[helene_t]

Suppose you have decided to play a natural system with a 12-14 NT and that
1♣-1♠
shows spades (probably Walsh style as discussed in the recent thread about Walsh and weak NT, but it doesn't matter so much for this discussion I think). The same issues which I am concerned about also apply to
1♣-1♥ and
1♦-1♠
albeit to a lesser extent.

Suppose opener holds
x-AKxx-Axx-Jxxxx
or
Kxx-KQxx-x-Kxxxx

and it starts
1♣-1♠
?

If 1NT now shows 15-17 you can't bid that. You might bid 2♠ with the second hand but since you would also make that rebid with a balanced 15-16 with 4-card support, partner will invite quite aggressively which isn't what you want, especially if they have only four spades.

If you rebid 2♣ with the first hand (and maybe also the second one), the system is not very well designed:
- The least space consuming rebid of 1NT is more well defined than 2♣.
- You bypass 1NT on the weak, potentially misfitting hands which are the ones where you are most likely to want to stop in 1NT.

So what I have been toying with is:
The 1NT rebid is either
- 15-17 bal with two spades
- 11-16(17?) with 5(4) clubs and shortness in spades
- 11-14 with 3-card support
- Possibly also 12-15 with 6♣4♥ but that is not essential
I assume that with 12-14, 2(24)5, a 1NT opening is mandatory so the 2♣ opening now shows 6+ clubs and denies three spades.

How responder now proceeds with 6+ spades depends on what a direct 2♠ response would mean so I will ignore that for the moment. With other hands, responder proceeds:
5-7 any: pass
8-10: 2♣/♦/♥, i.e. to play opposite 11-14 short in spades.
11-12: 3♣/♦/♥, i.e. invitational opposite 11-14 short in spades
13+: 2NT

Q1: I was wondering if 2NT should be any invite so that 3x becomes GF, or maybe 2NT and upwards as invitational+ transfers.

Q2: What should the jump shifts mean? Probably some kind of strong jump shifts are useful as bidding space will be cramped after responder's 2NT GF at second turn. But also we need some weak or semipositive 2♠ response to tighten the range of responder's 2♠ second bid

Q3: I understand that something like this used to be popular in Acol but has gone out of fashion. Maybe because it's a silly idea?

[DavidKok]

Having multiple cheap weak bids and jumping with strong hands is backwards. Checkback/NMF/XYNT is more efficient and should accomplish much the same. It is going to be relatively rare that you want to play 2♣ anyway - if partner has the balanced hand on 1♣-1♠; 1NT you have extra values, while if partner has the unbalanced hand there is a big club fit (in fact, that's one of the selling points of playing a weak notrump). I think copying existing structures goes a long step towards answering your first two questions.
As for the third one: I don't really believe in the idea. Responder will have difficulty evaluating their hand with 5(+) spades. With exactly 5 spades and a weak hand do you pass or go to 2♠? If partner has the big balanced, or even unbalanced with a doubleton support, bidding will often be better. If you include a host of singleton hands, not so much. In fact, this is a general weakness of the Walsh approach - with a weak hand, a 5cM and possibly a 4c♦, on 1♣-1M; 1NT you have to choose between 2M or NT, there's no way to bid the diamonds and let partner give preference (auctions showing diamonds at the 2-level are reserved for hands with 4M5(+)♦). The standard advice is to bid with a 5-card suit. If 1NT contains both 15-17 with doubleton (or even three-card) support and 11-14 with a singleton you practically have to give up on this. Responder will struggle in a similar way with a shapely hand with a 6-card major. That hand type should upgrade aggressively opposite the strong notrump, but has terrible prospects opposite unbalanced hands with a singleton in support.
Personally I prefer rebidding 2♣, even on a wimpy 5-card suit, rather than rebidding 1NT with a singleton (even if it is in range). The invitational(+) hands opposite minimum openings have plenty of space to recover, and the weak ones can now make an informed decision between 2♣ and 2M.

