[Kungsgeten]

In our current relay methods we play key card asking bids (and 4D as an end signal) after shape is resolved. We've previously played "control ask" (asking for aces and kings), followed by DCB, but gave that up because we weren't comfortable with it (the auctions took a lot of time, we would have to count on our fingers to see which steps should be skipped, etc).

A problem with our current approach is that we don't really have a slam invite available. We know opener's shape, but not if he's min or max. I've been thinking that we maybe should have some sort of strength inquiry (for instance QPs, or maybe control points) after shape is resolved. Would this be wise, when we're not playing DCB or parity check? Some ideas I'm considering:

- After shape is resolved, step one asks for QP. Other steps is RKC.
- After QP is shown, our bids are RKC (lowest step for longest suit, etc).
- If 4D is available after the QP ask, it is still an end signal.
- Some variation of Mulberry (slam invite in any suit, RKC in any suit, or sign off in any suit).
- Use optional RKC instead of normal RKC (if 4C asks for key cards, a 4D response would show a minimum and 4H+ shows key cards and extras).
- Simply "man up" and practice DCB style bidding until we're comfortable with it.

What's your experience playing QP asks? Is it often that slam is out of the picture after a response? If not playing DCB, the idea of QP ask would be to limit the hand further. We currently play relays after our 1M openings (5+M, 11-16) and our 1D opening (4+D unbalanced, 11-19), what QP ranges would you recommend for these?

[Zelandakh]

Kungsgeten, on 2017-September-12, 10:44, said:

Some ideas I'm considering:

Another option is to use 4NT any time before controls are asked for as a natural invite. With a fit, you generally get more from asking about controls (whether through scanning or RKCB) than using any sort of power ask.

Kungsgeten, on 2017-September-12, 10:44, said:

What's your experience playing QP asks? Is it often that slam is out of the picture after a response?

As I have posted previously, I am not a QP user but I would say that one of the big advantages of integrating QPs into your methods directly into the response structure is often knowing immediately whether slam is in the picture or not.

[shevek]

As in the thread I started, we use first step = SP, 2nd = kontrols.
Don't play end signal and the RKCs associated with that.
Not because it's no good, rather because I was concerned that one of us would forget and pass RKC. Plus a bit messy.

Anyway, we use a safer 4D, based on an idea devised by David Morgan.
After shape - but not strength - is out, 4D non-step says "I am about to make a slam invite in some suit, needing extras including good trumps"
Extras being defined as Base+2 SP &/or Base+1 kontrols. Good suits were Hx(x) or HHxx(x) where H = A/K/Q.
This has worked very well. 4♥ would show min, or max with bad hearts, etc. Sort of natural DCB. 4NT = resurrection strength.
If 4D non-step is available AFTER an SP ask, it says "raise my next bid with good trumps"

[nullve]

Have you considered structures (other than Ambra's) of the form

1M-2♣; ?:

2♦ = MIN, denies shape of type T
...2♥ = relay
......2♠+ = R(M)
2♥ = shapes of type T
2♠+ = MAX, denies shape of type T, R(M),

where R(M) is a symmetric relay structure and MIN/MAX is defined in terms of hcp, QP, controls or similar?

I don't play this myself, but I do like to limit Opener's hand before shape is resolved, since he's going to declare 4M a lot of the time and I want to avoid unnecessary information leakage. Also, transpositions to more standard symmetric should be possible in many cases. E.g. the bidding might go

1♠-2♣
2♦*-2♥
2♠**-2N
3♦***-3♥
3♠****

* MIN
** C or 3-suited
*** 2-suited, H shortage
**** 5134

instead of (e.g.)

1♠-2♣
2♦*-2♥
2N**-3♣
3♦***-3♥
3♠****.

* C or 3-suited
** 2-suited, H shortage
*** 5134
**** MIN

[Kungsgeten]

nullve, on 2017-September-14, 04:05, said:

Have you considered structures (other than Ambra's) of the form

...

where R(M) is a symmetric relay structure and MIN/MAX is defined in terms of hcp, QP, controls or similar?

