[Kungsgeten]

 straube, on 2017-February-26, 23:00, said:

Rev Flannery obviously works better if 1D includes balanced hands , but it seems to work here, too. It also solves lots of bidding problems, including leaving 1D-1S, X-2H as an artificial ask of some sort.

I have nothing against reverse flannery, I think its a good convention that do indeed solve bidding problems. I just feel that the majority of those problems aren't a problem in our structure. It would solve the 1-4-4-4 problem and make it easier when responder has 5-5 majors. We currently use the following response structure to 1D:

1M = Natural, F1. If GF then usually 5+ major.
1NT = 6+ clubs or 3-3-2-5, less than GF. Non-forcing but seldom passed.
2C = GF relay.
2D = Simple 3+ diamond raise.
2M = Invitational jump shift
2NT = Natural INV
3C = Diamond limit raise.
3D = Mixed raise
3M = Splinter
3N = 3-3-4-3, 13-15

In order to play Reverse Flannery we'd have to give up the invitational jump shift, or shuffle the responses somehow. We used to play:

1H = Nat or GF relay
1S = Nat F1
1N = Clubs, weak or GF
2C = "Bad" diamond raise or invitational jump shift in a major
2D = Constructive diamond raise
2M = Reverse Flannery
2NT = INV
3C = INV jump shift
3D = Mixed

When we changed our 1M structure so that 1M-2C was a GF relay we did the same for the diamond opening, to reduce memory load. The relays aren't quite as good anymore, but the bidding over 1D-1H is better (when responder has hearts).

Now with our current 1D-1S structure responder usually do not have a problem with spades and hearts, but you might end up in NT instead of a 4-3 / 5-2 major fit if responder has invitational values, or end up in a minor suit misfit instead of a major suit misfit.

1D-1S; 1NT (6+D or strong)---
2C = GF vs strong hand, opener rebids 2D with 6+D (about 11-14) and responder can then bid 2H "third suit GF" (have to bid 2NT or 3D with INV hand and both majors).
2D = Not enough to force game vs strong opener. Opener will bid 2H with the hearts reverse hand (same situation as 1D-1S; 2H, but it is known that responder is weak).
2H = Reverse Flannery hand, weak.
2S = Natural NF.

1D-1S; 2C (5-4 minors, 11-15/16, denies 4 hearts)
Pass = Preference.
2D = Preference.
2H = FSF.
2S = Weak.
2NT = INV.
3CD = INV.

1D-1S; 2D (4+D and 4H, 11-15. We've found an eventual heart fit.)

1D-1S; 2H (3S, 11-14 or 18-19. If we have a heart fit there's also a spade fit.)

[straube]

I like 1D-2S as WJS as well as 1D-2H as WRF because it reduces the load from 1D-1S. As for separating invitational hearts from WJS hearts, I can't always do that at the 2-level, but I certainly can after a 1D-1H, 1S start (I can also after a 1D-1H, 1N start). I have the feeling that 1D-2S as WJS and 1D-2H as WRF would work for you for the same reason it would work for most folks.

It's a pity that your 1D-1N can't be your GF relay. Seems like the relays would be entirely different from 1M-2C anyway so maybe not a memory load issue.

Could you make 1D-1M, 1N guarantee extras (16+?) or is it really needed as 2-way? If it were always strong I suppose responder could give a weak preference to either minor, rebid hearts weakly (1D-1H, 1N-2H) or bid something higher to GF. After 1D-1M, 1N-2m, opener could invite or correct (to 2D) or make some forcing call. It would be very "clean" if you could show the big hand immediately. Much better branching afterward.

I guess 1D-1S, 2D would be undesirable with x Axxx AQxxx xxx and you'd like to show more of your pattern, but with Rev Flannery at least you wouldn't be missing a heart fit. Plus 2H now is artificial GF and 2S would be invitational spades. Or you could rebid 2C with that hand which is what I do (with 14-15 anyway).

Btw, how do you handle 4M5D22 patterns? I assume 1N for one range. What about the other?