Taking a step back for a moment, the traditional claim of weak NT systems is that 1m-openings promise some extras, either shape or values. You can lump the unbalanced minimum raises in with the strong notrump raises of partner's major because these hands are supposed to not differ that much in playing strength. One of the downsides of this approach is that, compared to a strong notrump system, you are taking on more risk by opening light shapely hands. If partner doesn't have a good fit for your suits you will not have the promised extra playing strength that the system relies on. Your example hand ♠Kxx, ♥KQxx, ♦x, ♣Kxxxx illustrates the point well - even with spades opposite this hand never evaluates to more than a minimum opening. From best to worst I think your options are:

• Stop opening hands like this. Your 1m openings systemically promise some extras, and you don't have them. Keep in mind you can get burned not just on constructive auctions but also competitive ones - say, 1♣-(2♥)-X, where partner feels safe doubling a little light because you cannot have a weak notrump. Do you pass or run to 3♣?
  
• Open these hands anyway, rebid 2♠ on your example auction, have partner invite as if you have a 15-17 balanced most of the time, and get to failing games slightly more often. The hand type is rare and I wouldn't cater to it too much.
  
• Make minor adjustments to your system - shift the strong NT range to 15-16 (jumping with 17-19 rather than 18-19), or agree to not respond light to 1m, or accept that you might sometimes play in 2M while the field is in a good 4M game. In my experience it was common to upgrade balanced 17s and even nice 16s with four card support anyway, so that the simple raise showed something more like "unbalanced minimum raise, can be a three-card suit OR balanced four-card raise, 15-16^-".
  
• Make a major adjustment to your system of the sort you propose. This means overhauling most of your notrump ladder and its responses, possibly also changing your game try bids. Alternatively you could play transfers over 1♣ or some other artificial system.

[nullve]

Idea:

1♣-1♠: now with fewer than 4 hearts unless GF strength, and fewer than 6 spades unless GF (alternatively, INV+) strength opposite 15-17 BAL

1♣-1♠; 2♣ = 11-13, 0445/1345/1435 OR 12-14, 1444 [similar to Precision's 2♦ opening = "11-15, 4415 minus 1 card"]

and based on what I play over 1♣-1♥ in a T-Walsh context:

1♣-1♠; 1N = a) "11-13", 6+C2-S b) either "14-16", 5+C2-S, unBAL or 15-17, 1444 c) 15-17 BAL, 2-3 S

1♣-1♠; 1N-?:

P: would have passed if 1N had shown c) only
2♣ = to play opposite a), usually 5 S unless GF strength opposite c)
...P = a)
...2♦ = b), 4(+)D3-H* (=> 2N = ART GF; other = NAT NF)
...2♥ = b), 4(+) H* (=> 2N = ART GF; other = NAT NF)
...2♠ = c)*
...2N = b), 6 C, 1-suited* (=> e.g. 3♣ = NF; other = NAT GF)
...3♣ = b), 7+ C, 1-suited (=> e.g. 3♦+ = NAT GF)
2♦ = GF relay
...2♥ = c)
...2♠ = a)
...2N+ = b)
2♥** = 6+ S, GF (or, if you prefer, INV+) opposite 15-17 BAL
2♠ = 6+ S, IJS-strength
2N+: less important!

If 1♣-1♠; 1N is to include

d) 11-14, 3 S

as well, then maybe

1♣-1♠; 1N-2♣; 2♦ = as above OR d)
1♣-1♠; 1N-2♣; 2♦-2♥ = range ask opposite d) (then e.g. 2♠ = d), MIN; 2N = d), MAX; 3♣+ = b), NAT)
1♣-1♠; 1N-2♦; 2♠ = a) OR d), 6+C3S, 1-suited
1♣-1♠; 1N-2♦; 2N/3♣/♦ = b) [following my own system]
1♣-1♠; 1N-2♦; 3♥+ = d), 5+C4R3S

.

