestions, to wit: In analyzing a hand in a post-mortem when you have the luxury to mull over a hand without time constraints, a count of both is called for, allowing you to see where you went wrong, if you did indeed boot away a trick or two. Further, I would suggest that you give counting both a try during actual play. After all, we are talking about numbers going only up to 13 and back.

And if you get flustered trying to do both while others wait on you, I would still recommend that you look to where your winners are -- except with those slam-dunk hands where two or three suits are solid. And I would recommend counting from one hand OR the other, figuring on winners in the other hand, of course, but accounting for every card in one hand or the other.

Lemme use the above layout for an example. If declarer takes the first trick with the A of spades and looks to count out from the closed hand, then every card in the closed hand must be a winner or go on a winner in dummy or be looked on as a loser. He sees that spades are taken care of and clubs are solid, and the first two rounds of diamonds are going to be losers whatever he does, but that still leaves two more diamonds. What's he going to do about them. That wasn't part of his thinking. OR . . . he could count out from dummy, if he so chooses. In that event, he could dismiss clubs from mind (except for the entries they represent) and dismiss diamonds also, there being nothing he can do about those two in dummy, but now he has three spades under the A to take care of (in which case he might well consider finessing the Q, for after all it is a potential winner, and anyway you've gotta take care of that Q somehow -- or look on it as a loser).

One hand or the other. Account for every card in that hand, looking to how many can go on winners in the other hand, of course, but accounting for it anyway. But mixing them up, counting spade losers in the closed hand and diamond losers in dummy is just what did this declarer in.

I invite the reader to consult the following categories:
1. Countout. There are numerous hands where the reader can follow how an experienced declarer would view the hand when dummy comes down. Please remember that you could out from one hand OR the other. On the above 26 cards fr'instance. You count out from dummy OR the closed hand. If he wants to count out from dummy, he dismisses diamonds from thought. He can't do anything about those two. But he must take care of three spades under the A. This brings up the problem that he'd be ruffing with some top honors needed to draw trump, so declarer might think of taking a finesse, notwithstanding a singleton opposite an ace. If on the other hand, he wants to count out from the closed hand, clubs are solid, trump are solid, he has a singleton spade opposite the A and can dismiss spades. The first two rounds of diamonds are losers he cannot avoid. But he must do something about the 3rd & fourth rounds of diamonds.

Every card in the closed hand must either be a winner or go on a winner in dummy or be counted as a loser. By the same token, if you count out from dummy . . . but that brings me to the second category I invite the reader to explore:

2. Dummy Reversals. That's the term for hands where declarer counts out his winners and losers from dummy, and where every card must either be a winner or go on a winner in the closed hand or be considered a loser. Often it is simply easier to count out from dummy than from the closed hand.

3. Crossruffs. This is the exception to everything said above. You look for winners, pure and simple. You don't play "from" one hand or the other. Rather, you look for the number of winners you can realize by ruffing back and forth. Look for hands with balanced trump, a short holding in one suit in one hand, a short holding in another suit in the other hand . . . and not least, look for hands where you have a goodly supply of top trump -- where after, say, round two of each side suit, you can ruff with high trump. Well, look at the hands under that category.

And practice with various hands, some from this site, some from the newspaper, some from hands you play. But I do believe that counting winners is what will stand you well on touch-and-go hands, for that will tell you how to develop the winners you need.

	A Q 7 5
	J 8 6 4
	8 5
	A Q 5



	6
	A K Q 9 3
	9 6 4 3
	K 7 3