[nullve]

Idea:

A kind of (parity) key card investigation tool where

* the trump suit is not yet agreed on;
* Teller is the one to choose the trump suit.

Typical situation where it could be used in my system:

* There are two potential trump suits from Asker's perspective
* 4♥ asks
* The highest (normal) response to 4♥ is 5 of the highest ranking potential trump suit

So for example if the potential trump suits are the roundeds, then the (normal) responses are 4♠, 4N, 5♣, 5♦ and 5♥ and the larger structure could be something like

4♠ = even KC
...4N = TQ ask
......5♣ = no TQ(♣)
......5♦ = TQ(♣)
......5♥ = no TQ(♥)
......5♠+ = TQ(♥)
...(...)
4N = odd KC(♣), no TQ(♣)
5♣ = odd KC(♣), TQ(♣)
5♦ = odd KC(♥), no TQ(♥)
5♥ = odd KC(♥), TQ(♥)
(...)

[nullve]

I'll try to illustrate with the hands from the how do you bid this? thread:

---

With the new -wood and a bit of wishful thinking, maybe:

nullve(S)-nullve(N)

1♦(1)-2♣(2)
2♦(3)-2N(4)
3♣(5)-3♦(6)
3♥(7)-3N(8)
4♣(9)-4♥(10)
4♠(11)-4N(12)
5♣(13)-5♦(14)
5♥(15)-5N(16)
7N(17)-P

(1) "10+, NAT(ish) unBAL" or "20-22 BAL"
(2) "unBAL GF with primary C"
(3) relay
(4) "16-18, either (3)4+H5+C or 1-suited"
(5) relay
(6) 1-suited
(7) relay with slam interest and willingness to bypass 3N
(8) S length "special", not 3226
(9) relay
(10) "15-17", 7C3S(21), simultaneously acting as the -wood
(11) even # of key cards (but only Opener knows what the trump suit is at this point)
(12) trump Q ask (not knowing what the trump suit is)
(13) C is trumps, no trump Q
(14) grand slam interest (ergo remaining key cards + trump Q), not 3217 (ergo 3127)
(15) ♠K ask
(16) ♠K, no ♠J
(17) contract

Added, 17 May (= Norway's birthday! ) 2021:

Spoiler

I now think responses 4♥ as Norwood(♣,♠) should be:

4♠ = even # of KC(♣)
4N = odd # of KC(♣), no trump Q
5♣ = odd # of KC(♣), trump Q
5♦ = even # of KC(♠)
5♥= odd # of KC(♠), no trump Q
5♠ = odd # of KC(♠), trump Q
(...)

Then the auction on the above hands, given Opener's decision to relay over 2N, would rather be

nullve(S)-nullve(N)

1♦(1)-2♣(2)
2♦(3)-2N(4)
3♣(5)-3♦(6)
3♥(7)-3N(8)
4♣(9)-4♥(10)
4♠(11)-5♦(12)
5♥(13)-5N(14)
7N(15)-P

(1) "10+, NAT(ish) unBAL" or "20-22 BAL"
(2) "unBAL GF with primary C"
(3) relay
(4) "16-18, either (3)4+H5+C or 1-suited"
(5) relay
(6) 1-suited
(7) relay with slam interest and willingness to bypass 3N
(8) S length "special", not 3226
(9) relay
(10) 15-17, 3(21)7 (and simultaneously Norwood(♣,♠))
(11) even # of key cards, C is trumps
(12) grand slam interest (ergo remaining key cards + trump Q), not 3217 (ergo 3127)
(13) ♠K ask
(14) ♠K, no ♠J
(15) contract

A similar auction on the hands from the 3217 Progressive thread:

---

nullve(N)-nullve(S):

1♣(1)-1♥(2)
2♦(3)-2♠(4)
2N(5)-3♣(6)
3♦(7)-3♥(8)
4♥(9)-4♠(10)
5♥(11)-5♠(12)
6♣(13)-6♦(14)
6♠(15)-6N(16)
P

(1) "10+, NAT(ish) unBAL" OR "11-13/17-19/23+ BAL"
(2) "0+, 4+ S, < 4 H unless GF"
(3) "19-21 unBAL"
(4) GF relay
(5) "5+C(3)4+H or 1-suited"
(6) relay
(7) 1-suited
(8) relay, slam interest
(9) 18-20, 3(21)7 (and simultaneously Norwood(♣,♠))
(10) even # of key cards, C is trumps
(11) confirming all key cards, trump Q, 3217, no ♠K
(12) ♥K ask
(13) ♥K, no ♠Q
(14) ♥Q ask [cannot ask directly about the ♦K]
(15) ♥Q, no ♠J
(16) contract

[nullve]

About how I got the idea and what kind of problem it is supposed to solve:

In my relay system, if x and y are given suits and hi is the highest remaining suit, then 6x4y(21) shape is shown using the structure

3♥ = 5x4y22, 6x4y(21), 8+x3(!)y or 6+x4+y0hi
...3♠ = relay, either guaranteeing slam interest and either 2+ x or 4+ y, or just looking for a 6-2 or better major suit (x=M) fit
...(...)
......(...)
......4♦ = 6x4y(21)
......(...).

