[Kungsgeten]

I know several systems which use 1M-1NT as a GF relay (or INV+ relay in Zelandakh's case). I'm thinking that it might be possible to combine something like a forcing 1NT (usually looking for the right part score), with a game forcing relay? This would probably work best if using unbalanced major suit openings. I'm thinking something along these lines:

1S-1NT;
2C = 6+ spades. May be 6-4. Now 2D relays (maybe INV+).
2D = Natural. 2H and 2S are non-forcing. 2NT is the GF relay.
2H = Natural. 2S is non-forcing. 2NT is the GF relay.
2S = 4+ clubs. 2NT is the GF relay.
2NT+ = Not sure, some special hands.

Over a 1H opening we'd probably want to use Kaplan Inversion to get lower. Not sure if there's a plus in swapping the 2C and 2M rebid as above, but my guess is that some

Two-over-one could have some special meaning, perhaps transfers and certain invitational hands. Something like:

1S--
1NT = Forcing NT. Relay or "normal" negative NT type hand. Shapely weak hands usually bid something else.
2C = Sign off in a minor, or various invitational hands.
2D = 5+ hearts.
2H = Raise.
2S = Raise.
2NT = Raise.
3C = Natural INV.
Higher = Raises?

I guess it would be problematic if the opponents interfere, and there are probably other downsides too. I'm just curious if someone of you has experimented with this concept.

[awm]

Yeah I've looked at such systems before. There are a few basic issues:

1. There are a number of sequences where you either lose steps in the relay or lose the ability to correct the contract. Your use of 2NT as relay over two suited hands is pretty high for example (+2 compared to symmetric), and you're losing auctions like 1♠-1NT-2♣-2♥ in the natural sequence.
2. Pretty bad in competition, since relayer has neither shown a suit nor any values.
3. Opener's 3-level rebids get messed up. These are pretty useful opposite normal 1NT hands, but using them breaks the relays.

[straube]

On a recent thread, awm and I have been comparing our 1S methods to those of Zelandakh. I'm thinking of switching to awm's method. It does a really nice job of separating the weak/lightly invitational from strongly invitation from GF hands which makes the rebids natural and clear.

[Kungsgeten]

awm, on 2017-October-20, 14:14, said:

Yeah I've looked at such systems before. There are a few basic issues:

1. There are a number of sequences where you either lose steps in the relay or lose the ability to correct the contract. Your use of 2NT as relay over two suited hands is pretty high for example (+2 compared to symmetric), and you're losing auctions like 1♠-1NT-2♣-2♥ in the natural sequence.
2. Pretty bad in competition, since relayer has neither shown a suit nor any values.
3. Opener's 3-level rebids get messed up. These are pretty useful opposite normal 1NT hands, but using them breaks the relays.

Regarding 1, I think a possibility is to use the 2C rebid as 6+ major, with a possibility of being 6-4. Let's say that we decide that 2D should be GF over this:

2H = 4 card minor, or 6322. Then:
...2N = 4 clubs.
...3C = 6322
...3D = 6-4 diamonds, hi short
...3H = 7411, diamonds
...3S = 6421, diamonds, low short
...3N = 6430, diamonds, low short
...4C = 7420, diamonds, low short
2S = 7+ major + shortness
2N = 4oM.
3C = 6331, high short
3D = 6331, mid short
3H = 6331, low short
3S = 7222

In this case the two-suited table could be:

3C = High sho
3D = 5422
3H = 5431
3S = 5521
3N = 5530

A possibility could be that 1M-1NT; 2NT+ show 6-5 but perhaps that's a bad idea when having a minimum hand. Anyway offloading 6-4 from the normal two-suited table makes it somewhat more managable.

[nullve]

Kungsgeten, on 2017-October-20, 12:39, said:

I'm thinking that it might be possible to combine something like a forcing 1NT (usually looking for the right part score), with a game forcing relay? This would probably work best if using unbalanced major suit openings.

