[DinDIP]

One of the things almost all relay systems do poorly is ascertaining the quality of teller’s suits. Most systems have no mechanisms for showing how good (or not) a suit is until denial cue-bidding or other honour-showing methods are used. When something can be revealed about suit quality it is often limited to solid suits, often only with seven or more cards. (One exception is the Ultra system used by the late John Lowenthal and Philip Martin. This provided teller with a way to show a 6+card suit with 3+/top4 honours.)

Sometimes, however, asker needs to know at a lower level whether or not to look for a slam. Consider the problem asker faces when he has two small cards in a suit in which teller has shown six cards (or a small singleton opposite a seven-card suit). Unless teller has a strong suit that makes slam possible, asker wants to play in 4M (if a major-suit fit) or perhaps 3N (if a minor-suit fit). The latter is impossible in most systems unless asker can deny a good suit by 3N.

The original Kiwi Club tried to address the slam/suit game issue by making the first two scans by teller show the number of the top three honours in his 6+card suit. But often that is insufficient. Who wants to be in a slam depending on playing AQxxxx or KQxxxx opposite xx for one loser?

My partner and I use a suit quality inquiry that asks teller to accept only with a good enough suit to play for at most one loser opposite two small – but this operates only after shape is revealed.

However, some tweaking to the order in which shapes are shown can address this problem. Standard symmetric shows shapes like this when teller has a two-suiter:
3♦ = 5431
3♥ = 6421
3♠ = 6430
etc

What if we swap the 5431 and 6421 + 6430 responses? Now the responses are:
3♦ = 6421 or 6430
3♥ = 5431 with min+1 (or more) QP (or controls, if that’s your preference)
3♠ = 5431, min QP (or controls)
3N+ = as before

Now, after a 3♦ response, asker can bid 3♥ to inquire for shape. (While the responses could be 3♠ = 6430 and 3N+ = 6421, I think it’s better to play something like 3♠ = 6421 with min+2 or better, 3N = 6421, min or min+1, and 4♣+ = 6430.)

More importantly, asker can bid 3♠ instead to ask for shape to be resolved only if teller’s suit will play for at most one loser opposite two small cards. Yes, we are up a step compared with standard shape-showing – but asker has the valuable information he needed (and, unless you use an unusual DCB method, the step will usually be regained because teller has already shown a minimum honour strength in his 6+card suit).

The same principle can be applied when teller has a single-suiter. After a 3♣ bid to show 6223/6232 (or 6322/7222), asker can bid 3♥ (again, step+1) to ask that shape be resolved only with a suit that will play for at most one loser. And in the normal stream, where 3♦ shows a 5332, 3♥ a 6331 and 3♠ a 7321, a similar swap to that with a two-suiter can be made.

My testing shows this to be a winner. However, each relay partnership will have to look at the impact on the more frequent hands switched from 3♦. If your alternatives are RKC asks, for example – and you use these frequently – then the loss of a step may outweigh the additional information obtained about teller’s strength.

Partnerships using unlimited positive responses to 1♣ will also have to consider how teller shows hands that don’t have a strong enough suit but do have so many extras that teller cannot just bid a NF 3N over a 3♠ inquiry. My suggestion is to use 4♣ to show such hands, with good suits starting to show QP (or controls) from 4♦.

David

[hrothgar]

This is REALLY intriguing

Need to spend some time thinking about this

[Zelandakh]

My system has 2 ways of doing this. For fragment (2-3 card) suits there is usually an explicit stopper ask amongst the relay breaks. For real (4+ card) suits the relay break typically sets that suit and asks for an opinion about slam suitability. While this is not specifically referencing suit quality, that is obviously a factor and it would be easy to make it then primary factor if desired (although I currently think the general slam try is more useful). Similarly, if the bidding is at 3♦ and a 6430 (or whatever) has been shown, I like to use the only relay break below 3NT as the stopper ask for the fragment suit (as above) so you ought to factor in the loss of such alternative meanings before declaring this a winner. My testing suggests that such a stopper ask comes up much more frequently. And there are other options too - Adam uses various relay breaks for investigating shortages in quite a clever manner. Again, you need to see how these compare within the context of your system.

[PrecisionL]

Quote

(One exception is the Ultra system used by the late John Lowenthal and Philip Martin. This provided teller with a way to show a 6+card suit with 3+/top4 honours

What is your source? Thanks for clearing this mystery up as bridge guys refer to Ultra without any details.

Ultra Club Relay is a system authored by yours truly and KeyLime that is on www.bridgewithdan.com (link below in my profile notes)

Our upgrade, C3, not published yet, handles semi-solid and solid suits starting with a 2NT response to 1♣ (Artificial and forcing, 16+ hcp)

[DinDIP]

PrecisionL, on 2015-August-20, 08:54, said:

Quote

(One exception is the Ultra system used by the late John Lowenthal and Philip Martin. This provided teller with a way to show a 6+card suit with 3+/top4 honours.)

What is your source? Thanks for clearing this mystery up as bridge guys refer to Ultra without any details.

Ultra Club Relay is a system authored by yours truly and KeyLime that is on www.bridgewithdan.com

Yes, I've noticed the similarity in name . . .