When we played 1D-1H as two-way (natural or GF relay) we separated min/max hands early, both because it was needed (in case responder had the non-GF hand), but also because there was enough space to do it. When the relay start at 2C I think you're short on space if you want to show both shape and strength. On the other hand we're currently not playing "full shape", since we only can find out about the following hand types:

- 5422
- 5431
- 5440
- 4441
- 5332 (but not in which suit the doubleton is)
- 6+ suit, 4 in a side suit, and information about the short suit (but not if it is a singleton or a void). Not possible to show two singletons.
- 5+ 5+ in two suits, plus shortness, but not if one of them is six+ and not if it is a void or not. Not possible to show two singletons.
- 6 card suit without shortness
- 6 card suit with shortness, min or max
- 7+ suit with/without shortness.

The reason is mostly because of how "easy" the structure is, not because of effectiveness.

Showing strength first with two-suiters will probably get us past 3NT with some hands where we want to play 3NT?

[Zelandakh]

shevek, on 2017-September-14, 00:27, said:

As in the thread I started, we use first step = SP, 2nd = kontrols.
Don't play end signal and the RKCs associated with that.
Not because it's no good, rather because I was concerned that one of us would forget and pass RKC. Plus a bit messy.

There is an alternative way of playing 4-way KCB other than the 4♦ end-signal. It is using the 4m bids for key cards, with 4M and 5m thereby retaining their natural meanings. 4♦ is RKCB for clubs. 4♣ instead a puppet to 4♦, to be followed by 4♥ RKCB ♦, 4♠ RKCB ♥ or 4NT RKCB ♠. This obviously does not work once the bidding gets to 3♠ or higher though (there is a fix for 3♠/3NT but it involves an artificial 4♥, which might be a memory issue for you).

[nullve]

Kungsgeten, on 2017-September-14, 07:05, said:

Showing strength first with two-suiters will probably get us past 3NT with some hands where we want to play 3NT?

Well, here's a MIN/MAX-showing Ambra-like structure that resolves more or less like +2 symmetric:

1♥ = "5+ H", MAX w/ 5S5H possible
1♠ = "5+ S", MAX w/ 5S5H impossible

1M-2♣ = GF relay

1M-2♣; ?:

2♦ = MIN, 3- OM
...2♥
......2♠+ = A(M)
2♥ = 4+ OM
...2♠
......2N = MAX*, not 5S5H
.........3♣
............3♦+ = B(M,OM)
......3♣ = 5S5H (so MAX/MIN if M=♥/♠, respectively)
.........3♦
............3♥+ = C(M,OM)
......3♦+ = MIN*, B(M,OM)
2♠+ = MAX, A(M)

* Maybe better to switch MAX and MIN here?

A(M):

2♠ = 4+ D
...2N
......3♣ = 5M5D / ?
.........3♦
............3♥+ = C(M,♦)
......3♦+ = B(M,♦)
2N = 1-suited
3♣ = 5M5C / ?
...3♦
......3♥+ = C(M,♣)
3♦+ = B(M,♣)

B(M,y):

3♦ = 5+M4+y, high shortage, not 5M5y
3♥ = 5+M4+y, even shortage
...3♠
......3N = 5M4y22
......4♣ = 6M5y11
......4♦ = 7M4y11
......etc.
3♠+ = 5+M4+y, low shortage, not 5M5y
I.e. something like
3♠ = 5M4y31
3N = 6M4y21
4♣ = 6M4y30
4♦ = 6M5y20
4♥ = 7M4y20
etc.

C(M,y):

3♥ = 5M5y, high shortage
3♠ = ? [available for 5M4y04 or 5M4y40, if necessary, although there might be a better way to bid these shapes]
3N+ = 5M5y, low shortage.

So maybe the answer to your question is "no, not necessarily"?

[straube]

Kungsgeten, on 2017-September-12, 10:44, said:

We currently play relays after our 1M openings (5+M, 11-16) and our 1D opening (4+D unbalanced, 11-19), what QP ranges would you recommend for these?

I would use 6-10 for the 11-16 range. I'm not sure what ceiling I would use for 19 hcps.

It might be easier to answer some of your questions if we had examples of how high you are resolving patterns. It sounds like maybe you are +1 or +2 compared to symmetric relay? In any case, I think you're well-placed to guess whether slam is on if you just know pattern and QP total. If you need partner to have some perfecta (i.e. one or two combinations of cards that would produce slam) you just give up because you're too high to find out. If he has enough QPs so that many combinations produce slam, there you are.