(Teller's strength is already known to within 3 hcp, so this isn't quite as some relayers might think.)

For all possible values of {x,y} except one, Asker can then launch PKC(T) (= Parity Key Card Blackwood with T as trumps) at a level no higher than 4T+2. But when {x,y}={C,H}, 4♥ must be to play and 4♠ is already 4♣+3, which is one step above the safety level for PKC(♣).

So I started wondering if the needed space could be freed by letting Asker show instead of ask, as in

4♦*-?:

* 6x4y(21), x and y rounded suits

4♥ = to play
4♠+: same as 4♠+ in response to the -wood (4♥) described in the OP and used in example above

Here's an example of how this could work using the EW hands from

https://www.bridgeba...ost__p__1014475

---

East West

1♥(1)-2♣(2)
2♦(3)-2♥(4)
3♥(5)-3♠(6)
4♦(7)-4♠(8)
4N(9)-5♦(10)
5♠(11)-7♣(12)
P

(1) "10-21, 5+ H, unBAL"
(2) 3-way
(3) usually just a good MIN ("13-15"), so GF even opposite a LR (For a fuller description, see https://www.bridgeba...ost__p__1001309)
(4) GF relay
(5) "13-15", either 2542, 6H4C(21), 8+H3C or 6+H4+C0S
(6) relay
(7) 12-14, 6H4C(21)
(8) even # of key cards
(9) trump Q ask
(10) trump Q, C is trumps
(11) confirming all key cards, 2614, no ♥K
(12) contract

[Zelandakh]

The first hand should be easy for any relay system. It is particularly good for parity-cue style methods but even for basic ones you can get the job done with something like:-

1♦ = 10-17, 4+♦, unbal (if South decides to treat the hand as a simple 17)
... - 1♥ = INV+ relay
2♦ = extras, one-suited, GF
... - 2♥ = relay
2♠ = 3♥
... - 2NT = relay
3♣ = 3♠
... - 3♦ = relay
3♥ = 3361
... - 3♠ = relay
4♦ = 5 controls
... - 4♥ = relay
4NT = ♦ control, no ♠ control
... - 5♣ = relay
5♥ = ♥ control, no ♣A
... - 5♠ = relay
6♦ = 2 or 3 of top 3♦, !SQ, 1 of top 3♥
... - 6♥ = relay
7♦ = ♦AKQ, ♠J, ♥J, no ♦J

or

1♦ = 10-17, 4+♦, unbal (if South decides to treat the hand as a simple 17)
... - 1♥ = INV+ relay
2♦ = extras, one-suited, GF
... - 2♥ = relay
2♠ = 3♥
... - 2NT = relay
3♣ = 3♠
... - 3♦ = relay
3♥ = 3361
... - 4♣ = puppet to 4♦
4♦
... - 4♥ = RKCB(♦)
4NT = 0 or 3 kc
... - 5♣ = ♦Q?
5NT = ♦Q and no side king
... - 7NT (could ask about ♠Q but why bother?)

or

1♣ = 15+ nat/bal or 18+ any (if South decides to upgrade the hand)
... - 1♠ = GF with no 4+ major
1NT = 18+ (or an Acol 2)
... - 2♦ = one-suited with ♣
2♥ = relay
... - 2NT = 3♠, <3♥
3♣ = relay
... - 3♠ = 3127
3NT = to play
... - 4NT = extras, 6 controls
5NT = Q ask
... - 6♦ = ♣Q, no ♦Q
7NT

And if you check out IMPrecision auctions for this hand, you will see just how easily a well-designed parity cue relay system handles hands like this one.

Incidentally, I think you meant 5♦ rather than 5♣ for your auction on this hand as everyone seems to be assuming ♦ are trumps despite the description for #13.

