[nullve]

Suppose

1N-[2M-1]

is "transfer to M" and

1N-[2M-1]; 2M-[3m-1]

effectively a puppet to 3m. Then it makes sense to use the transfer also on weak hands with 4M6+m, expecting an auction like

1N-[2M-1]; 2M-[3m-1]; 3m-P,

1N-[2M-1]; [2M+1]*-[3M-1]**; 3M-P

or

1N-[2M-1]; [2M+1]*-[3M-1]**; 3M-4M; P,

since it's probably be better to play 3M than 3m if a 4-4 or 5-4 M fit exists.

But why not extend this idea (which I, and probably others, have mentioned before) and respond 2M-1 also on weak hands with 3M(!)3-OM6+m, since it's probably better to play 3M than 3m if a 5-3 M fit exists?

For example:

1N-[2M-1]; ?:

2M = 2-3 M
...(...)
...3m-1: now also on weak hands with 3M3-OM6+m***
...(...)
2M+1 = 4-5 M
...2N(M=♥) = relay
...3m = WK, 3M6+m****
......P = 4 M
......3M = 5 M
...(...)
...3M = WK, either 5+ M or 4M6+m*****
...(...)
...Alternatively, if M=♥:
...3m-1 = PUP 3m
...3♦ = PUP 3♥
...(...).

* superaccept promising 4-5 M
** retransfer
*** potentially wrongsiding 3♦ if M=♥ and m=♦
**** potentially wrongsiding 3m
***** potentially wrongsiding 3M

[hrothgar]

Like to see some discussion of competitive auctions

[hrothgar]

FWIW, the Scanian NT system uses transfers after their Stayman style bid which seems like it might be a better way to handle these hand types

[straube]

I think that both this and Inside Out Stayman are very bad ideas. You seem focused on finding better suit contracts for very weak hands and are sacrificing bids that have better use for finding game and slam.

[nullve]

straube, on 2019-May-19, 09:58, said:

I think that both this and Inside Out Stayman are very bad ideas. You seem focused on finding better suit contracts for very weak hands and are sacrificing bids that have better use for finding game and slam.

You seem to like IMprecision's NT structure a lot. The version I have in front of me seems to contain the following substructure:

1N-[2M-1]; 2M-?:

(...)
2N = GF, 4+M4+C, unBAL
...3♣ = mandatory relay
3♣ = GF, 5+M4+D, short minor-module
...3♦ = mandatory relay
(...).

So it seems that

1N-[2M-1]; 2M-[3m-1]

is indeed a puppet to 3m in that structure. So at least in this instance it's not clear that I'm sacrificing anything. But maybe you thought I meant to use those PUP 3m bids for weak hands only?

[straube]

IMprecision transfers promise either 5 of the major or 4M/6m GF. Some small fraction of the time opener may compete into a 4-3 major suit fit but at least this is opposite a hand that intends to force game. The large majority of the time, responder will have 5 of the major.

The relay bids allow responder to describe various hand patterns and opener cooperates (doesn't break relay to show fit) to preserve room for responder to show shortness, etc.

I don't see the IMprecision method as at all similar.

[nullve]

straube, on 2019-May-19, 11:18, said:

IMprecision transfers promise either 5 of the major or 4M/6m GF. Some small fraction of the time opener may compete into a 4-3 major suit fit but at least this is opposite a hand that intends to force game. The large majority of the time, responder will have 5 of the major.

The relay bids allow responder to describe various hand patterns and opener cooperates (doesn't break relay to show fit) to preserve room for responder to show shortness, etc.

I don't see the IMprecision method as at all similar.

One of us has a blind spot.

Suppose

1N-[2M-1]; 2M-?:

(...)
2N = WK, 3-4M6+C OR as in IMprecision
...3♣ = forced
......P = WK, 3-4M6+m
......3♦+: as in IMprecision
3♣ = WK, 3-4M6+D OR as in IMpresision
...3♦ = forced
......P = WK, 3-4M6+D
......3♥+: as in IMprecision
(...).

Do you still think my idea is incompatible with IMprecision?

[straube]

Yes.

As an almost aside, IMprecision doesn't require opener to super-accept with 4-cd support. I want to keep that. So your scheme would work maybe half the time, getting to the major with a super-accept but subsiding in the minor when not super-accepting. Still some wins, right? However, if I were to use 2M+1 as the super-accept (a good use), I'd want continuations more constructive than sorting out whether we had a 4-3 fit or not. These could be HSGTs or retransfers or whatever.

The main issue is you're proposing to transfer with a 3-cd major in hopes of finding a 5-3 major suit fit as opposed to a 6-? minor suit fit. This means that in a competitive auction, opener will frequently want to raise with even 3-cd support but will be reluctant to wind up in a 3-3 fit opposite a weak hand.