Obviously you're aware of parity cue bidding. I learned that from awm and sieong and it's worked great and seems easier to use than DCB.

[Kungsgeten]

straube, on 2017-September-21, 19:49, said:

It might be easier to answer some of your questions if we had examples of how high you are resolving patterns.

We resolve 5440 patterns very low (too low, but it is because we try to keep the system simple).

* 5-4-4-0 = 2NT
* 5-0-4-4 = 3D
* 5-4-0-4 = 3H

5422 pattern resolves at 3D.

5431 = 3H
5-5 + known shortness = 3S (can't find out if 6-5 or if void)
6-4 + known shortness = 3NT (can't find out if 7-4 or if void)

Any 5332 is at 2S. Now we do not relay for the doubleton, we use the following:

2S--2NT (bidding suit at the 3-level is natural);
3 new = Soft values in suit
3M = Weak opening suit
3NT = None of the above

With one-suiters we use:

2NT = 7+ suit, or max with 6322.
3C = Min with 6-card suit + singleton.
3D = Min with 6322 (could perhaps be bad 7222 too)
3H/S/NT = Max with 6-card suit and shortness up the line.

[straube]

You might consider lumping the first two QP steps together....so the first step would be 6 or 7 (if your base is 6) and then 8, 9, and 10 zooms. You can then re-ask for the 6 vs 7 if captain is still interested.

This fits pretty well with something I like if pattern is resolved at either 3S or 3N.

4C-asks QPs
.....4D-6 or 7
.....4H-8
.....etc

4D-asks QPs when captain is interested in hearts
.....4H-6 or 7
.....4S-8
.....etc

4H-is to play but slave can super-accept
.....4S-9 (a bit optional here as slave knows pard isn't interested in 8)
.....4N-zoom with 10

If pattern resolves at 3H or below, you wouldn't have any use for this idea and you could incorporate RKC or control asks or whatever.
I've been really satisfied with QPs and Parity Cue Bidding and pretty much would only consider RKC if I'm interested in a minor suit slam.

I've seen before (where pattern ends at 3S or 3N)

4C-QP ask
4D-terminator
4M-optional RKC ask

which makes sense in that if slave overrides, he thinks slam is on unless the partnership is specifically missing key cards. So that might be better since slave is limited, but if slave isn't limited (in the case of reverse relays) then I'd rather just stay with QPs.

Regarding the 5440s, I've given up the ability to even show those (I could tweak if I wanted, but they're rare) and my 5332s are shown at 3C and then the doubleton is shown at 3H, 3S, and 3N. For me, the 5332s are 11-13 so knowing the point range is so small and considering that they seldom produce slam, I don't mind resolving them high. I suppose you use 1M-2C as GF relay, too? If you're interested, I'll post what I do for that.

[Kungsgeten]

nullve, on 2017-September-21, 13:30, said:

Well, here's a MIN/MAX-showing Ambra-like structure that resolves more or less like +2 symmetric

I quite like this one. The downside is that you get quite high with single-suiters. The posted structure doesn't include 5332 and not 5440 shapes, but perhaps one could do something like this:

1M-2C;
2D = Any minimum, not 4oM and not 5-5. Could be 5440 min or max.
...2H = R.
......2S = 4C min or 5332 (rebids 3C)
......2NT = 6+M, min.
......3C = Any 5440, min or max (can not find out). Relay for void.
......3D+ = 4D, min.
2H = 4+oM, but not 5440.
...2S = R.
......2NT = 4oM, min.
......3C = 5(+)oM, min or max.
......3D+ = 4oM, max.
2S = 4+C, max or 5-5.
2NT = 6+M, single-suited and max.
3C = 5(+) diamonds, min or max.
3D+ = 4 diamonds, max.

Over 5-5 shape we could relay for min/max and shortness:

3H = High shortness, min.
3S = Low shortness, min.
3NT = High shortness, max.
4C = Low shortness, max.