[awm]

It seems like you could do something where:

1d-1s...

1nt = 6+d
2c = natural no 4H
2d = 1444 or 0454 or 1453 min
2h = 4H extras

1d-1s-1nt-2c = asking, either 4h or any GF
... 2d = no 4H no 3s
... 2h = 4-6
... 2s = 3s-6d not 4H

You can get full relays over these if you want.

[paulg]

Before we switched to reverse flannery, we happily rebid 2!d on these hands showing 4!H, 4+!d and less than a reverse. I don't really see anything particularly bad that can happen with this.

[straube]

I like Adam's suggestion. One of the nice things about it is that the 1N rebid isn't 2-way.

If I remember right, Adam uses 1D-1S, 2D similarly and then 2H is a correction but the 2S bid asks further shape.

[Kungsgeten]

 straube, on 2017-March-02, 11:24, said:

I like 1D-2S as WJS as well as 1D-2H as WRF because it reduces the load from 1D-1S. As for separating invitational hearts from WJS hearts, I can't always do that at the 2-level, but I certainly can after a 1D-1H, 1S start (I can also after a 1D-1H, 1N start). I have the feeling that 1D-2S as WJS and 1D-2H as WRF would work for you for the same reason it would work for most folks.

It's a pity that your 1D-1N can't be your GF relay. Seems like the relays would be entirely different from 1M-2C anyway so maybe not a memory load issue.

Could you make 1D-1M, 1N guarantee extras (16+?) or is it really needed as 2-way? If it were always strong I suppose responder could give a weak preference to either minor, rebid hearts weakly (1D-1H, 1N-2H) or bid something higher to GF. After 1D-1M, 1N-2m, opener could invite or correct (to 2D) or make some forcing call. It would be very "clean" if you could show the big hand immediately. Much better branching afterward.

I guess 1D-1S, 2D would be undesirable with x Axxx AQxxx xxx and you'd like to show more of your pattern, but with Rev Flannery at least you wouldn't be missing a heart fit. Plus 2H now is artificial GF and 2S would be invitational spades. Or you could rebid 2C with that hand which is what I do (with 14-15 anyway).

Btw, how do you handle 4M5D22 patterns? I assume 1N for one range. What about the other?

We open 1D with 5D4M22. My partner isn't a fan of treating 6322 or 5422 as balanced.

1NT could be always strong if we give up separating 3-card raises from 4-card raises. Another common alternative is ofcourse to play transfer rebids by opener, even though my guess is that Gazzilli is better. Some also play 1D-1M; 2C as 6+D 11-14 or a strong hand, and responder bids the other major to force vs the hand with 6+D.

1NT response as GF relay is a reasonable alternative. I think Fredin-Lindqvist played it that way in their Swedish Club system. They used the immediate 2C response as non-forcing, which I guess could work but I'd like it to promise six then, and then the 3-3-2-5 pattern becomes problematic (right now we transfer to clubs and then bid 2D with that pattern). We do not play full relays anymore, since we've tried to streamline the systems to decrease memory load. The relay scheme is very similar for 1D/H/S-2C:

1D-2C;
2D = 4+ clubs (after 1M-2C it could also be any 5332)
2H = 4(+) hearts (we bid suits bottom up, so after 1M-2C this would show diamonds)
2S = 4(+) spades (other major after 1M-2C)
2NT = 7+ diamonds, single suiter, relay with 3C and then as below
3C = 6 diamond, minimum + shortness
3D = 6 diamonds, no shortness
3H = 6 diamonds, extras, short clubs
3S = 6 diamonds, extras, short hearts
3N = 6 diamonds, extras, short spades

The two-suiters resolve like this:

2S = 5332 after 1M-2C; 2D-2H. Clubs longer/better than diamonds after 1D-2C; 2D-2H.
2NT = Three-suiter (4441 or 5440)
3C = High shortness
3D = 5422
3H = 5431, low shortness
3S = At least 5-5, low shortness
3N = At least 6-4, low shortness

I know that there are much more effective ways to do this, but this is fairly easy to remember.