* I would treat 14-16 hcp and either 2(42)5 or 2(32)6 as 15-17 BAL, though.
** Would have rightsided spade contracts opposite the strong BAL hand in my T-Walsh system, but here they are already wrongsided.

.

[helene_t]

 DavidKok, on 2023-February-27, 10:40, said:

Having multiple cheap weak bids and jumping with strong hands is backwards. Checkback/NMF/XYNT is more efficient and should accomplish much the same.

Maybe the 8-10 hands that want to play 2♣ opposite the 1(43)5 11-14 hands should bid 3♣ so that 2♣ is available for XYZ or something similar.

The problem is that
1♣-1♠
1NT-2♣*
isn't really a contract improvement, it is rather a courtesy bid, keeping the auction alive in case opener has 15-17 bal. So typically 5(332). So on a bad day we could be in a 5-2 fit. But mostly it will be 5-3 or 5-4, so probably I shouldn't worry too much about it.

As for
1♣-1♠
1NT-2♦*
being ambiguous about which of responder's two suits that is longer, my idea was not to worry about it, opener shows their 3-card spades support. I suppose they could decide to pass with a 3145 11-count, especially at IMPs. We will often play 2♠ in a 4-3 fit but we would do that anyway if we instead raised spades directly.

[helene_t]

 nullve, on 2023-February-27, 11:01, said:

Idea:

1♣-1♠: now with fewer than 4 hearts unless GF strength, and fewer than 6 spades unless GF (alternatively, INV+) strength opposite 15-17 BAL

1♣-1♠; 2♣ = 11-13, 0445/1345/1435 OR 12-14, 1444 [similar to Precision's 2♦ opening = "11-15, 4415 minus 1 card"]

and based on what I play over 1♣-1♥ in a T-Walsh context:

1♣-1♠; 1N = a) "11-13", 6+C2-S b) either "14-16", 5+C2-S, unBAL or 15-17, 1444 c) 15-17 BAL, 2-3 S

Thanks, I think I like that better than my structure. By making 6+ clubs the only weak option in the 1NT rebid, we don't need all those nonforcing 2-level rebids by responder.

[DavidKok]

I'm worried about the 4-2 fit, not the 4-3 fit. It is standard to give preference to responder's major with a doubleton opposite 5M4m. In a no-gadgets Walsh structure you are stuck with, say, 4♠5♦ on 1♣-1♠; 1NT. Do you bid 2♦ and hear partner give preference to spades, may be a doubleton? Do you bid 2♦ and forbid partner from correcting to spades on a doubleton, losing when you have 5♠4♦? Something else?

The standard Walsh solution is to play:
2♦: 4 spades, 5(+) diamonds, do not correct.
2♥: 5(+) spades, 4(+) hearts.
2♠: 5(+) spades, weak. Often has a 4cm.

The spades rebid is a really bad idea opposite a possible singleton, and you would need to invent a new rebid structure to make this work.

I don't like nullve's structure that much, although I'm problably misunderstanding it. On the auction 1♣-1♠; 1NT-? does responder have a way to show hearts without risking getting dropped in 2♣, say with 5=4=3=1, 5=5=3=0 or 6=4=3=0 and 8-11 HCP? Presumably any hand with long spades and weaker than an IJS makes a weak jump shift on the first round - does that include all two-suiters? Hands with secondary hearts are worth a lot more opposite a fit compared to a misfit, and hands with a fifth or sixth spade are worth a lot more opposite 3- or 2-card support compared to shortage. How can you tell whether your hand is weak or invitational with 2♣ NF? And how do the 4♠5(+)♦ hands bid, holding 9-11 HCP or so? Do you have a way for opener and responder to recover possible heart and spade fits on 1♣-1♠; 1NT-2♣; 2♠, where responder may have 4-5 spades, opener has 2-3, and neither side has said anything about hearts (if I understand correctly)?