I think Hand #2 just illustrates the inefficiency of the relays you want to play. Using 1♠ as INV+ or 1NT/2♣ as a pure GF relay would not run into such issues, for example:-

1♥ = 5+[hearts, unbal
... - 1♠ = INV+ relay
1NT = min, <4♠
... - 2♣ = GF relay
2♦ = 4+♣
... - 2♥ = relay
3♦ = 2614
... - 3♥ = relay
4♦ = 5 controls
... - 4♥ = relay
5♦ = controls in ♥, ♣ and ♠, no ♦A
... - 7♣/7NT = pick your poison (it is actually possible also to find out about ♥J below 6NT if that colours your decision)

7NT seems the better grand at MP; am not sure about IMPs but probably even there as if ♥ do not split we lack entries to ruff all the ♦. If we are worried about the split, it seems better to take an Ave+ board in a small slam rather than going for the in-between 7♣ option...but perhaps I am just looking at it wrong.

More generally, my view is that the answer here is really to arrange your relays such that you do not arrive at 4♦ without having shown some basic distribution beforehand, so that you do not need a crutch to handle hands that just want to sign off in 4M. That is not to say that the idea might not have legs, just that I think you should optimise the underlying system before you decide on the best way of introducing it.

[nullve]

Zelandakh, on 2021-February-23, 14:38, said:

The first hand should be easy for any relay system.

And not too hard for even a basic 2/1 system. (1♦-2♣; 2♦-4♣ etc.)

Zelandakh, on 2021-February-23, 14:38, said:

Incidentally, I think you meant 5♦ rather than 5♣ for your auction on this hand as everyone seems to be assuming ♦ are trumps despite the description for #13.

No, I meant 5♣. South knows about the 7-1 club fit but not yet about the 6-2 diamond fit.

Of course, South would often prefer a 6-2 diamond fit to a 7-1 club fit, so he could (should), knowing what I know about my system, have bid 3♦ (6+ D, slam interest) over

1♦-2♣
2♦-2N

or 3♠ (5+ D, slam interest) over

1♦-2♣
2♦-2N
3♣-3♦.

But then I couldn't have used these hands as an example. Hence the "wishful thinking".

Zelandakh, on 2021-February-23, 14:38, said:

I think Hand #2 just illustrates the inefficiency of the relays you want to play. Using 1♠ as INV+ or 1NT/2♣ as a pure GF relay would not run into such issues

Change the EW hands to

---

and my auction would start

East West
1♥(1)-1N(2)
2♣(3)-2♦(4)
3♥(5)-3♠(5)
4♦(5),

(1) "10-21, 5+ H, unBAL"
(2) "5-12, NAT, NF"
(3) "13-15", either 5H or 5S6+H OR "16-18", any
(4) 8+, relay
(5) same as in the auction starting 1♥-2♣; 2♦-2♥; 3♥-3♠; 4♦, but with "16-18" (15-17 hcp if 2614) instead of "13-15"

but Welland-Auken, for example, who use 1N as a GF relay in response their relatively standard 1♥ opening, would probably have to start something like

1♥-2♦*
2N**.

* 9-13, 2 H
** GF, 4+ C (iirc)

Change the EW hands again to

---

and my auction would start

East West
1♥(1)-1N(2)
3♥(3)-3♠(3)
4♦(3),

(1) "10-21, 5+ H, unBAL"
(2) "5-12, NAT, NF"
(3) same as in the auction starting 1♥-2♣; 2♦-2♥; 3♥-3♠; 4♦, but with "19-21" (18-20 hcp if 2614) instead of "13-15"

although West doesn't have values even for an INV+ 1♠ relay. (But I know East is too strong for a 1♥ opening in your system.)

I like the fact that even though my relay structure is not as efficient in the slam zone as many of the structures you see in this forum (for example, yours), it can be used "everywhere".

Quote

... - 7♣/7NT = pick your poison (it is actually possible also to find out about ♥J below 6NT if that colours your decision)

Instead of

(...)
5♠-7♣
P

I could have bid

(...)
5♠-6♠*
7♣**-7N
P

* ♥J ask
** ♥J, no ♠J

[not finished]

[nullve]

Another example, this time with 4♠ as a -wood:

---

1♦(1)-1♠(2)
1N(3)-2♣(4)
2N(5)-3♣(6)
3♥(7)-3♠(8)
4♠(9)-5♣(10)
6♦(11)-P

(1) "10+, NAT(ish) unBAL" or "20-22 BAL"
(2) "0+, (3)4+ H. < 4 S unless GF"
(3) 9-14, 2-H5D5C OR "10-12, 2-H6+D" OR "16-18, any" OR "20-22 BAL, < 5 H"
(4) "8+, relay"
(5) "16-18, either (3)4+ S or 1-suited"
(6) relay (GF)
(7) 4252, 6D4S(21), 8+D3(!)S or 6+D4+S0H
(8) relay [20 March 2021: Responder should just bid 3N(=T/P opposite 4252 or 6D4S(21)) here and miss slam opposite the perfecto AKxx-x-AQxxxx-Kx]
(9) 14-16, 5062 (and simultaneously a -wood)
(10) D is trumps, odd KC outside H, no trump Q*
(11) contract (happens to not be a very good one, but...)