[Kungsgeten]

 paulg, on 2017-March-02, 13:20, said:

Before we switched to reverse flannery, we happily rebid 2!d on these hands showing 4!H, 4+!d and less than a reverse. I don't really see anything particularly bad that can happen with this.

Well I guess the bad thing is when you have a club fit and miss that, playing in an inferior red suit fit. The probability of responder having 4+ clubs is higher than responder having 4+ hearts, since the only time responder will have 4+ hearts is when also having 5+ spades. Having 4+ clubs responder may be 4-4, or even have 5+ clubs and 4 spades.

[nullve]

 Kungsgeten, on 2017-February-23, 14:23, said:

A problem with responding 1H with 4H and 5S I guess is when opener has three spades and less than four hearts.

It might not solve all problems with RFR-type hands, but you can rebid 1♠ with 3+ S (a la Nightmare) and still have ridiculous amounts of space. Should be part of standard IMO.

(I know you played the 1♠ rebid as 4+ C a couple of years ago.)

[straube]

 Kungsgeten, on 2017-March-03, 08:20, said:

We open 1D with 5D4M22. My partner isn't a fan of treating 6322 or 5422 as balanced.

In that case, Adam's scheme works even better for you. If instead 1D-1S, 2D just showed diamonds, you'd have a real problem with 2452 as well as 1453.

For your relays how about...

1D-1N
.....2C-clubs
.....2D-hearts
..........2S-5H/5D
..........etc-heart reverser
.....2H-spade reverser
.....2S-5S/5D
.....etc-6D

[Kungsgeten]

I'm considering adding the Reverse Flannery (RF) hands to our 1NT response:

1D--1NT = 6+C or 3-3-2-5, less than GF / RF less than GF

It might be a bit overloaded, but I think it should be manageable if fourth hand enters the auction. Responses like this:

1D--1NT;
2C = Non-forcing. Responder bids 2D with weak 3-3-2-5, 2M with RF weak/INV, 2NT with INV 3-3-2-5 and 3C with INV 6+C.
2D = Non-forcing, short clubs and good diamond suit. 2M is RF weak/INV.
2H = Artificial force
...2S = RF, weak and NF
...2NT = 3-3-2-5, minimum and NF
...3C = 6+C, minimum and NF
...3D = 6+C, establishes a GF
...3H = RF, establishes a GF
...3S = 5-5 majors, establishes a GF
...3NT = 3-3-2-5, not min

[Zelandakh]

Does 2♣ show a real suit? If so, what does Opener rebid with a minimum 4M5♦ hand without good diamonds?

[Kungsgeten]

 Zelandakh, on 2017-March-13, 03:58, said:

Does 2♣ show a real suit? If so, what does Opener rebid with a minimum 4M5♦ hand without good diamonds?

You mean in the sequence 1♦-1NT; 2♣? No, 2♣ does not promise anything in clubs; it could be a club void (4-4-5-0 for instance, or perhaps 6430 with bad diamonds). Responder will only pass 2♣ with a minimum hand (less than invite) and 6+♣. I guess its a situation similar to 1♦-2♣ in your system, except I guess responder can be 3-3-2-5 in yours? The 3-3-2-5 hand is somewhat awkward, since we'll have to play clubs at the three-level if that's our best fit (or opener can choose to play a 5-2 diamonds fit or a 4-3 major fit, even if there's an 8+ club fit, if he feels more comfortable with that compared to playing 3♣).

We're going to try putting Reverse Flannery hands into 1NT and see how it goes. Its actually pretty nice that opener's 2♦ rebid will be well defined. 1♦-1♠; 2♦ will show 5♦, 4♥, 0-2♠ and 2-3♣. That's also favourable since FSF (3♣) consumes a lot of space in this sequence. A downside with Reverse Flannery in 1NT is that we previously could super-accept the club transfer by bidding 3♣, but that probably isn't advisable when responder can have both majors.