opportunities for encryption in bridge

[Cascade]

The point is that the disclosure is encrypted and unless you have the key you cannot decipher it.

 

This could be crucial as you might wish to lead through the king or ace in dummy but not around to the honour if it is held by declarer.

 

It is full disclosure it is just less useful information for the defenders.  It is not less useful for the bidding side since they have the key - they know who has the ♠A.

 

Now that is helpful....confirming that the whole idea is to beat the system (full disclosure), not to improve slam bidding. I am no longer ignorant.

That is a twist.

 

It is full disclosure it is just that the information we are choosing to exchange is more useful to our side than it is to the opponents.

 

Everyone practices or attempts to practice this. e.g. we don't determinedly give count when we judge that it is more benefit to the opponents than it is to partner.

 

This is just another tool to that same end.

[Mbodell]

One simple place for encryption that I've used is a form of 2-way drury where 2♣ shows a limit raise with 1 of the top 2 trump and 2♦ shows a limit raise with 0 or 2 of the top trump. Over 2♣ you can encrypt your game tries based on who has the A and who has the K (our simple rule was the person with the A bids "normally" while the person with the K reverses normal in a sort of 2-way shortness or length try). You can use 2♦ over 2♣ to say I don't have the other top card and want to make a "normal" game try non-encrypted. You can also decide to try the encrypted tries over the 2♦ initial response on the theory that it is rare that both players are missing both top honors.

 

Overall I'd say it is a net loss in that a natural nf 2♣ or 2♦ might be better, but it is fun.

 

It isn't about trying to get around full disclosure. It is about trying to not leak information. There is a difference as opponents are entitled to our agreements, understanding, and style issues. They are not entitled to what is in my hand.

[Cascade]

Drury a net loss - why am I not surprised?

[rbforster]

I am no longer ignorant.

Sadly I see we've already got an example of how poor understanding and common misconceptions about encryption that lead these methods to be banned. Full disclosure does not require showing your opponents your hand - sometimes they can figure things out, and sometimes they can't and that's just how it is.

 

1♠-3♠-4♣-4♦...6♠

 

opp "What does 4♦ show?"

me "cuebid showing A or K of diamonds"

opp "Which? You must tell me if you have the other so I can lead the suit when you don't since I'll know my partner has it"

 

How ridiculous is that? The bid shows one of two possibilities and if they can't figure it out by looking at their hand, too bad for them. This is exactly the same situation as the encryption examples:

 

4N-5♦

5N-6♦ showing either (A♠+K♦) or (no A♠, but K♣)

 

Again the response shows one of two possibilities. You give a full explanation of what the bid shows in your system, and if that information not helpful to them then that's not your problem. In fact, it's probably a sign that your system doesn't give away unnecessary information to aid the defense. Encryption is just a way of making this more likely.

 

Two-way game tries are a similar example of reducing information given to the defense during the bidding:

 

1♠-2♠

2N*-3x * what is the cheapest help suit you would accept?

4♠

 

Compare this with a "normal" auction where opener makes a natural bid showing values in a suit. The Kokish game try players are effectively hiding opener's hand. Are you going to demand that these players have opener reveal his best side suit to the defense just so they can know what not to lead because it offends your sense of fairness?

[hrothgar]

The point is that the disclosure is encrypted and unless you have the key you cannot decipher it.

 

This could be crucial as you might wish to lead through the king or ace in dummy but not around to the honour if it is held by declarer.

 

It is full disclosure it is just less useful information for the defenders.  It is not less useful for the bidding side since they have the key - they know who has the ♠A.

 

Now that is helpful....confirming that the whole idea is to beat the system (full disclosure), not to improve slam bidding. I am no longer ignorant.

Huh...

 

My reaction is that you are still quite ignorant and showing genuine signs of stupidity.

 

Disclosure has a very specific meaning. There is a fundamental difference between

 

1. Disclosing your agreements

2. Describing the contents of your hand

 

You don't seem to grok the difference.

 

If you frame your discussion in terms of "minimizing the amount of information disclosed to the opponents" rather than "beat disclosure" your posts probably won't get quite as vocal responses.

[akhare]

Now that is helpful....confirming that the whole idea is to beat the system (full disclosure), not to improve slam bidding. I am no longer ignorant.

Hmm..., there's a huge difference between "hiding information" and "hidden agreements / hidden methods".

 

In this case, the methods (the encryption algorithm if you will) are being completely disclosed to the opps, but the information that's encrypted (or specifically the key used to encrypt the information) isn't.

