A asks for Attitude; K asks for Count --discussion against suit cotracts

Most of you will run across this convention sooner or later. One of you asked me to put it on our card. I pretty violently refused that request. Here's why.

First off, this convention is seriously misnamed. It should not be called "A asks for attitude, K asks for count". Rather, this convention should be named (relatively nice) "A requires attitude and K requires count". Or, (not so nice) "A demands attitude and K demands count". Or, my true feeling "the Control Freak Convention".

Now that I've elicited your emotion, let's get to bringing forth some of your bridge logic. Presumably, this convention allows opening leader holding the A/K of a suit to pick a high honor and require partner to signal according to his/her presumed needs. Trouble is 1) opening leader will rarely know from the bidding what his/her needs are and 2) opening leader isn't always dealt the A/K of a suit, rather KQ(x) or Ax(x) only. And another special case is AK tight.

Here are some clarifications.

1) My partner bid diamonds. I held KQx in the suit. I led the K (having no A choice) and dummy decked xxx. Partner dutifully gave me a count card. Big deal, that's not what I needed to know. I needed to know where the A and the J were. I have the same problem if I had led from Kx. Note this is problematic even if pard hasn't bid the suit.

2) Suppose on some auctions you decide to lead the A from A(x)(x). Remember, partner is required by your convention to give an attitude card but here's the problem with that. Partner will not know if you hold the K so is he/she supposed to give you an atittude card for the K or Q? With standard methods you don't know that either but guesses are much more accurate.

3) With standard methods you can often show partner a doubleton holding. For example if you have agreed to lead A form A/K --holding AK only you would simple reverse that position, leading K form AK. Vice versa leading K from AK. With the above-named method one would not know whether partner were showing a doubleton or merely requiring you to give an attitude vs count card.

Suppose you held the golden suit combination for this method. AKx(x).Here's the problem. Often you would not know what you need from partner until you see the dummy. If dummy shows up with the Q, well you obviously don't need an attitude card. You know the attitude. If dummy shows up with xx, then you need an attitude card for the Q, wishing to know whether partner or declarer held the Q.

So with my expert partners I play thus. If one of us leads a high honor and dummy shows Qx(x)(x), partner gives a count card. If dummy does not show a Q, partner requires to give opening leader an attitude card for the Q.

How does this last treatment differ from the above-named convention? Simple, opening leader is not assuming he/she knows what is needed before seeing dummy. In otherwords he/she waits for one more piece of information to start the D going in the right direction.

Here's a perfect "mess up" using this convention. Opening leader holding Q85,AK954,Q4,J62 heard the auction go (1D) 1H (3D!) 4H, (5D) P (P) X at all nv imps. He decided that before the dummy came down he needed a count card from pard. HK. (side note, I don't know why, perhaps he thought there were a side suit somewhere). At any rate, the dummy decked J76,87,T9872,QT5 and partner dutifully gave him a "count" card from 942,JT3,6,AK8743. Not knowing what to do, opening leader continued hearts. I knew what to do, pitch a spade from dummy and claim down one -100 only. My hand: AKT3,Q62,AKJ53,9. This is clearly a mistake, the second heart should never have been cashed. With me? Partner would have led A or K per agreement. I would have offered the J denying the Q and partner finds a switch. Now the HQ
never sets up for a pitch.

Now, believe it or not, I've played this treatment with partners who have insisted on doing so.(Yes, even I can be that acquiescent). I'm convinced that it "works" when you hold AKQ in a suit and you "know" the attitude for the Q. So in this case, you clearly request a count card from pard. Is this slight feature, not available in standard methods, worth giving up the non-confusing and solid D that other more "traditional" methods provide.

I think not.

Let me know if you have any questions.

Isolde 8500