[helene_t]

 DavidKok, on 2023-February-27, 11:40, said:

I'm worried about the 4-2 fit, not the 4-3 fit. It is standard to give preference to responder's major with a doubleton opposite 5M4m.

Yes that's why you need to distinguish 4-5 from 5-4 in a normal Walsh system, but this system is different.

Opener won't take preference on a doubleton spades. With a doubleton spades they would have 15-17 which is enough to bid 2NT or higher opposite 8-10.

1♣-1♠
1NT-2♦
2♠*
would be something like a 3415 or 3136. Probably also most 3145, so sometimes we will miss a good diamond fit. But at least we will have 7 trumps.

[nullve]

 DavidKok, on 2023-February-27, 11:40, said:

I don't like nullve's structure that much, although I'm problably misunderstanding it.

I believe you missed this:

 nullve, on 2023-February-27, 11:01, said:

1♣-1♠: now with fewer than 4 hearts unless GF strength, and fewer than 6 spades unless GF (alternatively, INV+) strength opposite 15-17 BAL

[mikeh]

 helene_t, on 2023-February-27, 07:54, said:

Suppose you have decided to play a natural system with a 12-14 NT and that
1♣-1♠
shows spades (probably Walsh style as discussed in the recent thread about Walsh and weak NT, but it doesn't matter so much for this discussion I think). The same issues which I am concerned about also apply to
1♣-1♥ and
1♦-1♠
albeit to a lesser extent.

Suppose opener holds
x-AKxx-Axx-Jxxxx
or
Kxx-KQxx-x-Kxxxx

and it starts
1♣-1♠
?

If 1NT now shows 15-17 you can't bid that. You might bid 2♠ with the second hand but since you would also make that rebid with a balanced 15-16 with 4-card support, partner will invite quite aggressively which isn't what you want, especially if they have only four spades.

If you rebid 2♣ with the first hand (and maybe also the second one), the system is not very well designed:
- The least space consuming rebid of 1NT is more well defined than 2♣.
- You bypass 1NT on the weak, potentially misfitting hands which are the ones where you are most likely to want to stop in 1NT.

So what I have been toying with is:
The 1NT rebid is either
- 15-17 bal with two spades
- 11-16(17?) with 5(4) clubs and shortness in spades
- 11-14 with 3-card support
- Possibly also 12-15 with 6♣4♥ but that is not essential
I assume that with 12-14, 2(24)5, a 1NT opening is mandatory so the 2♣ opening now shows 6+ clubs and denies three spades.

How responder now proceeds with 6+ spades depends on what a direct 2♠ response would mean so I will ignore that for the moment. With other hands, responder proceeds:
5-7 any: pass
8-10: 2♣/♦/♥, i.e. to play opposite 11-14 short in spades.
11-12: 3♣/♦/♥, i.e. invitational opposite 11-14 short in spades
13+: 2NT

Q1: I was wondering if 2NT should be any invite so that 3x becomes GF, or maybe 2NT and upwards as invitational+ transfers.

Q2: What should the jump shifts mean? Probably some kind of strong jump shifts are useful as bidding space will be cramped after responder's 2NT GF at second turn. But also we need some weak or semipositive 2♠ response to tighten the range of responder's 2♠ second bid

Q3: I understand that something like this used to be popular in Acol but has gone out of fashion. Maybe because it's a silly idea?

1=4=3=5 with weak clubs has always been a difficult hand to bid in a naturally based method with a 1S response to 1C.

I’m not sure there is any perfect solution. In my experience these hands, and the uncontested start of 1C 1S are rare to the point that I think it best just to suck it up and rebid 2C, accepting the very occasional disaster in 2C.

To me, any attempt to design around this is going to be either or both of complex and inefficient….inefficient in that it stresses, and renders less effective, those sequences you now use to allow this hand to be sorted out.

For example, a structure that has multiple and significantly different meanings for a 1N rebid will inevitably lead to problems.

