Sunday, October 5, 2014

Signaling - suit or whole hand?

Signaling is easy when there's only one message to be sent and the same message to be received. In communications, we would say that the protocol is established. Here's a simple example: in a suit contract, partner leads the king of a side suit (king from AK). Dummy comes down with three to the queen. The queen is on view so that the protocol (assuming standard carding) established is that follower's carding will be: high-low if he has a doubleton and wants to ruff; otherwise low-high. High-low here is not count. It's attitude but of a specific type -- it doesn't say you like the suit (you don't) -- it says you can give partner a ruff in the suit. If you had J873 in the suit you would like the suit well enough. If partner continues with the suit, your jack will eventually become the high card. That might be nice in a notrump contract but is completely counter-productive in a suit contract! So, that's why the protocol doesn't cater to that situation. High-low says you can ruff the third round and that's all it says.

Unfortunately, there are also many situations where the protocol isn't so well established. Here's a case in point which arose in a club game against unknown opponents. Partner led a small heart and down came the dummy (click "Next").

I won the first trick with the ten and paused for thought. Partner was leading "my" suit and it looked like she had one or three as I could see all the low spots. If it's a singleton, she can get a ruff, but I need to cash some winners first in case it's declarer who has the singleton. So, I cashed the club king to which I received a discouraging signal. According to my interpretation of the protocol, that means that partner is looking for a heart ruff. But according to partner's interpretation, she just doesn't have anything good in clubs. I continued with the ace and then cashed the heart ace. It was declarer who ruffed but we had two more tricks coming: the trump king and the spade ace, for +500.

On this hand, it made no difference. The heart king that I mistakenly promoted in the dummy could never be used for a discard so all was well. But it got me thinking about the protocol in this situation. With no real clues from the auction, I really felt I needed to know about the heart situation. Why would I care about clubs? -- nothing in the dummy could go away on declarer's good clubs. In other words, I needed a signal that helped with the whole hand, not the suit (clubs) in question.

It turns out that we lost the hand in the auction which perhaps is my fault for not bidding hearts directly at some point. Because of partner's three small hearts, we can actually make four hearts, although several players did not. We scored a decent 8 out of 12, eight pairs having bid game our way with four of them making.

I've written before about signals that simply inform and signals that try to direct the defense: Show and tell -- more on defensive strategy. My thesis was that it is the degree of urgency, typically as evidenced by a strong side suit in dummy, which determines which message should be sent.

Here's another situation from the same session (also against unknown opponents):

You can click on the GIB button to see what's right. But I didn't have that luxury. Both of us could see that dummy had a threatening heart suit. So, to me that triggers the notion of urgency. Can we cash sufficient spade tricks before they can get their heart tricks? Or should we try to knock out the diamond ace early on so that it cannot be used later as an entry to the established hearts?

Partner encouraged spades and I had no reason to believe that she wouldn't be able to cash another three spades after getting in with her presumed entry. So, I continued spades at trick two. Unfortunately, the right defense was to give up on spades for now and knock that diamond out. This resulted in us suffering our only bottom board. Defeating the contract would have been a top.

Again, I probably should have known that a diamond was the most urgent. But, I suspect that opportunities for signaling about the whole hand (that's to say helping to direct the defense) arise quite frequently. We definitely need to be on the same page in these circumstances.


  1. Experts don't seem to agree on this either. If I understand Bobby Wolff correctly, he recommends that experts always signal count because a good player should be able to work out the rest. Richard Pavlicek seems to agree ( Frank Stewart (admittedly geared more towards advanced players than bona-fide experts) seems to say the opposite - for example don't echo when you don't want a ruff.
    This is one of those time to do long as you have an agreement.

  2. In theory, we also have an agreement that when one of us has bid a suit strongly (overcall, rebid, strong jump shift, etc.) then we show count (since attitude is already known). But I think that when matters are urgent, attitude should "trump" count. This stuff is hard!

  3. I prefer that third hand’s encouraging signal says I would prefer that you continue the suit you led rather than switch to another suit, and third hand’s discouraging signal says I would welcome a switch to another suit. What is “another suit”? Usually that is obvious, but when two suits could be the “another suit”, that is where having partnership agreements defining Obvious Shift come into play.

    Some textbooks and teachers say that third hand’s signal should be based upon the holding in the suit led. I think that is quite wrong, because the signal should be based upon the best defense for the hand rather than a particular holding in one suit.

    Except in a very few, generally easy-to-define, situations, third hand’s signal should be, IMHO, attitude and not count. When third hand signals high-low opposite a partner that seems to have led from an AK combination in a suit contract, and third hand is seeking a ruff on the third round of the suit, then third hand has signaled attitude: third hand wants partner to continue the suit led.

    Your defense, Phasmid, was, IMHO, definitely correct on Board 7 (where your partner should encourage clubs if not wanting you to switch back to hearts). And on Board 15, while you do not show your partner’s hand, if your partner preferred a diamond switch to a spade continuation, your partner should have discouraged on your lead of the SA. (In other words, diamonds – dummy’s shortest suit -- is the “another suit” to which your partner should expect you to switch; i.e., the Obvious Shift.) From where I sit, your partner in each of the two hands was corrupted by the tenets described in my second paragraph.

    Some experts espouse “signal what your partner needs to know”. To me, that seems like a potential accident in progress. But then, I am no expert.

    1. Always happy to find some common ground with you, Jeff.

      Still, I think that what the experts espouse is basically what we're talking about: think about signaling in the context the whole hand.

  4. Modify your comment to say "think about signaling ATTITUDE in the context [of] the whole hand" and we would definitely be in agreement.

  5. Although my classes haven't progressed to this stage, in my lesson plans for youth, I teach that attitude signaling is not designed to give partner a report card on the suit your partner led, but rather is designed to give partner a clue as to what partner should lead at his next opportunity.