[awm]

It seems to me that there are two basic approaches to carding in an expert partnership.

(1) Have a set of rules about your signals (i.e. when is it attitude/count/suit preference). These rules may have to be complicated. Follow these rules even though occasionally they will not be best... you accept that sometimes you cannot give the most useful signal, but in exchange partner will always know what you are signaling in each situation.

(2) Try to signal what you think partner needs to know. This may be a simpler approach, but there will be times when you and partner disagree as to which signal should be made or is most useful.

If you have some other approach feel free to mention that too.

[Phil]

I'm not sure 1 and 2 are mutually exclusive, but my style is certainly #1. Partnerships that practice #2 are frequently 'winging it', and invariably there are conflicts.

Even if you practice #1, there are positions that do not fit into a particular 'box', just like there are auctions where the partnership is in unfamiliar territory.

If I am discussing these matters, I also want to know how much I can trust my partner's signals on any given hand, or if I am frequently 'on my own'.

[dake50]

Have formal rules about what/when partner needs to know.
You are starting from the "I want to signal something"
instead of "We expect to need a signal here."
Eg. count to help holdup - I think is needed.
A 3rd count signal after a 2-suiter has been shown is not.

[JLOGIC]

2

[nigel_k]

1

I'm not that good at guessing what partner is guessing I have guessed.

[JLOGIC]

lol, guessing.

[the hog]

Prefer 1. Agree with Nigel. Have a preference for count.

[marcD]

Think 1 is better except for really top level partnerships . There are also a lot of tempo issues associated with 2 which are difficult to ignore even if you practice active ethics.

[JLOGIC]

How many new carding situations do you see at trick 1?

They are all thematic, and the problematic ones are well known (ie, partner leads the SK against a suit contract, dummy has Jxxx, and you have Qxx. Obv you discourage so that partner doesn't try to give you a ruff with AKxx setting up the jack. This is a problem the first time you see it but if you ever play bridge you have seen this numerous times and don't have to think about it. Or partner leads the ace against 3N and dummy has xx. Do you encourage with Jxxxx or not? If you do partner might underlead AKTx, if you don't partner might shift, etc etc.)

Obviously you cannot have no discussions or experience with partner, but you don't need a rule for all situations, most of the time it is clear what partner needs. I don't recall ever having a difference of opinion with my partners about what the best signal on a certain hand would be at trick 1. The ones where they are tough, there is no way you can have a good rule to cover it because it will be weird/obscure. Having a general approach is normal (for instance one poster said usually count. I don't know how that's playable but more power to them).

There are too many situations to cover, but assuming your partner and you can both use good bridge logic it is almost always obvious what signal is needed. Like, partner leads the ace vs 3N and dummy has QJTx. I would give suit preference with anyone, but I have not discussed this with anyone. I just don't see what else anyone would give. Or partner leads the king vs 4S and dummy has QJTx. If it is a random suit with no info about it, I will signal whether I can get a ruff or not. However, if it is partner's known 5+ card suit and I have raised, I will give suit preference. If partner has shown 4, and I could have 3 or 4, I would give count usually, but depending on the hand I might give suit preference if that is what's needed (and this all depends on the bidding, the contract, the form of scoring, and what dummy is. Maybe dummy will make it obvious which suit to shift to if there is a shift to be made, so I'd just give count. Etc Etc). Partner will know what he needs so it shouldn't be an issue. etc etc. If I gave a slow signal in the last situation, there is not much info given, because I am probably just thinking which signal is needed. I don't see the ethical problem there.

And let me re-iterate, of course at the beginning of a partnership it's good to discuss carding situations in general, and then when you play discuss specific situations that come up, and make sure you're thinking similarly. But I cannot imagine having a rule based system for every scenario which cannot cover all scenarios, and when one comes up where your rules are obviously suboptimal for that specific scenario, you are forced to make the suboptimal signal even though both you and your partner have a brain and can see IN THAT SCENARIO, WHILE IT'S HAPPENING, that your rule is wrong and you should be doing something else. And if you are overriding your rules in that scenario then you are playing #2 not #1 imo.

[gnasher]

1, but the rules define when and how we may deviate from the other rules.

Or 2, but there are still rules to cover what we might be signalling, what the normal signal is, and what might make us vary from the normal signal.

But most of all, you should discuss it properly.

[rhm]

(2) may look simpler, but is in fact much more complex. It is anything but easy in Bridge to anticipate your partner's problems in defense, some standard defense situations like an entry-less dummy notwithstanding. In practice most players are already overwhelmed finding the right defense for themselves let alone looking at it from partner's perspective.
My guess is that (2) is disaster prone unless two very good and experienced players form a long standing partnership and have gone through some signaling disasters already.

The trouble with (1) is of course that signaling like any form of information exchange in Bridge may be more helpful to opponents, in this case declarer, than it is to partner.

In my opinion there is also an ethical problem with (2) seldom discussed: Long standing partnerships will know when partner's signal will be honest and can be trusted. This is an implicit understanding coming from experience. In effect you have the same rules as in (1), only never formalized. Declarer is supposed to have the same information, but rarely does. Try to find that out as declarer at the bridge table. Asking defenders will already telegraph the problem you may have as declarer to them. If you ask you will often get an answer like: "If my partner would signal in this situation it would mean that and that, but he might not want to do that".
So as declarer you are often better off not asking at all, but it is not fair.

Rainer Herrmann

[Free]

1 is definitely easier on the mind, but more frustrating if you have to do something which you know will turn out badly. 2 is most flexible, which is desirable imo. You need some strict rules, but using logic in some cases should deliver the best results.