Consider the straightforward 1N rebid, with 2-3 spades and a balanced minimum…the prototypical hand described in all beginner texts.

Responder can and often will place the contract…3N with 4 spades and presumably 4333 or 4324/4234, or 4S with basically any non slammish game going 6+ spade suit. A short, relatively uninformative auction to what is usually the right contract.

If one has to cater to a variety of hand types, ranging from a bad 11 with short spades to 15-17, then responder can’t place the contract. He has to bid to allow opener to describe. The more bids one makes, the more one is helping the opponents, who will be listening. And that ignores the very real risk of one partner or the other having a ‘forget’.

To me, xyz cures the great majority of difficult sequences and is pretty simple.

Btw, here’s a nice and easy adjustment to xyz so long as 1C 1M 1N shows a weak notrump hand (it doesn’t work when the 1N rebid is strong). Play 2C as forcing 2D, to play or to show some invitational hand, but use 2D to promise 5 spades and 4+ diamonds.

Opener has rights over this 2D…he goes back to spades (it works equally well for hearts after 1C 1H 1N) with 3. No more worrying about an inadequate major suit fit.

Transfer Walsh also minimizes some of these problems, especially if you play a style in which opener ‘accepting’ the transfer shows 2-3 card support. But learning T-Walsh is, itself, somewhat burdensome, especially since there are a lot of variations. It’s not as bad as defining 2/1 GF, but I’d never play T-Walsh with a new partner, no matter how expert, without at least fifteen minutes of discussion.

As for nullve’s ‘solution’, it seems to involve a whole raft of other system kludges. If 1C 1S shows fewer than 4 hearts unless gf, I infer that one responds 1H with 5Scand 4H and less than gf values, as just one obvious issue. I’m sure he has a way of working all of these things out, but at what cost to ease of use, limited information leakage, and so on. Not to mention that imo most of these kludge methods have been designed for use against Trappist monks: opponents who are sworn never to bid, let along preempt

1C P 1H 3D

Good luck when responder has a modest 5=4 major hand and opener has 4S, no heart fit (or even a heart fit) and we belong in spades. I’m sure nullve will explain how he never gets this wrong😀

All of this to avoid a low frequency issue….opener has rebid 2C on some 1=4=3=5 Jxxxx suit, responder is too weak to bid over 2C yet is short in clubs, and neither opp has bid nor will they bid. Yes, it happens. But it doesn’t happen often and it’s not always a disaster (and virtually never a big disaster) when it does.

Final btw, for the 1=4=3=5 hands, one can assure finding hearts if one plays that responder bids 2D or 2H (whichever you choose to designate) with 5 spades and 4+ hearts if too weak to bid over 1C 1S 2C. I don’t…I use Meckwell on hands that top out a little below that range, so I can still miss a 4=4 or even 4=5 heart fit, but exceedingly rarely.

[nullve]

 mikeh, on 2023-February-27, 16:24, said:

As for nullve’s ‘solution’, it seems to involve a whole raft of other system kludges. If 1C 1S shows fewer than 4 hearts unless gf, I infer that one responds 1H with 5Scand 4H and less than gf values, as just one obvious issue. I’m sure he has a way of working all of these things out, but at what cost to ease of use, limited information leakage, and so on. Not to mention that imo most of these kludge methods have been designed for use against Trappist monks: opponents who are sworn never to bid, let along preempt

1C P 1H 3D

Good luck when responder has a modest 5=4 major hand and opener has 4S, no heart fit (or even a heart fit) and we belong in spades. I’m sure nullve will explain how he never gets this wrong😀

All of this to avoid a low frequency issue….opener has rebid 2C on some 1=4=3=5 Jxxxx suit, responder is too weak to bid over 2C yet is short in clubs, and neither opp has bid nor will they bid. Yes, it happens. But it doesn’t happen often and it’s not always a disaster (and virtually never a big disaster) when it does.

