[Valardent]

Hi,

I'm new to this forum, so sorry if post is unclear.

The system I’ve been playing for 5 years is basicly natural with a few starts for relay bidding (describing shape + specific honnor ask)

I’m trying to improve the auction termination mechanism once shape is known. Until now, I used what R. Hughes recommended in his book “building a bridge system”. The scheme works like this : 1st ask for aces (3 steps) : 14, 03, 2. 2nd ask for main kings (4steps) : 03, 1st or 2 others, 2nd or ..., 3rd or... (zooming on 1st ask if 2 aces so 6 steps possible on 1st ask)
3rd ask for last king and the two main queens and so on (groups of 3)...

Downsides of this method are that you must wait 3rd ask to know of queen of trumps and if relayer happens to have none of the honnors of a group of three (too frequent I’ve found), you will be stuck unless you can deduce it through the HCP total (often difficult if kings ask, if not impossible if queens and/or jacks ask)

I read all the topics about spiral scans/DCB of this forum for the last 7/8 years, learned a lot, but I’m still unsure which method would fit the better in my system structure because I’ve no experience at all using these tools.

In my system, there are 2(3) cases where a GF relay is initiated, asking for shape and strength followed by a specific ask for location honnors.

I would very much appreciate any help for any of the following cases from you all folks.

So this is how my system works :

1m openings are either natural unbalanced 4+m 11-22 HCP or any balanced 12-14 (1club), 18-20 (1diamond)

1st case : unbalanced minor hands

1m 1NT* *GF relay 12+ HCP without a 5 card major (so almost always balanced unless long minor or 5-4m with 15+HCP)

Shape and strength are described below or at 3NT in 95% so 4C (rarely 3S and almost never 3H) is often 1st step for scan. Shape is precisely known and strength is either 11-13, 14-16 or 17+ HCP.

A few issues I’ve been thinking about :

- the king in minor’s opener should be considered (almost) worth an ace since relayer has always at least 2 cards in opener’s minor

- I wonder if a DCB start (on the basis of AKQ SP) at 4C would be possible considering the fact that number of SP controls is not precisely known (I think a SP control ask followed by a dcb would take too much space and would so only be useful when one considers a grand slam which is not the first target). Would that be suicidal even with hands that are looking for a grand slam?

- Relayer should be able to ask for keycards and still be able to sign off in 4(5)NT , in 5m or in 5M in case of a 44 major fit.

- Should the method be different in function of opener’s strength?

- All in all, there will be maximum 3 kc asks possible since (at least) a kc in one of the 3 other suits (often a major) doesn’t make sense since opener has a single or a 2 card M and R hasn’t got any 5 card major.


2nd case : balanced hands

1 m 1NT*
2 NT 3 D** ** control ask A=2, K=1

In fact, I designed two ways to show a balanced hand : the one showed for “common” hands and an other that shows a max hand without jacks or maximum one in a four card suit if 14HCP (20HCP if 1D)

The idea is that if 3 or more controls are missing, slam will be at best on a finesse, so responder signs off in 3NT with 9 or less controls.

If the total is 10 (11 or 12), one can ask for shape (4C). When shape is known (around 5m), R should be able to ask for queens/jacks location.

(Problem here is that with a king missing, sometimes one can’t know which one it is. True, a good grand slam will be rare with only 11 controls and 2 balanced hands but some are even laydown)

For the QJ ask, would a DCB scheme work? Like : 1step is none or QJ, 2nd step is Q in first suit, none or QJ in 2nd or is a spiral more suited? How could the fact of using two ways to describe balanced hands improve on the precision of the chosen method?

If 3D asks for SP, will it be easier to locate honnors (at first glance I think it would be even more difficult)?


3rd case : 5+M opening hands

1M is 11-22, 5+M

1M 2♣* *GF relay 12+ HCP, no M fit (unless 18+ balanced) and no 5+ cards OM.

Again, shape and strength are described mainly below or at 3NT with 3 zones : 11-13, 14-16 and 17+

The main difference with the 1st case is that R can be single in opener’s M.

I was thinking of an ace asking bid followed by a DCB scheme where the 1st scan of the M would be none or KQJ (maybe only if 6+M?), other suits being none or KQ. Does this make sense?