If your proposal was to find major suit fits with weak 4M/6m hands and you required opener to super-accept or break relay to agree the major then I'd have to think about it.

Since you brought up IMprecision maybe you should ask Adam what he thinks of your idea in an IMprecision structure.

[Kungsgeten]

I think this could work if you don't have other alternatives showing these weak hands. Apart from complexity the downside doesn't seem to be that big, compared to not using the second round transfer as a non-puppet.

We play transfers after Stayman instead, but not exactly like the Scania system. A plus point for this treatment is that you can stop in 2M on a 4-4 or 4-3 fit, and that you also might "take-out" with some 5m-4M patterns.

The way we play our transfers requires that 1NT-2D is bid with invitational hands with 4 hearts. Also we play 1NT-2H; 2S-2NT as "standard", so invitational with 5S, instead of being a transfer. Unfortunately we can't use 1NT-2C; 2red-2S as invitational with 5S.

1NT-2C; 2D---
2H = We play this as non-forcing with both majors. Could be an invite. This is the same way as Scania does it.
2S = Range-ask or transfer to clubs. Could be INV with 4S, or weak with 4M and 5+C, or GF with 4M and 5+C (also 4-4-1-4).
2NT = GF relay style. Asks for more information about shape.
3C = GF transfer to diamonds (since we could pass with a weak hand). 4M and 5+D (also 4-4-4-1 if GF).
3D = Transfer to hearts. Could be 5-5 majors.
3H = Transfer to spades (Smolen). Could be 5440.
3S = Smolen. Could be 5440.
3NT = To play.


1NT-2C; 2H---
2S = Range-ask or transfer to clubs. As above. Always with 4S.
2NT = GF relay style.
3C = Transfer to diamonds, with 4S. Weak or GF.
3D = GF with heart support and shortness somewhere (Baze-style, but can clarify void or singleton).
3H = Invite.
3S = 4 hearts but suitable for 3NT. Choice of games.
3NT = To play.
4C = Slam invite with 4H, no suitable shortness (Baze-style).
4D = RKCB with 4H, no suitable shortness (Baze-style).


1NT-2C; 2S---
2NT = Puppet to 3C. Sign-off in a minor (with 4H) or GF with 4H and 5+C.
3C = GF relay style.
3D = GF with 4H and 5+D.
3H = Baze-style.
3S = Invite.
3NT = To play.
4m = Baze-style.

[awm]

In IMPrecision notrump systems, we don’t require super accept on four card support, and we also don’t have mandatory relays (usually opener breaks the relay with a known major suit fit unless the minor suit fit is bigger).

I’m not convinced that a 5-3 fit in a major even necessarily plays better than the 8+ card minor suit fit, so I don’t really see the need to find 5-3 major fit on these hands. It’s not like you are upgrading a non invite to game here. 4-4 fits are different because you can potentially set up the minor (with four trumps in the same hand); I think Qxxx x Axxxxx xx might well make game opposite a strong notrump with four+ spades but that Qxx x Axxxxx xxx is much less likely to make game opposite a strong notrump with five spades.

Overall, openers five card major is quite rare. Catering to this will cause you a lot of problems (slow auctions where opponents can double for the lead, information about shape that helps them defend, ambiguous sequences in competition where it’s hard for opener to know whether to bid on) for very little value. I think if you’re that concerned about playing in the major every time you have a 5-3 fit, you should open 1M and not 1nt and then play some
Gazzilli form to solve the rebid problems.

[nullve]

straube, on 2019-May-19, 13:11, said:

Yes.

As an almost aside, IMprecision doesn't require opener to super-accept with 4-cd support. I want to keep that. So your scheme would work maybe half the time, getting to the major with a super-accept but subsiding in the minor when not super-accepting. Still some wins, right?

Although Opener had to superaccept with 4-5 M in my example structure, my favourite super-accept scheme in a more standard NT structure is actually

1N-[2M-1]; ?:

2M = 2-3 M or 4M333
2M+1 = 4-5 M but not 4M333

because

* 4M333 behaves more like 3c M support wrt LoTT;
* having only one super-accept means less information leakage
* there is still enough room to relay out Opener's shape after 1N-2♦; 2♥/♠.

Of course, not super-accepting on all hands with 4 M means sometimes missing a 4-4 M fit after

1N-[2M-1]; 2M-[3m-1]; 3m-P,

but then it will also be missed by IMPrecision and some standard systems after

1N-[3m-1]; 3m-P,

right?

I actually had a look at IMprecision's INT structure before starting this thread, precisely because I thought it might be an example of a structure where Opener's one-step relay over 1N-[2M-1]; 2M-[3m-1] is mandatory. If it isn't, and I accept that awm is the final arbiter of that, it doesn't really matter, because there are other possible structures that are, and I thought you were making the more general point that my idea was incompatible with relaying Responder's shape as effectively as in IMprecision.

straube, on 2019-May-19, 13:11, said:

However, if I were to use 2M+1 as the super-accept (a good use), I'd want continuations more constructive than sorting out whether we had a 4-3 fit or not. These could be HSGTs or retransfers or whatever.