[Zelandakh]

Kungsgeten, on 2017-October-07, 01:09, said:

I quite like this one. The downside is that you get quite high with single-suiters. The posted structure doesn't include 5332 and not 5440 shapes, but perhaps one could do something like this:

You get more space if you use 1M+1 as an INV+ relay and then 1M+3 as the GF where partner makes the minimum response:-

eg 1♥ - 1♠
-----
1N = min with <4 spades
... - 2♣ = GF relay (then shape responses as per 2♦+ in first rebid)
... - 2♦♥♠NT/3♣ = nat, INV
2♣ = 4+ spades
... - 2♦ = GF relay
... - ... - 2♥ = min, 4 spades and not 3-suited
... - ... - 2♠ = 5+ spades or 3-suited
... - ... - 2N+ = max, 4 spades and not 3-suited
... - 2♥♠NT/3♣♦ = nat, INV
2♦ = max, 4+ clubs, GF
2♥ = max, 1-suiter, GF
2♠+ = max, 4+ diamonds, GF
--

This structure also has the benefit of handling weak hands more effectively than most 2♣ GF relay methods. The main downside is the potential for interference over the initial relay before being in a GF auction.

[straube]

I don't like the idea of separating min/max before shape resolution. It's arguably less important information but also you then have to have different continuations depending on whether partner has given you min/max information or not. Why not use a symmetric structure and then do your QPs after like most folks would do? You're usually +1 anyway which is acceptable. These min/max ideas leave you +2 or more which is very difficult. Here's Adam's structure from a prior thread...

awm, on 2017-August-15, 11:05, said:

Here's our relay structure over 1S-2C for what it's worth:

2D: clubs or one suited without shortness
... 2H relay and 2S one suited, 2N+ like below
2H: diamonds
... 2S relays and 2N+ like below
2S: one suited with some shortness
... 2N asks shortage
2N: 5-5 or more majors
3C: 4 hearts high shortage
3D: 5422 or 5440 or 7411
3H: 5431
3S: 6421
3N: 6430
4C+: 7420

Here's what I use (and it treats 5440s as 5431 which is probably distasteful, but 5440s are very costly). There are lots of these schemes floating around, so rearrange to your liking....

2D-4C, 5+C, 5332, 5+D
.....2S-4C
.....2N-5+C
.....3C-5332
.....etc-5+D
2H-single-suited with 6
2S-4D
2N-5OM
etc-5+M/4M

[nullve]

Kungsgeten, on 2017-October-07, 01:09, said:

The downside is that you get quite high with single-suiters.

What if

...2N(6+ M, 1-suited)-?:

3♣ = strong relay (+ 2 symmetric)
...3♦ = high singleton/void or 6M2(32)
......3♥ = relay
.........3♠ = 6M2(32)
............3N = relay(!)
.........3N+ = high singleton/void
...3♥ = middle singleton/void
...3♠ = 6M322 or 7M222
......3N = relay(!)
...3N+ = low singleton/void
3♦ = weak relay (uncertain about best game)
...Something like:
...3♥ = 6M133 or 6M(322)
......3♠ = wants to avoid 3N opposite 6M133
.........3N = 6M(322)
.........4♣ = 6M133
......3N = to play opposite 6M133
...3♠ = 6M313 (=> 3N = to play opposite 6M313)
...3N = 6M331
...4M = 7+ M
(...)

?

[nullve]

Zelandakh, on 2017-October-07, 05:46, said:

You get more space if you use 1M+1 as an INV+ relay and then 1M+3 as the GF where partner makes the minimum response:-

I've never quite understood how you survive when Responder has < INV values, but here are two ideas:

I) 1♠-?:

1N = your INV+ relay
2♣/♦ = same as in the Welland-Auken system, but with all INV hands removed
2♥ = freed up

The point is that if the 2m responses work for Welland-Auken, then these more limited 2m responses to your more limited 1M openings should work at least as well and probably better.

2) Using the 1N/2♣ responses to 1♥ the same way as the 2♣/2♦ responses to 1♠, respectively. So over 1♥, something like

P = normal
1♠ = INV+ relay
1N = "5-8, 2- H" or "9-10, 1- H" or "bad 3c raise"
...2♣ = "3+ S" (similar to 2♦ = "3+ H" (1-under transfer) over 1♠-2♣)
......(...)
......2♥ = "5-8*, 3-S2H" or "bad 3c raise"
......(...)
...2♦ = "3-S6+H" (1-under transfer)
...2♥ = "2-S5H4+m" (Muiderberg)
......P = "5-8*, 2 H" or "bad 3c raise"
......(...)
...(...)
2♣ = "9-10, 2 H" (must have a way to explore a S fit after this)
2♦: freed up
2♥ = "constructive 3c raise"
(...)