Final btw, for the 1=4=3=5 hands, one can assure finding hearts if one plays that responder bids 2D or 2H (whichever you choose to designate) with 5 spades and 4+ hearts if too weak to bid over 1C 1S 2C. I don’t…I use Meckwell on hands that top out a little below that range, so I can still miss a 4=4 or even 4=5 heart fit, but exceedingly rarely.

1♠ denying 4+ hearts unless GF is consistent with the use of Reverse Flannery by Responder. Just like in the Meckwell system over a nebulous 1♦ opening, I suppose.

[mikeh]

 nullve, on 2023-February-27, 17:37, said:

1♠ denying 4+ hearts unless GF is consistent with the use of Reverse Flannery by Responder. Just like in the Meckwell system over a nebulous 1♦ opening, I suppose.

You may know players who use reverse Flannery with invitational hands. Where I’ve played, and read about it, it always showed a ‘one bid’ hand: the point being to minimize the problems after 1D 1S 2D or, worse, 1D 1S 2C. I think it could be played as invitational, on the upper end, but I think it significantly reduces the efficiency of the bid if it becomes wide range

For example, if one uses 1M 2H as the bid, opener usually passes with a decent minimum and a fit, but would have to raise if responder could be near gf values…resulting in voluntarily reaching some very poor 3 level contracts when responder has the 5 count that most would include in the range.

[akwoo]

I've known people to have 2 different reverse flannery bids for different ranges.

[nullve]

Some pairs playing two RFRs (one invitational, one not) in the 2022 Bermuda Bowl:

Greco-Hampson (USA2)
Hurd-Bathurst (USA1)
Levin-Weinstein (USA1)
Piedra-Zimmermann (Switzerland)

[DavidKok]

For the brief time I played Precision we played 1♦-2♥ as exactly invitational Flannery and 1♦-2♠ as GF Flannery. You seem to have the GF range covered and presumably reserve 1♣-2♦ and 1♣-2♥ for invitational and weak Flannery respectively?
With the both-major hands mostly sorted the system looks strong, although mikeh's comment about interference does have me worried, as does leaking information on exactly game hands. I stand by what I said earlier - XYNT is relatively simple, very effective (more effective on some hands) and solves most of the issues.

[bluenikki]

 mikeh, on 2023-February-27, 16:24, said:

1=4=3=5 with weak clubs has always been a difficult hand to bid in a naturally based method with a 1S response to 1C.

I’m not sure there is any perfect solution. In my experience these hands, and the uncontested start of 1C 1S are rare to the point that I think it best just to suck it up and rebid 2C, accepting the very occasional disaster in 2C.

In Kaplan-Sheinwold, the simple rebid of the minor is defined as either a terrible suit or a terrible hand. So although you may be stuck in a 5-1 fit at the two-level, you won't get into big trouble from partner bidding too high.

(Of course, a minimum KS minor opening is always sound in quick tricks, so terrible means no extra quacks.)

[nullve]

 DavidKok, on 2023-February-28, 07:27, said:

You seem to have the GF range covered and presumably reserve 1♣-2♦ and 1♣-2♥ for invitational and weak Flannery respectively?

That's one way to do it, although personally I'd rather design the system around responding 1♥ on the < GF RFR hands. (Yes, that would make the system more vulnerable to interference over 1♣-(P)-1♥, as in mikeh's example.)

 DavidKok, on 2023-February-28, 07:27, said:

With the both-major hands mostly sorted the system looks strong, although mikeh's comment about interference does have me worried, as does leaking information on exactly game hands.

Could you be more specific? Assume that 1♣-2♦ and 1♣-2♥ are RFRs, if you like.

[DavidKok]

On the auctions 1♣-1♠; 1NT-2♣; 2♠-3NT and 1♣-1♠; 1NT-2♦; 2♥-3NT you have painted a detailed picture of the combined partnership assets, where standard bidders would bid 1♣-1♠; 1NT-3NT. This helps inform the lead (passive versus active).