I see what you mean, although it's possible to play

1N-2♦; 2♠-?:

(...)
3m-1 = PUP 3m
3♦ = PUP 3♥ (aka retransfer to H)
3♥ = shape relay (with H agreed)
(...)

and

1N-2♥; 2N-?:

3m = WK, 3M3-OM6+m
3♥ = PUP 3♠ (aka retransfer to S)
3♠ = shape relay (with S agreed)
(...).

straube, on 2019-May-19, 13:11, said:

The main issue is you're proposing to transfer with a 3-cd major in hopes of finding a 5-3 major suit fit as opposed to a 6-? minor suit fit. This means that in a competitive auction, opener will frequently want to raise with even 3-cd support but will be reluctant to wind up in a 3-3 fit opposite a weak hand.

I'm not sure what competitive auctions you have in mind, but

1N-[2M-1]-(3m)

is probably one of them.

I care about LoTT, so I'd never raise to 3M with only 3c support here even if 2M-1 were a standard Jacoby transfer promising 5 M. But I might double ("takeout") with 3M2m, leaving the final LoTT decision to partner.

straube, on 2019-May-19, 13:11, said:

If your proposal was to find major suit fits with weak 4M/6m hands and you required opener to super-accept or break relay to agree the major then I'd have to think about it.

Just allow

1N-[2M-1]

and

1N-[2M-1]; 2M-[3m-1]; 3m-P

on weak hands with 4M6+m. That's it, I believe.

[straube]

nullve, on 2019-May-20, 16:31, said:

Of course, not super-accepting on all hands with 4 M means sometimes missing a 4-4 M fit after

1N-[2M-1]; 2M-[3m-1]; 3m-P,

but then it will also be missed by IMPrecision and some standard systems after

1N-[3m-1]; 3m-P,

right?

That's what I said.

nullve, on 2019-May-20, 16:31, said:

I'm not sure what competitive auctions you have in mind, but

1N-[2M-1]-(3m)

is probably one of them.

I care about LoTT, so I'd never raise to 3M with only 3c support here even if 2M-1 were a standard Jacoby transfer promising 5 M. But I might double ("takeout") with 3M2m, leaving the final LoTT decision to partner.

It would have been helpful if you had stated your practice at the beginning. I haven't seen a poll, but I think most keep a penalty double here and compete with suitable 3-cd support.

nullve, on 2019-May-20, 16:31, said:

Just allow

1N-[2M-1]

and

1N-[2M-1]; 2M-[3m-1]; 3m-P

on weak hands with 4M6+m. That's it, I believe.

I don't like the tradeoffs.

After 1N-2C, 2M my 3C rebid shows 4OM and asks for a 3D bid or show of fit for the OM. This btw isn't IMprecision but is something awm designed.
This lets me (with 4M/6D) play fits in such contracts as 2D, 2M, 3D and (rarely) 3M.

True I can't find a major suit fit with a weak 4M/6C. I'd like to be able to do so, but at the risk of sounding sour grapes, a 3C contract is often the best spot here and often better than 3M when we have that fit.

I don't think transferring into 3-cd suits is a good method. It doesn't fit my NT structure. You can have the last word. Good luck.

[awm]

The four card major version of this actually seems pretty good to me, however. One could easily play:

1NT - 2♦/2♥ (transfer, includes some 4M-6m hands)
2M - 2NT (clubs)

... 3♣ = at most three in the major, can be passed by a suitable 4M-6♣ hand
... 3M = four-card support for major, but minimum (else super-accept) and not great fitting cards for the minor
... 3OM = four-card support for major, good fitting cards for the minor also

Continuations after 3M/3OM assume the major as trump.

The idea would be to add 4M-6m hands that are below traditional invite values but not total trash, such that you can upgrade to game when a four-card fit is found (at least when there are also fitting cards for the minor). I'm not so much worried about getting to 3♠ when responder has Jxxx x xx Jxxxxx and opener has four spades (you're probably not making here regardless and 3m may go down less anyway).

This doesn't cost so much in competition where opener normally competes after a transfer only with four-card support (I haven't seen many good results from 1NT-Pass-transfer-3x-3M with three card support TBH).

[straube]

awm, on 2019-May-21, 01:53, said:

The idea would be to add 4M-6m hands that are below traditional invite values but not total trash, such that you can upgrade to game when a four-card fit is found (at least when there are also fitting cards for the minor). I'm not so much worried about getting to 3♠ when responder has Jxxx x xx Jxxxxx and opener has four spades (you're probably not making here regardless and 3m may go down less anyway).

That's interesting. I think I will do this for lightly invitational 4M/6m. In that case, thank you nullve and awm.