For the ace asking bid, would you always recommend 14/30/2 (30/14/2?) or does some cases allow 2 steps (even or oneven number of aces) depending on opener’s strength?

Tx for reading till end ,

Patrick

[dake50]

In situations of captain-ask verses 'I got-you got?', relay for lowest missing ask.

BUT 2-step relay(if not agreed signoff) asks next rank. Eg. I got A answer I need, now skip K's, go to Q's.

[Free]

Hi Patrick,

I think it's better to provide some specific hands (and switch some honours to get different HCP ranges) and their bidding sequences up to the point you want to initiate slam bidding. This way we can see if this is really a slam bidding issue, or a problem with your relays.

For example, your 1M openings have 3 point ranges, I'm pretty sure you don't have exact distribution below 3NT in a lot of situations. Otoh you may have acceptable knowledge about the hand (for example, 6+M and a shortness, can be considered 6331 in most situations). I play relays over 1M openings with 2 ranges and even then we can't show exact shape in various situations.

My experience with AKQ-ask and denial cuebids is very positive, after 1 scan you usually have a very good idea of partner's hand because you take every top honour into account. It was however in a system which was designed for relays (MOSCITO), so the relay structure was VERY efficient. After initial slam bidding we could signoff at game level most of the time. We also had the possibility to ask for keycards, but that was seldom done because denial cuebids give so much more information. RKC was only used if you wanted to know keycards and specific tophonours in 1 suit (maybe even 2, but usually only 1 suit was the key factor).

[Ant590]

I haven't read R. Hughes's book, and this might be telling everyone what they already know, but here goes!

You can save a lot of space with Question & Answer relay-type auctions if you employ run-ons. In other words, if you structure your responses to the questions such that the most space consuming answers show better hands, then these hands can automatically answer the next question.

An example of this is actually RKCB, where the next question "do you have the Q of trumps" is answered automatically if partner has two keycards, the most space consuming response.

Spiral scanning and symmetric-relay type bidding follow this principle. Of course if your relay auction is not a linear series of questions, but allows partner to ask a variety of questions then you can't be too crazy with the run-ons.

Like I said, sorry if this was teaching Grandma to suck eggs.

[Valardent]

I'll post a serie of examples with more specific questions. I think it will make things easier and more clear in the long run.

First of all, let’s see if I understood the dcb scheme.

---

1C 1H* (spade transfer)
1S (3/4s) 2D** (GF relay)
2H (3s) 2S ®
3C (5c) 3D ®
3H (balanced 12-14) 4C (asking for SP controls, let’s assume base 6+)
4H (7 SP) 4S ®
5H (AKQ or 0 in diamonds) 5S®
6C (2 in clubs, only one in spades) 6D®
6H (only one in hearts) 6S®
7C (J of d, J of c, no spadeJ) 7NT (hoping for 3-2 spades or K♦ on)

Shape of opener is 5c3s(3)(2) (3S would hv asked full shape but no need here) ; in such a case I consider hearts and diamonds having the same length and so hearts gets scanned first (♥♠♣♦)

2 questions :

1) what should be the SP bases for 11-13, 14-16, 17+?

2) 6C finished showing all opener 7 SP, why can’t 6D directly asks for jacks? Is that because opener doesn’t know for sure that R knows all opener 7SP?

Patrick

[Free]

Valardent, on Nov 2 2009, 12:24 PM, said:

I'll post a serie of examples with more specific questions. I think it will make things easier and more clear in the long run.

First of all, let’s see if I understood the dcb scheme.

Dealer: ????? Vul: ???? Scoring: Unknown

♠ Qxx ♥ Kx ♦ J9x ♣ AQJ10x ♠ AKxxx ♥ Ax ♦ AQ10xx ♣ K

1C 1H*  (spade transfer)
1S (3/4s) 2D** (GF relay)
2H (3s) 2S ®
3C (5c) 3D ®
3H (balanced 12-14) 4C (asking for SP controls, let’s assume base 6+)
4H (7 SP) 4S ®
5H (AKQ or 0 in diamonds) 5S®
6C (2 in clubs, only one in spades) 6D®
6H (only one in hearts) 6S®
7C (J of d, J of c, no spadeJ)  7NT (hoping for 3-2 spades or K♦ on)

Shape of opener is 5c3s(3)(2) (3S would hv asked full shape but no need here) ; in such a case I consider hearts and diamonds having the same length and so hearts gets scanned first (♥♠♣♦)

2 questions :

1) what should be the SP bases for 11-13, 14-16, 17+?