 

Consider the following hypothetical relay auction:

 

1S - 1N* - 2C - 2♦* .... 6♦ * = relay

 

In effect, responder has skillfully used relays to improve slam bidding and has effectively obfuscated his own hand. Basically, at the end of the auction you have absolutely no information about declarer's hand.

 

The methods being discussed here are just an extension of the information obfuscation principle. As before, the methods themselves (relays, etc.) are being completely disclosed, the information that's transmitted isn't.

 

If you look around, you will find numerous instances of conventions tailored to obfuscate information and this is simply an extension of those methods.

[aguahombre]

I guess I am the only one who believes there is a difference between systems where one player asks questions without revealing to anyone, including partner, what his/her hand is --and systems where one player asks questions and the answers are encrypted so that he/she is the only one who knows what the answers are and doesn't have to tell the opponents --or gets to tell the opponents cryptically. If thinking that is against the spirit of disclosure is stupid, I proudly accept the label.

[akhare]

I guess I am the only one who believes there is a difference between systems where one player asks questions without revealing to anyone, including partner, what his/her hand is --and systems where one player asks questions and the answers are encrypted so that he/she is the only one who knows what the answers are and doesn't have to tell the opponents --or gets to tell the opponents cryptically. If thinking that is against the spirit of disclosure is stupid, I proudly accept the label.

In that case, do you have objections to DCB responses too? In a lot of cases the (correct) disclosure at the end of the auction goes something like "he either has AKQ of <blah> or nothing...either A/K of blah...etc".

 

Of course, relayer often has complete knowledge of the exact holding. In other words, the relayer knows the exact answer and is telling the information cryptically to the opps.

 

Is this against the spirit of full disclosure also? If not, how does it differ?

[oxyde]

4NT 5♣ = 1 or 4 keys

 

Do I have to tell it is 1 or 4 ?

If not, then door is open to encryption.

I am not fully sure but I guess we do not have to reveal, so...

 

------------------------

But to come back to initial aim of the topic

 

Such a method is based on the asumption that we are able to share the key :

1) Are we supposed to build the key based on our hand only ?

This I think is regulated and I guess answer is yes.

At least there must be a sentence somewhere in the laws stating that bidding shoud be based on our hand only.

 

If answer to 1) is Yes

It means we have to lose bidding space to first share between us the key.

2) Basically this means we have a less efficient bidding system.

Nothing to debate there.

3) How could we ensure that we both have understood the key ?

Tricky one. OK when talking about BW because we know we will stop if missing 2 keys (something like a CheckSum). But in other cases ?

[rbforster]

Such a method is based on the asumption that we are able to share the key :

1) Are we supposed to build the key based on our hand only ?

This I think is regulated and I guess answer is yes.

At least there must be a sentence somewhere in the laws stating that bidding shoud be based on our hand only.

I agree. I believe this falls under playing the same system throughout the event. This precludes things like basing your bidding on the prior hand's outcome (at the last table) and other such things I think everyone agrees are not allowed.

 

It means we have to lose bidding space to first share between us the key.

2) Basically this means we have a less efficient bidding system.

Nothing to debate there.

It's true you must allocate bidding space to attempting to establish a key. This may not be successful either, so space could be wasted. The efficiency of the bidding system however should include whatever advantages you gain from successful encryption, the probability of favorable leads or less good defense that might arise from giving away less information. Overall, I think that you're correct however that typically encryption will lead to a less efficient bidding system. That said, there are some opportunities (like Blackwood) where the costs are low or even zero and others where some bids didn't have useful meanings anyway so it's not a big cost to allocated some of those bids to attempting encryption.

 

3) How could we ensure that we both have understood the key ?

Tricky one. OK when talking about BW because we know we will stop if missing 2 keys (something like a CheckSum). But in other cases ?

Typically you'll need a response, perhaps the cheapest step, to confirm that the key has been understood and initiate an encrypted dialogue. When the key fails to be understood ("I have the A or K", "I don't have the other"), now you can either revert to natural bidding, having lost a little space, or perhaps try again to re-establish a key. In addition, you will have told the defense (at least whichever holds the other honor) information that your partner won't have.

[aguahombre]

I guess I am the only one who believes there is a difference between systems where one player asks questions without revealing to anyone, including partner, what his/her hand is --and systems where one player asks questions and the answers are encrypted so that he/she is the only one who knows what the answers are and doesn't have to tell the opponents --or gets to tell the opponents cryptically.  If thinking that is against the spirit of disclosure is stupid, I proudly accept the label.

In that case, do you have objections to DCB responses too? In a lot of cases the (correct) disclosure at the end of the auction goes something like "he either has AKQ of <blah> or nothing...either A/K of blah...etc".

