[fromageGB]

smerriman, on 2017-December-18, 14:31, said:

I find it hard to imagine a hand where you believe it's possible for you to only hold 2 keycards between the two of you, yet at the same time also believe it's possible you have enough combined strength for a grand slam. [Adjusted by subsequent comment]

The sort of hand where this happens is usually a distributional one, I think. I can never remember hands, but constructing a simple example that does not involve asking for Kings (which would be easier) :

Red against Green, partner opens 1♠ and RHO (a sensible player) overcalls 4♥. He can be expected to have about KQJT in 8 hearts. You too are distributional, with
♠ Kxxx
♥ x
♦ AKQJxx
♣ xx
and it is your bid. You have quite a few tricks. You have 2 of the 5 aces.

4NT RKCB is obvious. IF partner has 3 aces, then you can check for the trump Q, then bid 7♠.
If that Q is missing, 6♠ is probably solid.

If partner has 2 aces, then 6♠ is an excellent chance.
If partner has just one ace, then 5♠ is a good spot, although you hope his clubs are guarded.

You bid 4NT, it is passed to opener who bids 5♣. Your lucky day, so check for the Q (yes), bid 7♠.

* * *
That's quite reasonable, isn't it?

* * *
Partner has
♠QJTxx
♥Kx
♦xx
♣KQJx
and they take their 3 aces for 3 down.

It is that "0 or 3" ambiguity that is the problem. "0 or 4" is much safer.

[fromageGB]

smerriman, on 2017-December-18, 14:31, said:

But surely the 04/15/2/3 approach is going to cause much more regular problems when you have no space to explore after a 3 keycard response.

Exploring what? You have all the space above the "3" reply for Kings etc if you are happy with 6 and considering 7. This is the current "simple" method, and I would suggest retaining it, not going to flawed RKCB, but moving later to something better involving an asking bid of perhaps 4T+1 (next step up from trumps) that has no ambiguity.

Alternatively, if the trump Q is your immediate consideration, why not treat this as a 6th Ace? There is then no need to "ask for the Q". You can stop in 5T missing 2, bid 6T missing one, or look for 7T by checking side Ks and Qs if you have all 6 aces. It makes bidding easier. Ideally the ace asking bid should be the next step up from 4T, but without that you can still have 4NT as the ask, with "0" signing off in 5T, and otherwise in steps for 1, 2, 3. You use the same modulus 4 duality, so for example with hearts as trumps and 4NT asks :
5♣ = 1 or 5
5♦ = 2 or 6
5♥ = 0 or 4
5♠ = 3.
Because of the 4NT starting point, you would not ask unless 3 were enough for small slam. (Note, with RKCB 2+Q commits to slam.)
If the asking bid was 4♠, with hearts as trumps, then you can cope with any reply.
Much simpler than RKCB and the same basic method as used now.

[johnu]

smerriman, on 2017-December-17, 18:22, said:

Isn't it standard that if you are in any doubt as to the response, you sign off - and partner raises to slam with 3 or 4 keycards? So it seems very unlikely you'd ever run into any difficulties.

I don't know about standard, but I would rather make a grand slam try than raise directly to 6.

[johnu]

fromageGB, on 2017-December-19, 15:44, said:

You bid 4NT, it is passed to opener who bids 5♣. Your lucky day, so check for the Q (yes), bid 7♠.

* * *
That's quite reasonable, isn't it?

* * *
Partner has
♠QJTxx
♥Kx
♦xx
♣KQJx
and they take their 3 aces for 3 down.

It is that "0 or 3" ambiguity that is the problem. "0 or 4" is much safer.

You bid 5♦ to check for the trump queen and get the 5♥ response that shows the trump queen. Why would you jump to slam when you still don't know if partner has 0 or 3 aces? A 5♠ signoff has nothing to lose since opener should automatically pass with 0 keycards and automatically continue with 3 keycards.

Admittedly, this is a best case scenario where you should have no problems. Spades are trump and you have lots of room to make intermediate bids.

[Zelandakh]

johnu, on 2017-December-19, 17:13, said:

Why would you jump to slam when you still don't know if partner has 0 or 3 aces?

This was the whole point of the post. Fromage was responding to the sub-thread suggesting that there is no hand where there can be ambiguity between 0 and 3 key cards.

[johnu]

Zelandakh, on 2017-December-19, 17:47, said:

This was the whole point of the post. Fromage was responding to the sub-thread suggesting that there is no hand where there can be ambiguity between 0 and 3 key cards.

And my point was that there was no problem in the example given.

[smerriman]

fromageGB, on 2017-December-19, 16:35, said:

Much simpler than RKCB and the same basic method as used now.