2) 6C finished showing all opener 7 SP, why can’t 6D directly asks for jacks? Is that because opener doesn’t know for sure that R knows all opener 7SP?

Patrick

1) It's hard to say which base you should use. For MOSCITO we used 6+SP for a HCP range of 10-14, 9+SP for a 15+HCP strong C, and 10+SP for a 17+HCP strong C. Seems a bit like 3*min_HCP/5. The best is probably to run a big simulation to see which handtypes fit (distribution and all included) and accept a certain amount of hands that need to pass although they have enough HCP. We had hands with 8-9HCP which were opened, and we also had hands with 12HCP that needed to pass.

2) Relay responder never knows when relayer has enough information. The only thing relayer can say is "tell me more", but no more details.

[Valardent]

Back to 1st case of first post :

Partner opens 1C and you have :

Axx
Kxx
K
QJ10xxx

1♣ 1NT* (balanced GF relay without a 5 card major, long minor possible)
….
3NT (1435 11-13) ??

What now?

If you decide to use 4C as SP ask, what should 4D, 4H, 4S and 4NT be for?

(5C or higher would be sign off?)

If 4D is terminator and 4H,S,NT keycard asks in the 3 possible suits (spades excluded since R has at most 4 spades)

I imagine that best would be 4H kc for clubs, 4S and 4NT kc for the reds.

But should 4S be kc for ♥ since hearts is second longest suit or should 4S be kc for ♦ since the sign off in 5D is cheaper than the sign off in 5H?


I thought (wrongly I think) that using 4C as a SP ask followed by DCB would almost always commit you to play slam. In fact, if R is borderline for slam, asks for SP ctrl and gets a minimum response, all bids except 1st are to play (right?). So the risk of getting overboard is small. Although signing off in 4H won’t be possible if RR has a minimum SP base, in 4H and 4S if mini+1 base and so on.

Back to the example :

What is best ? 4H kc in clubs or 4C SP ask?

1) If 4C :

We get 4S (8 SP) as an answer, what now?

Slam could be cold with x Axxx Axx Kxxxx or dreadfull with x Axxx (J)xx AKxxx (same hand with JH or AQxx Qxx Axxxx and odds are +/- 50%)

2) If 4S :

We get 5C (0 or 3 kc), see 1)

In both cases, 5C is still a possible contract…
Am I missing any important matters?

P's hand : x Axxx Axx Kxxxx

About the SP bases for 11-13,14-16 and 17+…

Let’s start with a 6+ base for 11-13. Upper limit should be 9 because

AAKK or AAAQ is 10 SP and 14 HCP

For 14-16, I get through Fred’s formula : 42/5 = 8.4

So 2 choices :

1) base is 8+ and hands like KQJx QJ(x) Q(J) KJxxx (16 HCP) have to be downgraded to 11-13 (balanced or semi-balanced 15-17 hands can open 1NT)

2) base is 9+

x RDVx ADx DV10xx (15HCP) 8 SP should be downgraded. That’s way too much of a distortion so 8+ base is fine

Upper limit is 12 ? : AAAA

To keep some symmetry, a hand with just the 4 aces should be upgraded to 17+ so that 11-13 is 6 to 9 SP and 14-16 8 to 11

For 17+, 51/5 = 10.2

10+base seems logic.

Any comments are welcome.

Patrick

[Valardent]

Back to 3rd case of first post :

Partner opens 1H and you have :

AKQ
AQxx
Kxx
Axx

1♥ 2♣* (GF relay; no 5+ ♠ and no major fit unless 18+ balanced)
…. R
…. R
3NT (2542 11-13)

What now?