 

Of course, relayer often has complete knowledge of the exact holding. In other words, the relayer knows the exact answer and is telling the information cryptically to the opps.

 

Is this against the spirit of full disclosure also? If not, how does it differ?

the difference is that if the answer was --in fact -- an either/or answer, and the asker knows which it is because of his own hand, disclosure that the answer showed either/or is fine. But if the answer was exact, based on some code related to a holding in a different suit, and that code is only known to the side with the holdings in that suit --then full disclosure should be "his answer showed the Club King". this would not violate the bidding side's right to keep their holding in the code suit to themselves. They simply don't reveal the encryption method, only what the answer meant.

[hrothgar]

I guess I am the only one who believes there is a difference between systems where one player asks questions without revealing to anyone, including partner, what his/her hand is --and systems where one player asks questions and the answers are encrypted so that he/she is the only one who knows what the answers are and doesn't have to tell the opponents --or gets to tell the opponents cryptically.  If thinking that is against the spirit of disclosure is stupid, I proudly accept the label.

You're welcome to whatever beliefs you want. However, you should be very careful not to accuse people of practicing poor disclosure just because you disapprove of their methods.

 

As an example, I might (privately) believe that you like to bugger small children. However, I don't call you "aguahombre the pedophile" on this newsgroup. My subjective beliefs are quite irrelevent to the discussion at hand and labelling you as a pedophile really isn't a good way to engage in polite discourse.

 

In much the same vein, you have skewed and inaccurate ideas about the rules regarding disclosure. You should probably refrain from labeling people based on these, equally subjective, theories.

[hrothgar]

the difference is that if the answer was --in fact -- an either/or answer, and the asker knows which it is because of his own hand, disclosure that the answer showed either/or is fine.  But if the answer was exact, based on some code related to a holding in a different suit, and that code is only known to the side with the holdings in that suit --then full disclosure should be "his answer showed the Club King".  this would not violate the bidding side's right to keep their holding in the code suit to themselves. They simply don't reveal the encryption method, only what the answer meant.

As interesting as this answer might be, it is the complete opposite of how the rules of disclosure actually work...

[akhare]

But if the answer was exact, based on some code related to a holding in a different suit, and that code is only known to the side with the holdings in that suit --then full disclosure should be "his answer showed the Club King".  this would not violate the bidding side's right to keep their holding in the code suit to themselves. They simply don't reveal the encryption method, only what the answer meant.

Not revealing the encryption / encoding method is a violation of Full Disclosure. The "how" of the transmitted information must be revealed, the "what" i.e., specific contents of the message may not be obvious to all involved parties, depending on their specific holdings.

 

You can think of the encoded information a boolean expression, i.e., if (X == true) then Y else Z.

 

The boolean expression and the possible output values (Y or Z) must be revealed, but there's no obligation to reveal whether X is true or false...

[OleBerg]

It means we have to lose bidding space to first share between us the key.

2) Basically this means we have a less efficient bidding system.

Nothing to debate there.

Sometimes you might get the key for free.

 

Assume a relay-system with this response-scheme to ace-asks (Taken from Viking-club):

 

1.step = 0 or 3

2.step = 1 or 4

3.step = Suited aces

4.step = Minor or Major aces

5.step = Hard or soft aces

 

Anytime a 16+ hand has asked for aces, and partner ha replied step 3, 4 or 5, you have a free encryption for your further relays.

 

 

Come to think of it, this corresponds to the 4NT - 1 Ace, 5NT scenario described earlier.

 

But since I've made the post, I'll post it.

[rbforster]

Here's a link to my version of encrypted Drury, as I described previously. It shows how you can use encryption to both help conceal game tries, as well as to make it less clear whether a 3rd seat hand has opened light or not. Note in particular that the two "sign off sequences" that include the light opening hand could also be a sound opener showing a balanced minimum and declining partner's invitation. Balancer beware if you guess wrong which one opener has shown.

[awm]

I'm not that convinced by the encrypted drury argument. The issue is that opponents have already passed in second chair, passed over the third seat opening, passed over the drury call (even though it allows for some cheap bids below 2M). In some cases they have also passed over opener's 2♦ rebid. The odds that opponents want to balance here after passing at so many safe opportunities are not really all that high, and the nature of opener's hand will be clear after the opening lead in any case (since encryption will be broken when dummy is exposed).

 

As for the encrypted game tries, do you find that opponents typically make much better opening leads based on the knowledge of what your game try was? I don't find that this matters all that much. A "long suit" game try can be based on a strong side suit missing an honor or on a much weaker holding. It's not like slam bidding where knowledge of "who holds which ace" often helps the opponents substantially on lead.