* assuming

4N = even KC outside H
...5♣ = trump Q ask [I'm not sure exactly when it should be used. Suggestions?]
......5♦ = no trump Q, D is trumps
......5♥+: I have to think more about this. Suggestions?
......[If 5♣ can be bid with sufficient KC if D is trumps but insufficient KC if S is trumps, then the following scheme, which is similar to the one in the OP over 4♥(-wood)-4♠(even KC); 4N(trump Q ask), will not work:
......5♥ = trump Q, D is trumps
......5♠ = no trump Q, S is trumps
......5N+ = trump Q, S is trumps]
5♣ = odd KC outside H, no trump Q, D is trumps
5♦ = odd KC outside H, trump Q, D is trumps
5♥ = odd KC outside H, no trump Q, S is trumps
5♠ = odd KC outside H, trump Q, S is strumps

---

I wonder if there's a place for this kind of -wood outside my system (which I realise very few will ever play).

How about

(2♦)-4♣*-(P)-4♦**
(P)-4M-(P)-?

* Leaping Michaels, so 5+M5+C, F1
** asking for the major

?

[shyams]

A small sidetrack inspired by this (https://youtu.be/PEgk2v6KntY)

"Drop the -wegian. Norwood -- it's simple"

[nullve]

shyams, on 2021-March-04, 05:12, said:

"Drop the -wegian. Norwood -- it's simple"

Done!

[Zelandakh]

nullve, on 2021-March-04, 04:43, said:

Another example, this time with 4♠ as a -wood:

Is there some sort of bailout from this auction for ♠xx ♥KQJxx ♦Kx ♣xxxx that also keeps slam in the picture opposite ♠AKQ ♥Ax ♦ QJTxxxxx ♣-? This issue, having the known hand making the final decisions, is one that I feel you have not yet really addressed. On grand hand you can pass the ball back - no problem - but I am not sure how you can consistently get 5 vs 6 decisions right. I still think you have significant efficiency savings to find within the shape relay structure to limit the hands for which you need to use this method.

[nullve]

Zelandakh, on 2021-March-04, 06:22, said:

Is there some sort of bailout from this auction for ♠xx ♥KQJxx ♦Kx ♣xxxx that also keeps slam in the picture opposite ♠AKQ ♥Ax ♦ QJTxxxxx ♣-? This issue, having the known hand making the final decisions, is one that I feel you have not yet really addressed.

Responder would have to bid 3N over 3♥, intending to play there opposite (16-18 and) 4252 or (15-17 and) 6D4S(21). (Yes, 4♥ could then be a much better contract. )

I haven't really thought out a system other than "natural" (which is hardly optimal!*) after that, but the idea has always been that Opener should be able to bid again with every other shape, including (15-17 and) 4063.

* There is also more space available for the remaining hand types after ...3♥-3N(T/P opposite 4252 or 6S4D(21)) than after ...3♥-3♠(relay) now that Opener can no longer have 6D4S(21).

Zelandakh, on 2021-March-04, 06:22, said:

On grand hand you can pass the ball back - no problem - but I am not sure how you can consistently get 5 vs 6 decisions right. I still think you have significant efficiency savings to find within the shape relay structure to limit the hands for which you need to use this method.

I can't!

Asker will usually only know about

* the (usually) exact shape and (non-walrus) 3-point range;
* (the number of) key cards and the trump queen.

(Yes, a suit could be wide open.)

Zelandakh, on 2021-March-04, 06:22, said:

I still think you have significant efficiency savings to find within the shape relay structure to limit the hands for which you need to use this method.

I hope so!

My relay structure started as a GF 2N+ structure over

1M*-1N**

* "10-21, 5+ M, unBAL"
** "5-12, NF"

and was inspired by Bocchi-Duboin's 2001(?) card.

I've long wondered what 2N+ structure others would come up with given that Teller has already shown

* what his primary suit is (for some definition of 'primary suit' that deals with suits of equal length);
* an unbalanced (but possibly semibalanced) hand;
* approximate strength.

I guess most regulars in this forum would start with a (+4?) SR-like structure, i.e something like

2N = 5+x4+u or 3-suited
3♣ = 5+x4+v, not 3-suited
3♦ = 1-suited
3♥ = 5+x5+w
3♠ = 5+x4w, high shortage, not 3-suited
3N = 5x4w22 or 7x4w11 [unplayable, cannot F1 with 5x4w22 shape]
4♣ = 5x4w31 [unplayable, cannot bypass 3N with this shape]
4♦ = 6x4w21
4♥ = 6x4w30
4♠ = 7x4w20
etc.,

and work from there.