Not sure by whom (I think John Lowenthal as I corresponded with him about Borel, his hand generation program) or exactly when (haven't kept emails from the 90s), I was given a copy of (some of) the notes for the Lowenthal-Martin system (and asked not to distribute further). They called it Ultra Club as it was an improvement, in their eyes (and mine), on the Ultimate Club. It shares some features with the Ultimate Club, especially the strong club with strength-showing responses (but in QP rather than AKQP). However, it reflects Lowenthal's preferences from his Canary Club partnership with Paul Heitner: canape openings when maximum (and MAFIA otherwise), Reverse Flannery and Roman 2M.

In this sophisticated system teller used different paths in response to 1♣ to show single-suiters with solid, good and bad suits, and asker had the option of breaking the chain (below 3♥) to ask for QP in the last-shown suit before resuming shape-showing one level higher. After a 1♦/1M opening and 2♣ relay teller could show a solid suit, and asker could use Ogust or relay for shape if teller showed a non-solid single-suiter.

David

P.S. Very interesting ideas for DCB as well, with a control parity step and showing/denying controls, as well as only scanning n-1 non-singleton suits.

[PrecisionL]

David, thanks for the reply. Canary Club is one of my favorite systems, I have a copy of the original publication in the ACBL bulletin. There is a version on the Internet. http://www.cjkinn.ne...asny/System.htm
I have never seen a reference to Ultra but I do have paper and electronic versions of Ultimate Club.

Edited 8/23/15:
C3 (Copious Canape Club) shows semi-solid and solid suits at the 3-level in response to an opening bid of 1♣ (Strong & Artificial). 2NT and higher = 1-under Transfers.

Example: 1♣ - 2NT (♣s) - 3♣ (Beta) -

3♦ = Semi-solid ♣Suit and No outside Controls
3♥ = Semi-solid ♣Suit and K outside Control
3♠ = Semi-solid ♣s and A outside Control
3NT = Solid Suit♣ and No outside controls (AKQxxx)
4♣ = Solid Suit♣ and K outside Control
4♦ = Solid Suit♣ and A outside Control ...

[PrecisionL]

Update on suit quality methods:

I have been synthesizing a new strong club system combining elements of Fantunes & Nightmare & Millennium Club. Just recently we decided to use 1♣ - 2♣ as the catch all G.F. response which can include 5-cd suits without 2 of the top 3 honors. We also have solid 7-cd suits with and without extras in the design.

With 2 honors or more we transfer at the 2-level for majors and at the 3-level for 6-cd minors (with A or better). We treat xxxxx as a 4-cd suit and xxxxxx as a 5-cd suit.

[straube]

DinDIP, on 2015-August-17, 19:56, said:

What if we swap the 5431 and 6421 + 6430 responses? Now the responses are:
3♦ = 6421 or 6430
3♥ = 5431 with min+1 (or more) QP (or controls, if that’s your preference)
3♠ = 5431, min QP (or controls)
3N+ = as before

Thought I'd throw this out there...


3♦ = 5431
3♥ = 6421 or 6430, bad suit
3♠ = 6421 or 6430, good suit
3N+ = as before

With this opener doesn't have to skip a step to make the next ask. He makes a single step and responder first resolves shape (S1 is 6430 and etc is 6421). OTOH, if opener has a big fit, he won't care as much about suit quality so sometimes this will cost room.

Not sure how your balanced shapes with 6 resolve. Mine goes like....

2N-higher short or lower fragment (6223 or 6232)
3C-middle short
3D-7222 or 6322

which I could revise to

2N-higher short or bad suit (6223, 6232, 7222, 6322)
3C-middle short
3D-good suit (6223, 6232, 7222, 6322)

or vice versa

[shevek]

David,

This is a bit tangential ....
Our focus is on finding the right game with a 7-card major fit, or 8-card minor fit.
I recall an Acolish idea I heard from you:
1♥ - 2♦
2♥ - 3NT, shows 2 ♥s, to cater for opener having a 5-carder.
With a stiff heart, responder bids another suit then 3NT to play regardless.
This looks okay, though opener can't raise responder's second suit to the 4-level.

We have similar chain breaks in relay auctions, when asker thinks that finding shape may not be enough to decide strain.
Say responder shows a positive with a 6-card major.
A break to 3NT is a choice of games, typically xx or stiff honour.
This doesn't help much with your issue of a strong relayer seeking trump quality but helps a strong responder with the same doubts.
Plus often 3NT is better with xx opposite Kxxxxx, or similar.

Need a different approach with 5-carders. Comes up often when describer is balanced.
After
1♣ - 2♣ (8+ balanced)

we can stop in 2NT if no fit and a 24-count. Is that sensible?

Opener can break with a 4-carder or a hand like this:
♠K65 ♥AJ965 ♦75 ♣AK8

Easy to see relay finding partner with 4-2-3-4 or similar.
We have ways for responder to show some texture.


As you know, the other tweak we have is a tweak of your 4♦ trump ask.
This shows a hand needing "good" trumps in a suit the asker is about to disclose.
For 4+ suits, "good" = 2 of top 3, (or maybe 3 of 5). For shorter suits, "good" = 1 of top 3.
Had this recently:

---

3NT = 0-3-3-7
4♦ = bid 4♥ with any minimum, or max with no heart honour.
4♠ = max (base +2 SPs &/or +1 kontrols) ♥ honour
5♣ = raise with "good" trumps
6♣ = Hope KJT is enough.