 

Honestly I'd rather have a natural 2/1 bid available in one of the two minors than have this encryption.

 

This problem is pretty common with encryption schemes for things other than slam bidding. Encrypted bidding only helps during the auction and on the opening lead (since the key is blown when the dummy is exposed). But in a real competitive auction you are often cramped for space (unlike a drury sequence where opponents are passing and passing again) so establishing a key is really hard. So it's primarily an opening lead thing, and a lot of the information that can potentially be hidden doesn't compellingly effect the opponents on lead anyway.

[Cascade]

I guess I am the only one who believes there is a difference between systems where one player asks questions without revealing to anyone, including partner, what his/her hand is --and systems where one player asks questions and the answers are encrypted so that he/she is the only one who knows what the answers are and doesn't have to tell the opponents --or gets to tell the opponents cryptically.  If thinking that is against the spirit of disclosure is stupid, I proudly accept the label.

In that case, do you have objections to DCB responses too? In a lot of cases the (correct) disclosure at the end of the auction goes something like "he either has AKQ of <blah> or nothing...either A/K of blah...etc".

 

Of course, relayer often has complete knowledge of the exact holding. In other words, the relayer knows the exact answer and is telling the information cryptically to the opps.

 

Is this against the spirit of full disclosure also? If not, how does it differ?

the difference is that if the answer was --in fact -- an either/or answer, and the asker knows which it is because of his own hand, disclosure that the answer showed either/or is fine. But if the answer was exact, based on some code related to a holding in a different suit, and that code is only known to the side with the holdings in that suit --then full disclosure should be "his answer showed the Club King". this would not violate the bidding side's right to keep their holding in the code suit to themselves. They simply don't reveal the encryption method, only what the answer meant.

I assume therefore that everytime you bid Blackwood and then continue to 5NT or its equivalent confirming possession of all key-cards that you disclose explicitly which aces partner has if asked.

 

If not then the only difference between what you do and what another pair do who use encryption is that they utilize the fact that the opponents do not know their holdings while you do not. Your disclosure is identical to theirs.

[rbforster]

I'm not that convinced by the encrypted drury argument. The issue is that opponents have already passed in second chair, passed over the third seat opening, passed over the drury call (even though it allows for some cheap bids below 2M).

Of course all of this is required for a normal Drury rebid by opener to be made uncontested. I agree we're talking about a less likely auction at this point, but I don't think that's a flaw in the methods, just that the methods apply to a somewhat unlikely situation.

 

...In some cases they have also passed over opener's 2♦ rebid. The odds that opponents want to balance here after passing at so many safe opportunities are not really all that high, and the nature of opener's hand will be clear after the opening lead in any case (since encryption will be broken when dummy is exposed).

 

Honestly I'd rather have a natural 2/1 bid available in one of the two minors than have this encryption.

Consider a "slightly too weak to X" takeout of hearts hand in 4th chair:

 

[hv=s=sqxxxhxxdajxckxxx]133|100|[/hv]

 

After (P)-P-(1♥)-?, you judge to pass (or adjust the hand accordingly to be a maximum pass with takeout shape). The opponents continue:

 

P-P-1♥-P

2♣*-P-2♥*-?

 

If 2♣ is regular Drury and 2♥ shows a light opener, you are probably willing to double here, at least at some colors. However, if 2♣ is encrypted Drury and now 2♥ shows either a light opener with the A♥ or a balanced 11-13 with the K♥ (not interested in game), you're risking a lot more by doubling. Furthermore, you know that since you're the only one from your side with shortness, the auction will end in 2♥ if you don't double. Note that here the side wanting to bid hasn't really had more than it's initial opportunity to pass and this is clearly a much more difficult position for the 4th chair than normal Drury.

 

All of this is not to say I don't agree with you that a natural 2/1 in a minor might not be better or that there's little value gained in the lead, but I just want to make the point that encryption can create harder bidding problems for other side even if it doesn't do anything else.

 

As for the encrypted game tries, do you find that opponents typically make much better opening leads based on the knowledge of what your game try was? I don't find that this matters all that much. A "long suit" game try can be based on a strong side suit missing an honor or on a much weaker holding. It's not like slam bidding where knowledge of "who holds which ace" often helps the opponents substantially on lead.

Certainly the slam case is more clear, but I wouldn't typically lead a long or help suit bid by opener as a defender, and I would expect that if I lead it (by accident against a less informative auction) it would probably not work out as well as other choices. I certainly haven't done a rigorous study of this, but this is my impression.