Of course, I don't have to play essentially the same structure in all these other situations...

[Zelandakh]

nullve, on 2021-March-04, 09:34, said:

I've long wondered what 2N+ structure others would come up with given that Teller has already shown

Personally I simply would not start from there. You have information that needs to be shown for both hand types so it can certainly be arranged more efficiently than splitting out the GF hands to 2NT+. The obvious solutions are either a forcing multi-way 2♣ rebid or (probably better) transfer rebids. I am confident that either could create a GF on the third round in a more efficient way than the immediate GF rebids. But of course, you will know by now that if going for relays I think it is better still to combine INV+GF hands together rather than weak+INV hands. This allows for SR+1 relays for 1♠ and standard relays for 1♥ [+2/+1 if you also include 18+ hands], a considerable improvement over +4!

[nullve]

Zelandakh, on 2021-March-04, 14:25, said:

Personally I simply would not start from there. You have information that needs to be shown for both hand types so it can certainly be arranged more efficiently than splitting out the GF hands to 2NT+. The obvious solutions are either a forcing multi-way 2♣ rebid or (probably better) transfer rebids. I am confident that either could create a GF on the third round in a more efficient way than the immediate GF rebids.

Are you talking about a multi-way 2♣ rebid over 1♦-1♠ (when "both hand types" might refer to

AKQ95
---
QT8732
A8

and

AKQ
Ax
QJTxxxxx
---

) or 1M-1N ("immediate GF rebids")?

Maybe it doesn't matter, since I play a kind of limited Gazzilli 2♣ (inspired by what Bocchi-Duboin played over 1M-1N) and an immediate GF 2N+ structure in both cases. The difference, which has to do with the fact that 1♦ is "10+" (with responses starting at 0) and 1M only "10-21" (with positive responses starting at ~ 5), is that

1♦-1♠; 2♣ = "19-21, any" if strong
1♦-1♠; 2N+ = "22+" (GF)
1M-1N; 2♣: "16-18, any" if strong
1M-1N; 2N+ = "19-21" (GF).

Regarding transfer rebids: I played transfer rebids over 1M-1N in a regular partnership a long time ago. (See here, inside the spoiler) Then nullve-nullve started playing transfer rebids also over

1♣-1R*,

* T-Walsh

1♦-1M

and

1♥-1♠.

At first I thought this was a huge improvement over standard, but then I realised

1) that I frequently ended up in 2N after

1x-1y;
[2z-1]-2z
2N-P

when standard players would have been able to stop (but sometimes at the risk of missing game) in 2z after

1x-1y
2z-P.

2) how hard bidding can get after e.g.

1x-1y
[2x-1]-GFR

when Opener has not further limited his hand.

Then I learned about Gazzilli and how it helps solve both of these problems.

I now use two Gazzilli-like rebids (1N and 2♣) over 1♦-1M to better be able to handle 1M responses on tradtionally subpostive hands. The focus is then on partscore bidding ratther than slam bidding, of course.

Zelandakh, on 2021-March-04, 14:25, said:

But of course, you will know by now that if going for relays I think it is better still to combine INV+GF hands together rather than weak+INV hands. This allows for SR+1 relays for 1♠ and standard relays for 1♥ [+2/+1 if you also include 18+ hands], a considerable improvement over +4!

My relay structure is not quite as bad as +4 since Opener has already shown his range. The ranges are, for each given shape, always defined in terms of (non-walrus) hcp, but I could have (although I suspect the system would suffer) defined them in terms of (say) controls or AKQ points instead. And a range-then-shape approach is not inherently less economical than a shape-then-range approach.

This post has been edited by nullve: 2021-March-05, 10:46

[Zelandakh]

nullve, on 2021-March-05, 05:53, said:

Are you talking about a multi-way 2♣ rebid over 1♦-1♠ (when "both hand types" might refer to

Yes, methods based on conventions like Gaz, Riton and the like seem like they would offer greater efficiency here.

nullve, on 2021-March-05, 05:53, said:

Regarding transfer rebids: I played transfer rebids over 1M-1N in a regular partnership a long time ago.

At first I thought this was a huge improvement over standard, but then I realised

1) that I frequently ended up in 2N after

1x-1y;
[2z-1]-2x
2N-P

Opener's 2NT here should show around 16-17hcp. This should not be a different result from a pair playing natural rebids.

nullve, on 2021-March-05, 05:53, said:

1x-1y
2z-P

This auction shows where there really is a notable difference. Here Opener with the same 16-17 hand will make an additional try that would not have been available in the natural auction, so you end up in 2NT rather than 2z with ~22-24hcp and a 4-3 fit in z. I personally do not see this as a significant flaw, particularly given that the defence in 2z after this auction tends to be rather easy.

nullve, on 2021-March-05, 05:53, said:

2) how hard bidding can get after e.g.