[rhm]

Free, on 2011-March-28, 03:01, said:

You need some strict rules, but using logic in some cases should deliver the best results.

If there would exist only a single logic I would agree. Unfortunately from many postmortem discussions, logic can be applied from many different perspectives.

As an example

JLOGIC, on 2011-March-28, 02:17, said:

How many new carding situations do you see at trick 1?

(ie, partner leads the SK against a suit contract, dummy has Jxxx, and you have Qxx. Obv you discourage so that partner doesn't try to give you a ruff with AKxx setting up the jack.

Why is this obvious? You might just as well consider it obvious to give count and if partner deciphers it as count the problem is solved by different logic. But if one considers it obvious to give attitude and the other considers deciphering it as count you may be in trouble.

Another contentious one: Playing a suit contract you lead an ace and dummy comes down with a singleton.
I like to play suit preference here. But as usual and no surprise many disagree. It might well be right to continue the suit and force the dummy. So they prefer attitude. Again it is not clear which logic should apply.

Rainer Herrmann

[helene_t]

Isn't it similar to how a partnership should define doubles? I.e. the default meaning of a double/signal is "whatever would be most useful in this situation". And then you make specific agreements for a lot of situations where it is not obvious what is best. But you still use the default rule whenever you are in a situation not covered by the agreements.

[bluecalm]

Quote

Do you encourage with Jxxxx or not? If you do partner might underlead AKTx, if you don't partner might shift, etc etc.)

So what do you do ? Bidding 1NT - 3NT

[awm]

To Helene: the analog to (1) would be that we discuss a lot of situations, and then have some general rule like "undiscussed doubles of suit contracts below 2NT are takeout" -- this occasionally gets us in trouble when there is some unusual auction we have not discussed and we really want to double for penalty but can't because of this agreement. The analog to (2) is basically that for any undiscussed double, partner is supposed to figure out the meaning by looking at the auction and his own holding in the suit. This will usually let you double for takeout when you want and double for penalty when you want, but very occasionally there will be a situation where partner really cannot tell what you mean (i.e. opponents have a much bigger fit than implicit in the bidding) and you have a disaster.

Anyway, perhaps here's a common situation. Say declarer wins the opening lead in 3NT and leads the ♦K, dummy having ♦QJTx. If dummy has no entry, everyone gives count (I hope!). If dummy has a lot of entries then count is arguably less useful, so maybe you give suit preference. But what if dummy has exactly one side entry? What if dummy has some side cards which may or may not be entries and only one defender can really tell? A type (1) agreement might be to say that we always give count in these situations, or that we give count unless dummy has a side card which is a totally obvious entry.... and accept that occasionally this means there is a situation where dummy has a lot of "stuff" and a suit preference signal might be more useful but you signaled count. A type (2) approach would be that you look at dummy and decide whether dummy has "enough" clear/possible entries that you can signal suit preference, and if not you signal count. Obviously if the type (1) players have a decent rule you will usually get the same result. But there will be some cases where the rule causes the type (1) player to give a less useful signal. And there will be an even smaller number of cases where the type (2) players have a serious accident because one person thinks the signal is suit preference and the other thinks count.

[Lurpoa]

Signalling what partner needs... that is the first thing...

No doubt.

[bluecalm]

We should have separate thread with specific examples. Would be very educational to have discussion on those things.

[JLOGIC]

rhm, on 2011-March-28, 03:26, said:

If there would exist only a single logic I would agree. Unfortunately from many postmortem discussions, logic can be applied from many different perspectives.

As an example


Why is this obvious? You might just as well consider it obvious to give count and if partner deciphers it as count the problem is solved by different logic. But if one considers it obvious to give attitude and the other considers deciphering it as count you may be in trouble.

Another contentious one: Playing a suit contract you lead an ace and dummy comes down with a singleton.
I like to play suit preference here. But as usual and no surprise many disagree. It might well be right to continue the suit and force the dummy. So they prefer attitude. Again it is not clear which logic should apply.

Rainer Herrmann

Lol. Do you read what other people write? I do not advocate not knowing whether you play COUNT or ATTITUDE as your primary signal, or whether you play SUIT PREFERENCE at trick 1 with a singleton on the board.

I know it is super obvious that I do not advocate those things by advocating 2 instead of 1, however I added in this paragraph for people like you:

Quote

And let me re-iterate, of course at the beginning of a partnership it's good to discuss carding situations in general, and then when you play discuss specific situations that come up, and make sure you're thinking similarly. But I cannot imagine having a rule based system for every scenario which cannot cover all scenarios, and when one comes up where your rules are obviously suboptimal for that specific scenario, you are forced to make the suboptimal signal even though both you and your partner have a brain and can see IN THAT SCENARIO, WHILE IT'S HAPPENING, that your rule is wrong and you should be doing something else. And if you are overriding your rules in that scenario then you are playing #2 not #1 imo.

Giving count is inferior in general to attitude for many reasons, in my example it would be because with 4 small you want to discourage, and with QYxx where Y can beat dummies 2nd card you will often want to encourage. But if my general agreement was COUNT, I would give count and it would be a non problem situation. The point was, that these are all thematic and come up all the time, and are non problems to an experienced bridge player. They will not cause tempo/ethical issues.

[akhare]

My preference is somewhere between 1 and 2.

It's good to have some fixed rules for say opening leads and T1 situations (example: obvious shift). However, partner is expected to apply bridge logic and treat the signal as a "suggestion" and not a command.

At the same time, I don't believe spoon feeding partner (like always providing an inveterate "true count" on declarer's leads all the time)...