I can see 2 plausible ways, 4C SP ask or ♥ kc followed by a suit asking bid (SAB) in ♦

If 4D is terminator, 4H should be kc in ♥ (but what about 4S and 4NT? Kc in ♣ and ♦ or ♦ and ♣? I think I will stick with the suit order logic ♥♠♣♦ to still be able to sign off in 5m in both cases on borderline hands if opener has less than 2 keys)

….. 4♥
5♣(2 without) 5♠ (5D wld ask for HQ, 5H is sign off, so 5S is SAB in D)
6♣(K or AQ) 7♥

P’s hand : xx KJxxx AQ10x Jx (not a mandatory opening)

Imagine we change that hand into xx KJxxx A10xx KQ (7NT cold) :

….. 4♥
5♣(2 without) 5♠
6♦ (A or KQ) ?

If I get it right, 6♠ would be a SAB in ♠, 6NT and higher sign offs. So no ♣ SAB available (response would probably be too high anyway)

Let’s see to where a 4♣ SP ask + DCB would have led :

…. 4♣
4♦ (6SP) 4♥
5♣ (no SH) 5♦
5♥ (no CH) 5♠
5NT (one HH) 6♣ (to be sure )
6♥ (2 DH, no SJ) 7♥

With second hand :
… 5♦
5♠ (one HH) 5NT
6♣ (one DH) 6♦ (12 tricks in NT in any case*)
6♥ (no SJ) 6♠
7♦ (QC+HJ, no DJ) 7NT

*except that opener could have Jx KJxxx AJxx QJ, pbbly marginal enough to ask further

So in this case SP + DCB was a better choice than kc + SAB it seems.

Patrick

[Free]

Many questions, I'll try to answer them all

When 3NT shows the exact distribution, you have the following continuations:
4C = SP ask
4D = terminator
4H = RKC suit 1
4S = RKC suit 2
4NT = RKC suit 3
5C = RKC suit 4

Suits are ordered by length, in case they have equal length you use priority H>S>C>D. So here 1=C, 2=H, 3=D and 4=S. It's indeed pretty useless to ask for keycards in S, but these are the standard rules (and you see it's the highest step anyway because it's so rare). However if partner showed a 2-4-2-5 at 3NT you might want to make exceptions: 3=S and 4=D although RKC D would have more merrit. Obviously one can come up with some general rules of exception.

When starting slam auctions, it's important to know what information you need. If you need general top honour power, SP ask is most useful. If you need specific cards, RKC is better.
There are around 24SP. For balanced vs balanced hands you need around 21SP, this means missing either an Ace or both a King and a Queen (or 3 Queens). With unbalanced hands you can go a lot lower (suppose you have the Ace in partner's singleton, then 18 combined SP will be enough - K and Q of partner's singleton won't be counted). If you're way below that limit, it's usually better to use RKC, as in this case: you have 8 (you can count your singleton K to evaluate the situation) and partner has 6-9. Even if he has maximum, he can still have a useless hand. You'll never know in time if he has the quick tricks or not because 17 (8+9) is still below the required 18.

In this case, RKC is definitely better, because you need to know if partner has the missing keycards, not if he has some Queens and Kings on the side, and you need to know this fast! Anyway, lets compare both cases:
1) SP ask: you get 4S which shows 8SP. You don't know anything, so you have to signoff in 5C. Partner always has to pass.
2) RKC: 4H = RKC C, then 4NT = 0-3. If you signoff now, partner is allowed to continue the auction since you don't expect him to have 3 with 11-13HCP. You'll get to 6C, but also if partner has x-Axxx-xxx-AKxxx...

About the SP base, my suggestion:
11-13 = 6+SP
14-16 = 8+SP
17+ = 10+SP
It's symmetric and it uses my original ranges (10+-14=6+ and 17+=10+). Upper limits are not important imo.

It's indeed possible that you have to downgrade some hands. Note that these are usually hands with lots of Jacks, or with singleton K or Q, which deserve to be downgraded anyway. And opps have to pass, and partner needs a GF to begin with,... which makes this situation very unlikely.

A note about your relays:
Personally I think 3NT is way too high for frequent hands like 5431. If you'd show this hand with 3D (like moscito), then you can RKC with 3S! =>4C=0/3, followed by SAB D which is 4H. Now opener can show his top honours, if he has Qxx or xxx you'll be below 5C, otherwise he has the Ace, which means 3 keycards and slam.
If you'd show 5431 with 3H you lose 2 steps (because 3NT is not an asking bid). Now you're too high, because 4NT would be the SAB D and with Qxx you'll pass 5C.
Why do you use so many HCP ranges anyway? If you want to use SP ask with DCB's, a range of 4-5HCP is good enough (for example 11-15 and 16+). It might even be useful to not show HCP ranges anyway, and fall back on zooming to show significant extra's. You can signoff, and partner with 3SP extra's than the minimum base can zoom. You'll base your slambidding not on HCP but on SP anyway. I think this would improve your relays significantly, and zooming allows your slambidding to be more efficient (as I showed you in an email).