1x-1y
[2x-1]-GFR

I am not sure I want to get into this one as we were previously talking specifically about a SF 1NT response. Whether it still makes sense to go for a relay structure after Responder has started to describe their hand is an interesting question but somewhat distinct from the main one. Most pairs that use full relays have Responder use only artificial calls on the way to that relay auction. This avoids inefficient usage of bidding space.

nullve, on 2021-March-05, 05:53, said:

My relay structure is not quite as bad as +4 since Opener has already shown his range. The ranges are, for each given shape, always defined in terms of (non-walrus) hcp, but I could have (although I suspect the system would suffer) defined them in terms of (say) controls or AKQ points instead. And a range-then-shape approach is not inherently less economical than a shape-then-range approach.

When I spoke about SR+1 (for 1♠) and SR+0 (for 1♥) I was including strength division between 2 bands (for me 10-13 and 14-17 but you would presumably adapt that to 12-14 and 15-17). That is why I added the proviso that including 18+ hands would add an additional step. You do indeed get the step back versus SR methods that use an "extras" bid after shape resolution but I find the comparison with standard SR for the shape resolution stage nonetheless useful.

[nullve]

Zelandakh, on 2021-March-05, 09:38, said:

Opener's 2NT here should show around 16-17hcp. This should not be a different result from a pair playing natural rebids.

Aargh, I meant to write

1x-1y
[2z-1]-2z
2N-P.

Corrected now.

[nullve]

Hands from the Take over the bidding.. thread:

---

South-North:

1♥(1)-2♣(2)
3♥(3)-3♠(4)
4♦(5)-4N(6)
6♣(7)-P

(1) "10-21, 5+ H, unBAL"
(2) 3-way
(3) "16-18", either 2524, 6H4C(21), 8+H3(!)C or 6+H4+C0S
(4) relay
(5) 15-17 hcp, 6H4C(21), simultaneously Norwood(♣,♥)
(6) C is trumps, odd # of key cards, no trump Q
(7) contract

Added, starting 20 April 2021:

Spoiler

Hands from the I have a hand for you thread with a very similar auction:

---

North-South:

1♣(1)-1♥(2)
2♦(3)-2♠(4)
2N(5)-3♣(6)
3♥(7)-3♠(8)
4♦(9)-4N(10)
5♦(11)-6♣(12)
P

(1) "NAT(ish) unBAL or 11-13/17-19/23+ BAL"
(2) "0+, 4+ S"
(3) "19-21 unBAL"
(4) GF relay
(5) "(3)4+H5+C or 1-suited"
(6) relay
(7) either 2425, 6C4H(21), 8+C3(!)H or 6+C4+H0S
(8) relay
(9) 18-20 hcp, 6C4H(21), simultaneously Norwood(♣,♥)
(10) C is trumps, odd # of key cards, no trump Q
(11) remaining key cards, the trump Q, not 2416 (hence 1426)
(12) contract

9 September 2021: Hands from the Here's a nice hand (bid it MY way) thread:

---

(rotated 180 degrees for convenience)

1♣(1)-1♠(2)
2♣(3)-2♠(4)
2N(5)-3♣(6)
3♥(7)-3♠(8)
4♦(9)-5♣(10)
5♠(11)-7N(12)
P

(1) "10+, NATish unBAL" OR "11-13/17-19/23+ BAL"
(2) "D or BAL, T-Walsh style"
(3) "13-15, 5+ C, unBAL" OR 14-16, 4414
(4) GF relay
(5) "(3)4+ H or 1-suited"
(6) relay
(7) 2425, 6C4H(21), 6+C4+H0S or 8+C3(!)H
(8) relay
(9) 12-14, 6C4H(21)
(10) odd # of key cards, trump Q, C is trumps
(11) all remaining key cards, not 1426 (ergo 2416), ♥K, no ♠K
(12) contract


5 September 2021: Hands taken from the Easy Squeezy? thread:

---

(rotated for convenience)

1♣(1)-1♥(2)
2N(3)-3♣(4)
3♥(5)-3♠(6)
4♦(7)-4N(8)
5♦(9)-5♥(10)
5N(11)-6♦(12)
6♥(13)-7♣(14)
P

(1) "10+, NATish unBAL" OR "11-13/17-19/23+ BAL"
(2) "0+, 4+ S"
(3) "GF, either 5+C(3)4+H3-S or 1-suited"
(4) relay
(5) 2425, 6C4H(21), 6+C4+H0S or 8+C3(!)H
(6) relay
(7) 21+ hcp, 6H4C(21)
(8) odd # of key cards, no trump Q, C is trumps
(9) remaining key cards, trump Q, not 2416 (ergo 1426)
(10) ♥K ask
(11) ♥K, ♦K
(12) ♥Q ask
(13) no ♥Q
(14) contract