Doesn't this now make it impossible to proceed when holding a 10 card trump fit when you don't care about the queen? Maybe it works for you, but seems much more complicated than RKCB for me compared to the rare occasion of double ambiguity.

[Vampyr]

OP, you have found sources that demonstrate that "KCB" is in fact known outside your local club, but that doesn't mean it is a good idea. If you really want to play it, but eventually switch to RKCB, as long as you use the responses listed in post #22 you will be able to switch without causing too much difficulty for your partner.

But is RKCB really that complicated? You don't have to play the full system. In fact, most people ask for kings (usually specific) and the trump queen and nothing else (they may have added responses to show voids).

[Zelandakh]

johnu, on 2017-December-19, 18:05, said:

And my point was that there was no problem in the example given.

If you have the agreement to ignore a sign off with 3 or 4 key cards. This is not a bad idea (I suggested it earlier in this thread) but it is not something that every pair plays.

[Vampyr]

Zelandakh, on 2017-December-19, 20:06, said:

If you have the agreement to ignore a sign off with 3 or 4 key cards. This is not a bad idea (I suggested it earlier in this thread) but it is not something that every pair plays.

Are you sure?

[Zelandakh]

Vampyr, on 2017-December-19, 22:08, said:

Are you sure?

Sure that it is not a bad idea or sure that not every pair plays it this way?

[Vampyr]

Zelandakh, on 2017-December-20, 06:33, said:

Sure that it is not a bad idea or sure that not every pair plays it this way?

The latter.

[Zelandakh]

Vampyr, on 2017-December-20, 07:27, said:

The latter.

I am reasonaby confident that below a certain level the majority do not play it that way.

[Tramticket]

Vampyr, on 2017-December-19, 22:08, said:

Are you sure? [that players systematically ignore a sign-off with three or four cards after a 4♣/♦ response]

We certainly don't have that agreement. We don't sign-off, unless two key cards are missing. We expect to be able to tell whether partner is showing zero/one key card or three/four key cards (and therefore know whether two key cards are missing). If we don't know then we have no business bidding RKCB - other slam investigation techniques are available, such as cue bids.

I suppose it might be possible to construct a freak deal (probably in a competitive auction), where a response would be ambiguous. If that ever happens (it never has), I guess that we will have to sign off and hope that partner figures it out. (She will).

[rmnka447]

Liversidge, on 2017-December-16, 15:41, said:

I have just found a source for Key Card Blackwood as I have described it on Richard Pavlicek'website - the Advanced Partnersip Bidding section - under Suit Slam Bidding.
http://www.rpbridge.net/3m00.htm

What follows is a direct cut and paste. he does not mention RKCB, queen ask etc., just the basic system I have been describing.

"The best of the many ace-asking conventions is “key-card” Blackwood. The basic concept is to count the king of trumps — always an important card — just like an ace. Thus there are five key cards, and your partnership needs at least four of them to warrant bidding a slam".

From comments made on BBO in the past I gather that he is well respected, which is odd given the comments on this thread about Key Card Blackwood. All very confusing.

FYI, Richard Pavlicek is, indeed, a very well respected American Bridge player. He is an ACBL Grand Master and an WBF World International Master. He has won 12 major events (11 Team, 1 Pairs) at North American Bridge Championships (national tournaments) and finished runner-up in such events about an equal number of times. He has over 17,500 master points which puts him in 175th place in the all-time list of North American Masterpoint winners. He has devoted significant time to teaching bridge and writing about it. He coauthored perennial bestseller "Modern Bridge Conventions" with Bill Root.

In short, he's one heck of a bridge player and teacher.

[johnu]

Zelandakh, on 2017-December-19, 20:06, said:

If you have the agreement to ignore a sign off with 3 or 4 key cards. This is not a bad idea (I suggested it earlier in this thread) but it is not something that every pair plays.

Let's see. On the sample sequence, you have shown 0 or 3 keycards, partner asked you for the trump queen, you have shown the trump queen, and now partner has signed off in 5♠. Whether you have an agreement or not, you should bid on since the only thing partner can be worried about after using RKC is whether you have 0 or 3 keycards, and you have 3 keycards.

If instead partner signed off in 5♠ after your 0 or 3 keycard response, you may not be sure you aren't missing 2 keycards since partner "may" (???) have bid RKC with 0 keycards. Can't you see that this is a different level of problem?

[Zelandakh]

johnu, on 2017-December-21, 02:53, said:

Let's see. On the sample sequence, you have shown 0 or 3 keycards, partner asked you for the trump queen, you have shown the trump queen, and now partner has signed off in 5♠.