If you'd use wider ranges, I'd suggest:
11-15 = 6+SP
16+ = 9+SP

Another way of improving relays is to start with a lower call. Suppose you use 1C-1D as natural or GF relay. Now opener will show his hand type (example 1M = unbalanced natural, 1NT = balanced, may have 4M (Walsh-style)) and now you just need 1 more call to show the GF relay (2D for example when playing 2-way checkback). Opener already showed a lot of his hand. Suppose he bid 1H showing unbalanced and 4H. This means 4441 or 5+C-4H. After the 2D GF relay, you could bid:
2H = short S
2S = 5422/4441 (now relay for shortness => exact shape on 3C(5422)/3D/3H)
2NT+ = short D, exact distribution
This way your 5431's are shown on 2NT, 6421's at 3C, 6430 at 3D,... Extreme efficient imo! If you want to cater for 5H-6C you'll need some extra step somewhere, but you'll still have your 5431's at 3D or lower.
Ofcourse, if opener happens to be extremely strong (18+HCP reverse for example) you'll lose accuracy. You might want to make 1M rebid forcing for example, and make reverses 6-4.
After 1C-1D-2C, you can use 2D or 2H as GF relay (depends if you want 2D natural or not).
After 1C-1D-2D, you can use 2H as GF relay.

You can do the same after 1D openings: 1D-1H = natural or GF relay. I've actually played this after a nebulous 1D (showing 0+D) without much problems. Here you might want to play 4-card D however, to make further bidding after 1D-1H-2H more precise. Otherwise opener only showed 3D and 4H, after 2S relay you still have a lot of possible distributions (balanced, 4441, 5+D-4H). This is the worst case scenario for your relays, any other response is more efficient:
1D-1H-1S-2D GF relay (like XYZ) is acceptable (2 steps lower than with H-D)
1D-1H-1NT-2D GF relay
1D-1H-2C-2S GF relay (I suppose you want 2H to be natural). This is pretty high, because you need to show 5-5 hands as well. You might want to use 2D as relay here, usually this isn't a problem. (I already play 1D-1M-2C-2D as relay (not GF), opener can have longer C, this works ok but we haven't set it up to be a full relay scheme)
1D-1H-2D-2S GF relay (I suppose you want 2H to be natural)

[Valardent]

Free, on Nov 3 2009, 05:27 AM, said:

Many questions, I'll try to answer them all

When 3NT shows the exact distribution, you have the following continuations:
4C = SP ask
4D = terminator
4H = RKC suit 1
4S = RKC suit 2
4NT = RKC suit 3
5C = RKC suit 4

Suits are ordered by length, in case they have equal length you use priority H>S>C>D. So here 1=C, 2=H, 3=D and 4=S. It's indeed pretty useless to ask for keycards in S, but these are the standard rules (and you see it's the highest step anyway because it's so rare). However if partner showed a 2-4-2-5 at 3NT you might want to make exceptions: 3=S and 4=D although RKC D would have more merrit. Obviously one can come up with some general rules of exception.

When starting slam auctions, it's important to know what information you need. If you need general top honour power, SP ask is most useful. If you need specific cards, RKC is better.
There are around 24SP. For balanced vs balanced hands you need around 21SP, this means missing either an Ace or both a King and a Queen (or 3 Queens). With unbalanced hands you can go a lot lower (suppose you have the Ace in partner's singleton, then 18 combined SP will be enough - K and Q of partner's singleton won't be counted). If you're way below that limit, it's usually better to use RKC, as in this case: you have 8 (you can count your singleton K to evaluate the situation) and partner has 6-9. Even if he has maximum, he can still have a useless hand. You'll never know in time if he has the quick tricks or not because 17 (8+9) is still below the required 18.