[Zelandakh]

nullve, on 2021-April-15, 01:31, said:

(3) "16-18", either 2524, 6H4C(21), 8+H3(!)C or 6+H4+C0S

Again, here you have reached 3♥ with very little useful distributional information conveyed. Compare that with a regular relay auction:-

1♥ = 5+♥
... - 1♠ = relay
2♦ = extras 4+♣
... - 2♥ = relay
3♣ = 0634 or 1624

where we are 2 steps lower and Opener's hand is described between 2 very similar hand types. This remains the basic issue with your system. Get the fundamentals right first and then decide if you need an additional crutch such as "Norwegian Wood".

[nullve]

To indicate why I play 3♥ this way, here are some structures between the (obviously unplayable) sym+4 structure in post #10 and my own structure:

1)

2N = 5+x4u or 5+-5+ or (4441) (if possible)
3♣ = 5+x4v
3♦ = 1-suited
3♥ = 5+x4w, high uneven shortage
3♠ = 5+x4w, even shortage
3N = 5x4w31 (5x4w40)
4♣ = 6x4w21
4♦ = 6x4w30
4♥ = 7x4w20

where 5x(440) is temporarily treated as 5x(431), which frees up 2N-3♣; 3♦.

2)

2N = 5+x4+u or 5-5 (if possible) or (4441) (if possible)
3♣ = 5+x4+v, not 5-5
3♦ = 1-suited
3♥ = 5+x4+w, high uneven shortage, not 5-5
3♠ = 5+x4+w, even shortage
3N = 5x4w31 (5x4w40)
4♣ = 6x4w21
4♦ = 6x4w30
4♥ = 6x5w20
4♠ = 7x4w20

3)

2N = 5+x4+u or 1-suited, not 5-5
3♣ = 5+x4+v, not 5-5
3♦ = 5-5 or (4441)
3♥ = 5+x4+w, high uneven shortage, not 5-5
3♠ = 5+x4+w, even shortage
3N = 5x4w31 (5x4w40)
4♣ = 6x4w21
4♦ = 6x4w30
4♥ = 6x5w20
4♠ = 7x4w20

(I think it's much better to have the 1-suiters in 2N in case Asker wants to break out of the relays.)

4)

2N = 5+x4+u or 1-suited, not 5-5
3♣ = 5+x4+v, not 5-5
3♦ = 5-5 or (4441)
3♥ = 5x4w22 or 6+w4+w and high uneven shortage
...3♠ = relay
......3N = 5x4w22
......4♣ = 6x4w(21)
......4♦ = 6x4w03
......4♥ = 6x5w02
......4♠ = 7x4w02
...3N = to play opposite both 5x4w22 and 6x4w(21)
...(...)
3♠ = 5x4w13 (5x4w04)
3N = 5x4w31 (5x4w40)
4♣ = 6+x4+w, even shortage
4♦ = 6x4w30
4♥ = 6x5w20
4♠ = 7x4w20

5)

2N = 5+x4+u or 1-suited, not 5-5
3♣ = 5+x4+v, not 5-5
3♦ = 5-5 or (4441)
3♥ = 5x4w22, 6x4w(21) or (6+w4+w, high uneven shortage)
...3♠ = relay, willing to bypass 3N opposite 6x4w(21)
......3N = 5x4w22
......4♣ = 6x4w03 (allows 4♦ = stop signal; 4♥/♠/N = PKC in the lowest/middlemost/highest suit, respectivly, of x, w and fragment)
......4♦ = 6x4w(21) (both PKC(x) and PKC(w) avaiblable at a sufficiently low level unless {x,w}={♣,♥})
......4♥ = ? (this is where I put 8x3w, which might be silly)
......4♠ = 6x4w02
......4N = 7x4w20
...3N = to play opposite both 5x4w22 and 6x4w(21)
...(...)
3♠ = 5x4w13 (5x4w04)
3N = 5x4w31 (5x4w40)
4♣ = 6x4w30 (allows 4♦ = stop signal; 4♥/♠/N = PKC in the lowest/middlemost/highest suit, respectivly, of x, w and fragment)
4♦ = 6x5w11
4♥ = 7x4w11
4♠ = 6x5w20
4N = 7x4w20

But if I knew how to use 1♠ or 1N instead of 2♣ as the relay response to 1M without hurting my partscore bidding*, then I would certainly adopt a more symmetric structure.