 

This problem is pretty common with encryption schemes for things other than slam bidding. Encrypted bidding only helps during the auction and on the opening lead (since the key is blown when the dummy is exposed). But in a real competitive auction you are often cramped for space (unlike a drury sequence where opponents are passing and passing again) so establishing a key is really hard. So it's primarily an opening lead thing, and a lot of the information that can potentially be hidden doesn't compellingly effect the opponents on lead anyway.

I agree that opportunities are not that common to apply encryption in competitive bidding, but the Rosencrypt double is a good example. Once you show a high honor in partner's suit, you can now encrypt your strong raise and your competitive raise during the bidding. Of course partner won't know if he doesn't hold the other honor, but he often will most of the time for his overcall.

 

Perhaps if people knew more about the benefits of encryption, they would arrange to have bidding situations come up that show encrypted meanings.

[oxyde]

I think I have a good one :

- if partner showed 12H+ in the bidding and the answer to our BW was the low choice (0 out of 0/3 for instance), then we answer to King question the first King we do NOT have. Otherwise we answer the King we have.

 

You might very well advocate that this is for COMPRESSION purposes as your method assume partner is more supposed to have 2 than 1 when having the low choice before.

Don't see why this would be not allowed.

[Finch]

As we seem to have got on to the disclosure discussion, here's a poser for you.

You are playing a pairs tournament with two-board rounds. You don't bother looking at the opponents' convention card. One of your opponents is declarer on the first board, and it is played out without comment.

 

On the second board, you are declarer. You ask your LHO about their carding. You are told

"We play standard count & attitude if the first card below a ten played from dummy on the previous board was even; otherwise we play UDCA"

 

You can't remember what the first non-honour played from dummy was on the previous board.

 

Or even, the key is based on the parity of the first pip played from dummy on this board. Lots of people will have difficulty remembering this.

 

(Note this isn't the same as basing your key on a hand played against different opponents)

 

Is this legal? Should it be legal? [Obviously it won't work more than once.]

 

I can't see anything in EBU regulations that makes it not legal, but it makes me uncomfortable.

 

And this is coming from someone who thinks that 'normal' encrypted signalling should be legal.

[oxyde]

1) As you are playing a non-standard carding system, you should inform your opponents as soon as you come to their table.

 

2) Again, you are using a key that is not based on the current board, so I guess it is not allowed - or at least it will never be allowed if some encryption was to become allowed.

[Finch]

1) As you are playing a non-standard carding system, you should inform your opponents as soon as you come to their table.

 

2) Again, you are using a key that is not based on the current board, so I guess it is not allowed - or at least it will never be allowed if some encryption was to become allowed.

1) There is nothing in the EBU regulations that says I have to inform the opponents verbally, as long as it is clearly marked on my convention card.

 

2) The method I propose is not encrypted, in that the key is in principle freely available to declarer. So I can't see any regulation that forbids it. The EBU does have a regulation that I think could be used to prevent a similar agreement in the auction (e.g. we play weak or strong NT depending on a similar key) but that only applies to calls, not to carding agreements.

[helene_t]

Is this legal? Should it be legal? [Obviously it won't work more than once.]

My guess is that it would be illegal to use a key from the previous board, but probably legal to use a key from the current board.

 

At least, that is how I think it should be. If the key is not related to the current board, you must say what carding you are playing. Even though the first board was played against the same opps, it should be considered external information. Like saying that our carding depends on whether the 357th decimal in pi is odd or even.

[Echognome]

1) As you are playing a non-standard carding system, you should inform your opponents as soon as you come to their table.

 

2) Again, you are using a key that is not based on the current board, so I guess it is not allowed - or at least it will never be allowed if some encryption was to become allowed.

1) There is nothing in the EBU regulations that says I have to inform the opponents verbally, as long as it is clearly marked on my convention card.

 

2) The method I propose is not encrypted, in that the key is in principle freely available to declarer. So I can't see any regulation that forbids it. The EBU does have a regulation that I think could be used to prevent a similar agreement in the auction (e.g. we play weak or strong NT depending on a similar key) but that only applies to calls, not to carding agreements.

I first heard about this as a story about Brian Senior either doing this or saying this would be a funny method. I cannot verify that it actually came from Brian or that he ever played it.

 

However, I can say that my partner and I tried it out for a little while. We did it so that we could concentrate on pips and, I'm sure, for the comedic/annoyance factor to the opponents. We did put the method in bold and highlighted both on the front of our cards and in the signalling area in the back.

 

We played it for about a month or two. Most people never noticed or cared. We often forgot at least one board a session. It did have an upshot of having to figure out the best way to defend without being able to rely on your partner's signal. The one or two times opponents did notice, they just rolled their eyes and played on. So we just scrapped it and went with upside down signals.