* the "point" of the 2♣ response?

[Zelandakh]

nullve, on 2017-October-09, 14:45, said:

I've never quite understood how you survive when Responder has < INV values,

Weak hands are simplicity itself - bids are essentially natural and non-forcing except that 1♥ - 1NT shows spades (but is still non-forcing).

1♥
--
1♠ = INV+ relay
1N = weak, 4+ spades
2♣ = weak, 4+ clubs
2♦ = weak, 5+ diamonds
--

1♠
--
1N = INV+ relay
2♣ = weak, 4+ clubs
2♦ = weak, 4+ diamonds
2♥ = weak, 5+ hearts
--

What I discovered here during playtesting is that it is generally better to bid up-the-line, even for example showing a 4 card minor ahead of a 5 card heart suit over 1♠ unless holding extreme (at least 5-5) shape. This in turn means that a sequence like 1♠ - 2♥; 2♠ - 3m shows at least 5♥-5m. The other point to mention is that I play my 1M openings as unbalanced, meaning that Opener's 2NT rebid is free; I use this as a strong 3 card raise, essentially covering "death hand" types.

The end result is a system that handles weak hands well, not only in isolation but also for the extra pressure exerted on 4th seat. The bigger weakness of the method imho comes with INV hands, which have limited sequences available and are the cause of not being in a GF after the initial relay. I do think that the positives outweigh the negatives though, at least opposite the semi-limited range for which the method was originally designed.

[straube]

I would want to test something like below. I've kind of done so already for a similar situation and found that staying out opener's way was useful. Opener is encouraged to make his normal rebid and responder often gets a chance to raise.

1♠
--
1N = INV+ relay
2♣ = weak, 3+ clubs
2♦ = weak, 5+ diamonds and 2-card preference to clubs (so 5D/4C would respond 2C)
2♥ = weak, 6+ hearts

On a side note, it's always nice how you always list both pro and con for your methods.

[Zelandakh]

straube, on 2017-October-10, 01:43, said:

I would want to test something like below. I've kind of done so already for a similar situation and found that staying out opener's way was useful. Opener is encouraged to make his normal rebid and responder often gets a chance to raise.

1♠
--
1N = INV+ relay
2♣ = weak, 3+ clubs
2♦ = weak, 5+ diamonds and 2-card preference to clubs (so 5D/4C would respond 2C)
2♥ = weak, 6+ hearts

On a side note, it's always nice how you always list both pro and con for your methods.

Thank you. That just seems like common sense on a public forum. If anyone actually tries it they will find out the awkward parts fairly quickly so hiding them brings nothing. What I personally like best about the structure is the efficient division of strength ranges, approximately 10-13/14-17 opposite INV and GF hands but 10-15/16-17 when Responder is weak. This would carry over into your structure so it ought to retain a similar efficiency. What I am unclear on is how you deal with a weak 2542 or 1552 though. The latter in particular seems to be difficult to handle - perhaps 2♥ should include such 5♥-5♦ hands in addition to 6+♥?

[awm]

A few question/comments for Zel:

1. Part of why this structure never seemed playable to me is openers 5(332) patterns. Removing them from 1M fixes a lot of issues on the responses but may cost you when opener has these hands (may miss major fits for example).
2. What is openers rebid with 51(43) after 1S-2H? How about with 5314 or 5224 after 1S-2D? It seems like passing could be really bad, bidding 2S sounds like six, but bypassing 2S could miss the last playable spot?
3. When does opener pass 1S-2C? Does he ever raise?
4. If responder has an invite with 3S, does he respond 1NT to 1S? If so, and openers rebid could be minimum, what’s his next call? It seems like distinguishing 2 vs 3-fit here matters in case opener has extra shape, but of course you often want to play 2S and not 2NT on the 5-2 and you don’t necessarily want to play 3S on the 5-3?