Oh, you meant a 5♠ sign-off after 5♥? I had assumed you meant over 5♣, which is the standard way of differentiating between 0 and 3 in such circumstances. After a successful queen ask, it is not certain that 5♠ is a sign-off any more. Many play this as the king ask or even as a special sequence to fond out information that would otherwise be difficult. There have been many BBF threads over the years on continuations after a trump queen denial or sign-offs after a successful trump queen ask. I am fairly sure Justin made the case for 5♠ being forcing in this precise sequence many years back for example.

But yes, if your agreement is that 5♠ over 5♥ checks back between 0 and 3 keycards then great. We found out about the ♥K extra over the alternative method. You should not pretend that this is a universal treatment though.

[johnu]

Zelandakh, on 2017-December-21, 06:43, said:

Oh, you meant a 5♠ sign-off after 5♥? I had assumed you meant over 5♣, which is the standard way of differentiating between 0 and 3 in such circumstances. After a successful queen ask, it is not certain that 5♠ is a sign-off any more. Many play this as the king ask or even as a special sequence to fond out information that would otherwise be difficult. There have been many BBF threads over the years on continuations after a trump queen denial or sign-offs after a successful trump queen ask. I am fairly sure Justin made the case for 5♠ being forcing in this precise sequence many years back for example.

But yes, if your agreement is that 5♠ over 5♥ checks back between 0 and 3 keycards then great. We found out about the ♥K extra over the alternative method. You should not pretend that this is a universal treatment though.

And your suggested method is??? Sure, if you have agreed in advance to play this, but otherwise??? Without any agreements, doubt over 0-3 or 1-4 has to be in the back of the mind. I think it would by crazy to bid 5♠ as a grand slam try (or investigate 6NT?) when partner may play it as signoff.

[fromageGB]

smerriman, on 2017-December-19, 18:38, said:

Doesn't this now make it impossible to proceed when holding a 10 card trump fit when you don't care about the queen? Maybe it works for you, but seems much more complicated than RKCB for me compared to the rare occasion of double ambiguity.

This is much simpler than RKCB. The steps are a logical and simple 1,2,3, and there is no asking for the Q, because you have already included this. What could be simpler? Yes, if you have an undisclosed 10 card fit you can no longer ignore the Q, but this has never been useful to me when playing RKCB, as far as I recall. A partner with a good memory says he has only once done that, and the Q turned up in 3 and lost. If that ability is your most important priority then choose your poison, but I have never needed it, and simplicity with no practical ambiguity is more important for me.

The method also has a number of other advantages you do not get with RKCB:

• Assuming you use the space between 5T and 6T to discover all the side Ks and the ability to ask for or show "extras", a 1 or 2 reply also enables you to discover some agreed specifics, such as the Qs of the suits bid apart from trumps, in the space you would have otherwise used for the trump Q.
  
• It can be used without change for void asking (exclusion asking) where you ask with a bid >4T in the suit to be ignored. Say with hearts trumps and a club void, 5♣ asks; partner with the 2 aces you need replies 5♠, and you can still discover both side Kings and extras before deciding on 6 or 7.

And also bear in mind that when you do have a 10 card fit, sometimes that is known by partner so you can agree to ignore the Q if you wish, and if unknown, three quarters of the time you will have the Q anyway.

[blackshoe]

My understanding of the development of Roman Keycard Blackwood is that original Blackwood specified responses showing 0 or 4 aces, 1 ace, 2 aces, 3 aces; then the Italians came along with Roman Blackwood, with responses showing 0 or 3 aces, 1 or 4 aces, 2 aces; then somebody (Kantar?) built on that to Roman Keycard Blackwood, responses showing 0 or 3 keycards, 1 or 4 keycards, 2 or 5 keycards* (two steps, the first denying the trump Q, the second showing it), where the keycards are the four aces and the king of trump. IMLE "Keycard Blackwood" is a shortened form of "Roman Keycard Blackwood" (the name), not a separate convention. But I suppose a separate such convention does exist, since somebody found a description of it somewhere. However, I do object to the OP's characterization of this "Keycard Blackwood" convention as "standard".

In Kantar's Roman Keycard Blackwood: The Final Word he takes some 85 pages to describe the basic convention, and an additional 230 or so pages (from memory, the numbers might be off a little) to describe exceptions, add-ons, and other nonsense. In all of that he never mentions Kickback, which IMO is a superior approach. I have a book on it around here somewhere. It's about a hundred pages, maybe less. Kickback has its own problems and idiosyncrasies, of course.

All in all, I'd suggest starting off with no Blackwood. Learn Italian style cue-bidding instead.

* Most people forget about the 5 keycards. After all, would anyone ask holding no keycards himself? But it is part of the original system.