Based on what I play now:

1♠ = "10-21, 5+ S, unBAL"

1♠-1N; ?:

P = not allowed! (instead of "10-12", 5S3-H4+m, which I play now)
2♣ = "10-15", 4+ H OR "16-18", any
...2♦ = "8+", relay, "10+" if 1-S3H
......2♥ = "10-12" (=> 2N = GF relay)
......2♠ = "13-15" (=> 2N = GF relay)
......2N+ = "16-18", relay structure (see R(♠) below)
...(...)
2♦ = "10-15", 5S3-H4+m (instead of "13-15", 5S3-H4+m, which I play now)
...P = WK, usually 6+ D
...2♥ = INV+ relay (instead of NAT NF, which I play now)
......2♠ = "10-12"
.........P = < GF, 2 S (and probably not both minors)
.........2N = GF relay
............E.g.:
............3♣ = 5S4D
...............3♦ = relay
..................3♥+ = U(♠,♦) (see U(♠,y) below) w/ lots of idle bids
...............(...)
............3♦ = 5S5m
...............3♥ = relay
..................3♠ = 5 C (=> 4♣/♦ = slam try in C/S, respectively)
..................3N = 5 D (=> 4♣/♦ = slam try in D/S, respectively)
...............(...)
............3♥+ = U(♠,♣) (see U(♠,y) below) w/ lots of idle bids
..........3♣ = P/C
..........(...)
......2N+ = "13-15", relay structure (can even be R(♠) (see below) w/ lots of idle bids)
...2♠ = to play
...2N = "bid your minor"? (instead of GF relay (opposite "13-15"), which I play now)
...3♣ = WK, long C? (instead of P/C, which I play now)
...(...)
2♥ = "13-15", 6+S3-H (=> 2N = GF relay)
2♠ = "10-12", 6+S3-H (=> 2N = GF relay)
2N+ = "19-21", relay structure (see R(♠) below)

R(♠) (= a simplified version of what nullve-nullve play*):

2N = 5S4*C, 6+S4C or 1-suited
...3♣ = relay
......3♦ = 1-suited
......3♥+ = U(♠,♣)
...(...)
3♣ = 5S4*D or 6+S4D
...3♦ = relay
......3♥+ = U(♠,♦)
...(...)
3♦ = 5S5O*
...3♥ = relay
......3♠ = 5 C (=> 4♣/♦ = slam try in C/S, respectively)
......3N = 5 D (=> 4♣/♦ = slam try in D/S, respectively)
......4♣+ = 5 H
...(...)
3♥+ = U(♠,♥)

* The biggest difference is that my structure allows me to treat 53(50) as 53(41) initially, so that 3♦ can show 5S5H/5S2-H5m instead of 5S5O.

U(♠,y):

3♥ = 5S4y22 or 6S4y
...3♠ = relay
......3N = 5S4y22
......4♣+ = 6S4y
...(...)
3♠ = 5S4y13 (5S4y04)
3N = 5S4y31 (5S4y40)
4♣+ = more extreme 2-suiters

The relay structure has more in common with GAR than with Kerr-style symmetric relay, so I expect many regular posters here will hate it for that reason alone. But from a practical point of view I'm pretty sure Opener's 2♦ rebid is the worst part of this structure over 1♠-1N, although not obviously worse than, say, Opener's 2♠(="11-15", 5S3-H4+C) rebid over 1♠-1N; 2♣-2♦ in standard Gazzilli, pretending the comparison makes sense. Also, since the two-over-one responses to 1N are now partly freed up, they might be used also on some of the hand types that could cause problems over 1♠-1N; 2♦, such as 5-10, 14(53)/15(43).

[straube]

Deleted

[yunling]

My structure over a 11-20 1S:

1S-?
1NT=forcing, weak without much shape or any GF
2C=weak 6+D or invite w 4+H
2D=weak 6+H or invite w/o 4H
2H=weak 5H and another place to play(either 2S or 5+m)
2S=constructive

1S-1NT-?
2C=riton
……2D=GF relay, else=natural weak
2D/H=4+, natural limited
……2NT=GF relay
2S=4+C limited
……2NT=GF relay
3x=5-5 invitational


http://www.bridgebas...__1#entry861584

Relay at +2 is not that much a problem since that's I'll pay for splitting the range.
But the 2-level response gets somewhat mixed up opposite a maximum opener. This structure don't work well with limited openings since too many hands are thrown out of 1M(minimum balanced and 1-suited) and using 2C to split the range is not ideal.

[Phil]

In our Fantunes variant, 1M is as wide ranging of an opening imaginable. Thus, initially we wanted to take as much pressure off of a FNT (+KI) as possible so we lumped in invites instead into relays.

A typical auction might start

1♥ 1N ®
2♣ (spades)

If we'd break right there the call would natural and NF.

What we found was that a forcing NT, even a wide ranging one, worked just fine. Plus we play an elaborated Gazzilli treatment that lets us unwind these hands

Additionally, we wanted to steer clear of the non GF relay cc problems.