* In the Z Club, canapé version thread I suggested that this might be easier to do in a canapé context because 1M-2♣ could be used as P/C rather than NAT, NF.

---

I know the above auction to 6♣ isn't very impressive, but how do you think it compares to the 2/1 auctions in the Take over the bidding.. thread?

This post has been edited by nullve: 2021-April-16, 06:18

[Zelandakh]

nullve, on 2021-April-16, 02:06, said:

But if I knew how to use 1♠ or 1N instead of 2♣ as the relay response to 1M without hurting my partscore bidding*, then I would certainly adopt a more symmetric structure.

* In the Z Club, canapé version thread I suggested that this might be easier to do in a canapé context because 1M-2♣ could be used as P/C rather than NAT, NF.

Sorry I thought that thread was primarily about the 2♣ opening and stored it away for some thought later.

Some years ago I had a discussion with Adam (arguably the best system designer at BBF) about expanding the range of a 2♣ response to the 1♠ opening (I think to 3+♣). No doubt you can find that thread so if you like his ideas, by all means use them. When I went through all of the possibilities I found that for each advantage there was a corresponding disadvantage and therefore did not adopt it, But sure, pairs such as Welland-Auken play 2♣ as their NFNT response so 2♣ as a more limited version of the same is surely more than playable. The real key to the NF responses, at least according to the fairly extensive playtesting I did during system development, is to bid suits up-the-line regardless of length. While there remain some cases that are problematic, such as 4441 opposite 3325 (1♦ - 2♣; P), these are typically matched by problem cases in 2/1 (1♦ - 1NT; 2♦ - P?).

Where part-score bidding can be hurt is when Responder has an invitational hand and Opener is minimum. Now a minor suit fit can easily get buried so we play 2M or 2NT rather than 3m. This can also happen in 2/1 but somewhat less often. As I mentioned, I did a large amount of testing to make sure that the disadvantages were less than the advantages. For this specific system, with openings limited to 17 and always unbalanced, this was true. It is quite possible that a different opening bid structure, particularly something like allowing balanced hands in the openings, would lead to more disadvantages than advantages.

In any case, it might be a worthy discussion, either open or private, to go through your specific concerns. The time when responding 2♣ rather than 2♦ gains is typically when Responder has something like 1(54)3 and Opener 5(21)5 (1♠ - 2♦; 2♠ - 2NT; 3♣). It typically loses when Opener has 3 clubs, or in your 2+ version 3-4 clubs (1♠ - 2♣; P). So you have to decide which hands you want to stop at the 2 level and which are forced to the 3 level. I suspect you will like Adam's structure though; it might be just the compromise you are looking for.

[nullve]

nullve, on 2021-March-04, 04:43, said:

I wonder if there's a place for this kind of -wood outside my system (which I realise very few will ever play).

In this thread on Bridgewinners, Kit Woolsey claims that 4N in the standard 2/1 sequence

1♠-2♥
3♦-3♥
4♦-4N

is RKC, not natural. (Thanks for mentioning this, Benoit!)

But if not natural (I don't care), then it seems that Norwood could be used to check for key cards and the trump queen also in the case where Responder has a slam going hand with self-sufficient hearts:

4N = even KC
...5♣ = trump Q ask
......5♦ = no trump Q, D is trumps
......5♥ = no trump Q, H is trumps
......5♠+ = trump Q (details?)
5♣ = odd KC, no trump Q
...5♦ = insufficient KC if D is trumps
...5♥ = sufficient KC if D is trumps, insufficient KC if H is trumps
...5♠+ = sufficient KC (details?)
5♦ = odd KC, trump Q, D is trumps
5♥ = odd KC, trump Q, H is trumps

A similar Norwood, but this time with three potential trump suits, could be used to solve another problem in my own system: How to check for key cards and the trump queen in any of the potential trumps suits D, H and S after

...4♥ = <3-point range>, 5530

if 4♠ would to play.

I.e. something like:

4N = even KC outside C
...5♣ = trump Q ask
......5♦ = no trump Q, D is trumps
......5♥ = no trump Q, H is trumps
......5♠ = no trump Q, S is trumps
......5N+ = trump Q (details?)
5♣ = odd KC outside C, no trump Q
...5♦ = insufficient KC if D is trumps
...5♥ = sufficient KC if D is trumps, insufficient KC if H is trumps
...5♠ = sufficient KC if either D or H is trumps, insufficient KC if S is trumps
...5N+ = sufficient KC (details?)
5♦ = odd KC outside C, trump Q, D is trumps
5♥ = odd KC outside C, trump Q, H is trumps
5♠ = odd KC outside C, trump Q, S is trumps

This post has been edited by nullve: 2022-February-13, 08:47