In this case, RKC is definitely better, because you need to know if partner has the missing keycards, not if he has some Queens and Kings on the side, and you need to know this fast! Anyway, lets compare both cases:
1) SP ask: you get 4S which shows 8SP. You don't know anything, so you have to signoff in 5C. Partner always has to pass.
2) RKC: 4H = RKC C, then 4NT = 0-3. If you signoff now, partner is allowed to continue the auction since you don't expect him to have 3 with 11-13HCP. You'll get to 6C, but also if partner has x-Axxx-xxx-AKxxx...

About the SP base, my suggestion:
11-13 = 6+SP
14-16 = 8+SP
17+ = 10+SP
It's symmetric and it uses my original ranges (10+-14=6+ and 17+=10+). Upper limits are not important imo.

It's indeed possible that you have to downgrade some hands. Note that these are usually hands with lots of Jacks, or with singleton K or Q, which deserve to be downgraded anyway. And opps have to pass, and partner needs a GF to begin with,... which makes this situation very unlikely.

A note about your relays:
Personally I think 3NT is way too high for frequent hands like 5431. If you'd show this hand with 3D (like moscito), then you can RKC with 3S! =>4C=0/3, followed by SAB D which is 4H. Now opener can show his top honours, if he has Qxx or xxx you'll be below 5C, otherwise he has the Ace, which means 3 keycards and slam.
If you'd show 5431 with 3H you lose 2 steps (because 3NT is not an asking bid). Now you're too high, because 4NT would be the SAB D and with Qxx you'll pass 5C.
Why do you use so many HCP ranges anyway? If you want to use SP ask with DCB's, a range of 4-5HCP is good enough (for example 11-15 and 16+). It might even be useful to not show HCP ranges anyway, and fall back on zooming to show significant extra's. You can signoff, and partner with 3SP extra's than the minimum base can zoom. You'll base your slambidding not on HCP but on SP anyway. I think this would improve your relays significantly, and zooming allows your slambidding to be more efficient (as I showed you in an email).

If you'd use wider ranges, I'd suggest:
11-15 = 6+SP
16+ = 9+SP

Another way of improving relays is to start with a lower call. Suppose you use 1C-1D as natural or GF relay. Now opener will show his hand type (example 1M = unbalanced natural, 1NT = balanced, may have 4M (Walsh-style)) and now you just need 1 more call to show the GF relay (2D for example when playing 2-way checkback). Opener already showed a lot of his hand. Suppose he bid 1H showing unbalanced and 4H. This means 4441 or 5+C-4H. After the 2D GF relay, you could bid:
2H = short S
2S = 5422/4441 (now relay for shortness => exact shape on 3C(5422)/3D/3H)
2NT+ = short D, exact distribution
This way your 5431's are shown on 2NT, 6421's at 3C, 6430 at 3D,... Extreme efficient imo! If you want to cater for 5H-6C you'll need some extra step somewhere, but you'll still have your 5431's at 3D or lower.
Ofcourse, if opener happens to be extremely strong (18+HCP reverse for example) you'll lose accuracy. You might want to make 1M rebid forcing for example, and make reverses 6-4.
After 1C-1D-2C, you can use 2D or 2H as GF relay (depends if you want 2D natural or not).
After 1C-1D-2D, you can use 2H as GF relay.

You can do the same after 1D openings: 1D-1H = natural or GF relay. I've actually played this after a nebulous 1D (showing 0+D) without much problems. Here you might want to play 4-card D however, to make further bidding after 1D-1H-2H more precise. Otherwise opener only showed 3D and 4H, after 2S relay you still have a lot of possible distributions (balanced, 4441, 5+D-4H). This is the worst case scenario for your relays, any other response is more efficient:
1D-1H-1S-2D GF relay (like XYZ) is acceptable (2 steps lower than with H-D)
1D-1H-1NT-2D GF relay
1D-1H-2C-2S GF relay (I suppose you want 2H to be natural). This is pretty high, because you need to show 5-5 hands as well. You might want to use 2D as relay here, usually this isn't a problem. (I already play 1D-1M-2C-2D as relay (not GF), opener can have longer C, this works ok but we haven't set it up to be a full relay scheme)
1D-1H-2D-2S GF relay (I suppose you want 2H to be natural)

Hi Fred,

Tx for your replies, it helps a lot.