[Zelandakh]

awm, on 2017-October-10, 03:44, said:

A few question/comments for Zel:

1. Part of why this structure never seemed playable to me is openers 5(332) patterns. Removing them from 1M fixes a lot of issues on the responses but may cost you when opener has these hands (may miss major fits for example).
2. What is openers rebid with 51(43) after 1S-2H? How about with 5314 or 5224 after 1S-2D? It seems like passing could be really bad, bidding 2S sounds like six, but bypassing 2S could miss the last playable spot?
3. When does opener pass 1S-2C? Does he ever raise?
4. If responder has an invite with 3S, does he respond 1NT to 1S? If so, and openers rebid could be minimum, what’s his next call? It seems like distinguishing 2 vs 3-fit here matters in case opener has extra shape, but of course you often want to play 2S and not 2NT on the 5-2 and you don’t necessarily want to play 3S on the 5-3?

Hi Adam, as always your feedback is welcome and absolutely on point!
1. I treat 5M(332) as balanced, meaning these hands are either opened 1NT or 1♣. This was a design decision and therefore the whole response structure is predicated on the 1M openings being unbalanced.
2. In all cases the rebid is 2M for hands in the lower (10-15) range precisely as in Standard. Typically going beyond 2M requires extra strength (for Opener) or extra shape (for Responder). This is essentially the reason why suits are bid up the line - the idea is to find any sort of fit and stop. If the auction starts 1♠ - 2♥ then Responder either has a heart one-suiter or is 5♥-5m. Either way, Responder should be in a position to know what to do between passing 2♠ or continuing. 1♠ - 2♦ with 5314 is more awkward and it is possible to miss a heart fit here - but then a similar situation can occur in the corresponding F1NT auction too. Since the structure for weak hands is based around natural bidding, it is common for similar concepts to come up as occur in regular methods. Most of the time limiting the hand more fully and showing a suit gives a small advantage but there are a few cases where Responder bidding a suit preempts our natural resting place, such as the above 5314 opposite 1543 (we might have to play 2♠ instead of the better 2♣ (or 2♥)).
3. Opener passes 2♣ with any hand having 3+ clubs and no game interest. A raise to 3♣ shows 16-17 and 4+ clubs. A "raise" to 2NT shows 16-17 with 3 clubs and 6+ spades. In theory Opener could also raise to, for example, 4♣ with slam interest but that is unlikely enough to be ignored.
4. Yes, and this is an area where I have followed your threads on the subject with some interest. At present I use the sequences 1♠ - 1NT; 2♣ - 2♠ and 1♠ - 1NT; 2♦ - 2♠ as equivalent to Drury with a 3 card limit raise. That means that hands with a 5-2 fit often have to play 2NT. This was again a design decision. It would be a fairly simple matter to change this so that these sequences showed a doubleton spade and the 3 card limit raise made some other response (probably committing us to 3♠). At the moment I have not been convinced that the change is worth it but it is something I have on my review list.

All of the points you raise are good ones. The first is basically fixed but something I am actually quite happy with, particularly given that the 1NT structure (as well as 1♣ - 1♦; 1NT and 1♣ - 1♦; 1♥ - 1♠; 1NT) is built around 2♣ Puppet Stayman.

The second is an area where there might be some small gains to be had, for example it might turn out to better to respond 2♥ on some or all hands with 1543 shape. I am fine with exceptions providing I can can come up with a simple rule for covering them. On the other hand, doing this loses the chance to stop in 2♦ when it is right so it is not clear that it is actually better. There is no method I am aware of that can find absolutely every fit here and imho actually not too many that do a better job than this one.

The last point is pretty much one of pure maths. There are pretty clear gains and losses to taking the 3 card limit raise out of the relay. I know (from the BBF threads) that you have invested considerable effort in this area and consider it a major factor. At the moment my appraisal is lower, which is essentially why I have stuck by the original base. This part is less about the merits of the INV+ relay as about how best to use it though. I would assume that if you were to use it you would prefer 1♠ - 1NT; 2m - 2♠ to show an invite with 2 spades and use a different sequence, such as 1♠ - 3m or 1♠ - 3♥ or even 1♠ - 1NT; 2m - 3♠, for the 3 card LR. It would actually please me a lot if some pairs were to start using such alternative treatments - this is how systems grow and improve, something that I would love to see happen within my lifetime.