Some feedbacks to your last post :

Quote

However if partner showed a 2-4-2-5 at 3NT you might want to make exceptions: 3=S and 4=D although RKC D would have more merrit.

3=S doesn't make sense since R has at most 4 spades so 3 should definitely be kc in ♦

Quote

In this case, RKC is definitely better, because you need to know if partner has the missing keycards, not if he has some Queens and Kings on the side, and you need to know this fast!  Anyway, lets compare both cases:
1) SP ask: you get 4S which shows 8SP.  You don't know anything, so you have to signoff in 5C.  Partner always has to pass.
2) RKC: 4H = RKC C, then 4NT = 0-3.  If you signoff now, partner is allowed to continue the auction since you don't expect him to have 3 with 11-13HCP.  You'll get to 6C, but also if partner has x-Axxx-xxx-AKxxx...

I agree, RKC is better (safer) in this case. But getting an 8 SP response allows you to deduce the different possibilities for slam partly tx to the 11-13 range.

When p bids 4NT, you know he shows 3 tx to his shape.

So in this case, one way doesn't get the edge on the other.

Quote

A note about your relays:...

Indeed, I could design my system to have only 2 ranges and start the auction termination lower but :

-It would make it more artificial (with many 2-ways bids). Altough I use a lot of transfer responses to gain some space, I much more prefer the base remains natural (one reason is in case of preempt)

- I think having more narrowed ranges can be an advantage

- I'm not going to change all the relay structure which is mostly natural and pretty easy to remember (don't have the time nor the will)

- Those auction termination are pretty rare because much of the time R has to be very strong. With 2 hands that don't have a priori the potential for slam, I prefer to engage a natural sequence and eventually, if the hands meshes really well, bid slam in a more classical way. So I use this auction termination mainly when small slam is odds on or better and certainly to find some pin-point grand slams which are much more difficult to find on natural bases.

Quote

Personally I think 3NT is way too high for frequent hands like 5431.  If you'd show this hand with 3D (like moscito), then you can RKC with 3S!  =>4C=0/3, followed by SAB D which is 4H.

Is not 4D asking for Queen of trumps, 4H SAB in ♥ and 4♠ SAB in ♦?

Patrick

[Valardent]

Another example (1st case) gave me an idea :

---

bidding would go

1C 1NT
...
3S (11-13 4135) 4C (SP ask 6+base)
4H (7SP) ??

Much of the time when R is (very) heavy and RR is minimum and unbalanced, R can often deduce all AK&Q in opener's hand after SP response like in this example (if we don't count SP for K or stiff Q).

So DCB loses some intrest because we can bid 6C directly. But 7D is odds on with DJ in opener's hand. If we continue with DCB, hoping to get the information about the DJ :

... 4S
5S (only 1 in clubs*) 5NT
6C (only 1 in spades*) 6D
6H (only 1 in diamonds*) 6S *information already known
7C (CJ, no SJ)

we go way overboard without even getting the desired information

On 4H, everything above 4S is signoff, so no way to bypass the DCB scheme to ask for jacks unless... a sign off bid can't possibly be for sign off like 5H since opener is single and R has at most 4.

So if 5H becomes some sort of Jacks DCB, we get :

4H (7SP) 5H (jackask)
5NT (CJ, no SJ) 6C (6D will be end contract if no DJ)
6H (DJ) 7D

Too far-fetched?

Patrick

[Free]

Valardent, on Nov 3 2009, 01:01 PM, said:

~snip~

Quote

Personally I think 3NT is way too high for frequent hands like 5431.  If you'd show this hand with 3D (like moscito), then you can RKC with 3S!  =>4C=0/3, followed by SAB D which is 4H.

Is not 4D asking for Queen of trumps, 4H SAB in ♥ and 4♠ SAB in ♦?

Patrick

Yes you are right, I forgot a step

I've also thought about impossible signoffs, but they are so rare that I didn't go into detail. Your suggestion is definitely playable, has some merrit, but I don't know if it will ever come up. It's true that sometimes just the number of SP is enough for R to place all the missing top honours, so there's nothing wrong with giving a meaning to impossible calls. You'll usually start at 5H or above I think (5m over 4m is still signoff, but a jump in H is not) so you have to be pretty sure what you're doing, and space is limited. Immediately asking to scan Jacks is probably the most useful meaning.