[awm]

Sam and I have been trying an experiment with new relay-based responses to 1M. I thought I'd describe them here since some people seem to be interested in our methods. The main reason I've resisted playing relay methods in the past is that we play fairly light and wide-ranging openers (many 9s open 1M) and opposite such a range I think it's essential to have two different invites (one that's 11 to bad 12 and another that's good 12 to bad 14), and to be able to stop in 2M when an invite is decline whenever possible (2NT not a great contract). This is hard to arrange with many of the relay schemes I've seen proposed, which often have (at most!) one way to invite. Anyway here's what we're trying:

Our general style of major suit openings is that we play five-card majors; we generally open on the rule of 18 (so many 5-4 9-counts open) although we'll pass minimums where points are not in the suits. Hands with 16+ points open a strong club. We open many hands with a seven-card major at the four-level. Hands at the bottom end of our opening range with a 6-card major and NOT holding four cards in the other major will normally open a weak two (so 1M with a six-card suit generally has 11-15 high). We'll typically pass with balanced hands with doubleton in partner's major and less than around 9 points.

1♠ - 1NT = semi-forcing, up to a bad 12 points. Will not have three-card support for spades.
... Mostly natural rebids; if responder bids 2♠ next it shows 9-bad 12 (so we can stop there on a bad invite). 2NT rebid is 11-bad 12.
1♠ - 2♣ = either GF (any) or a good 12-14 with 4+♥
... 2♦ = clubs, or a (semi)-balanced one-suiter [now 2♥ is GF relay, others are natural and NF with good 12-bad 14 and imply 4+♥)
... 2♥ = diamonds
... 2♠ = six-plus spades and singleton or void somewhere; GF because this is always 11+ hcp
... 2N+ = hearts; this is considered GF because we must have a heart fit and shape (unless responder had a GF anyway)

We lose a step on most relay sequences here.

1♠ - 2♦ = forcing; shows a good invite (12+ to 14-) without three spades and without four hearts
... 2♥ = normal asking bid; responder makes a natural call next with 2♠ being balanced with doubleton, 2NT showing both minors, 3m a six card suit
... 2♠ = GF 6+♠
... 2N = GF 5+/5+ majors
... 3m = GF 5+/5+

1♠ - 2♥ = good spade raise, often three-card limit raise; also includes four-card limit raise with side shortage (this hand bids shortness if opener bids a simple 2♠)
1♠ - 2♠ = normal raise
1♠ - 2NT = four-card mixed or balanced limit raise
1♠ - 3 new suit = wide-ranging "weak" something like 0-9 hcp and 6+ suit

Our sequences over 1♥ are similar but a little different:

1♥ - 1♠ = like a forcing notrump, can include a spade suit; in any case up to a bad 12 hcp, will not have three-card support for hearts
... 1NT = 3-4♠; responder can pass with 31(45), otherwise 2♣ shows both minors, 2♦ shows 4+♠ and asks about spade length, 2♥ is about 9-12 natural, 2♠ is to play
... 1NT-2♦-2♥ shows three spades, 2♠ shows four spades and not super-max, 2NT+ shows four spades and very maximum
... 2m = natural, 2♣ can be three if 2533
... 2♥ = 6+ hearts
1♥ - 1NT = GF (any) or 4+♠ and at least a good 12
... 2♣ = diamonds, or (semi)-balanced with hearts only
... 2♦ = clubs
... 2♥ = 6+ hearts only, side shortness, 11+ points and GF
... 2♠+ = 4+♠, GF because we must have a spade fit and some shape (or responder has an initial GF)
1♥ - 2♣ = good 12 to bad 14, without four spades or three hearts
... 2♦ = normal relay, and responder makes a natural bid; we use responder's 2♠ here to show 5+/5+ minors, whereas 2NT is generally 31(45)

[straube]

I also like to group GI hearts with GF relay, but I've only used it for exactly 4 hearts. Like you, opener's higher rebids promise hearts.

http://www.bridgebas...n-1s-structure/

I like to get rid of the GI 6m hands right away via 1S-3m. Then 1S-2C, 2X-3m is GI 4H/6m

Also like to get rid of 6+H right away. Then 1S-1N, 2m-2H is Lebensohl. It's really pretty useful.

[Kungsgeten]

It looks playable. My main concern would be contested auctions, since it may be hard to establish a GF and/or showing extra shape? Also on partscore deals (or fit based game deals) it may be hard to find a spade fit after the hearts opening. I like the symmetry of 1NT as the relay after 1H, but 2C as the relay after 1S. Perhaps the "intermediate" relays could be used to show specific limited GF hands with the same pattern, where its easier to leave the decision to partner instead of relaying? For instance:

1S-2D; 2H---
3H = Semi-solid club suit, 0-1 spades, 15-17 hcp
3S = Same, but with diamonds
3NT = 5-5 minors, 15-17 hcp, 0-1 spades

Same principle after 1H-2C; 2D

[Zelandakh]

Did you consider my scheme Adam? Basically that would mean your strong invites use an INV+ relay while your weak invites make a weak response. Opener then makes an invite with a maximum that Responder accepts with the weaker invite hand. Obviously weaker invites do not use relays in this method and I think more sequences end up in 3 of a suit. It is probably worth the comparison though.

[awm]

I've looked at a bunch of schemes, and we tried a bunch of things ourselves. Here's our view:

We tried a lot of methods where 1♠-1NT was the relay, because this obviously gains a step on the relay auctions. The problem is that the hands where we don't relay usually end up with pretty lousy results here. For example, suppose responder has a less than invitational 1453 hand. If he responds 2♣ (NF) how does opener know not to pass with 5233? If he responds 2♦ (5+ NF?) we can't easily scramble to a good spot when opener has 5314, etc. This only gets worse if some of the invitational hands are bidding 2m, or if some hands with doubleton spade are bidding 2m, or if we need 2♥ as a raise, etc. Trying to put TWO ranges of invite through 1NT along with all the GF hands just doesn't seem to work; for example if the auction is 1♠-1NT-2♥ now we want 2♠ as relay and so you need 2NT for both the stronger and weaker invite not to mention you cannot play in 2♠ on either invite, etc. I think Zel's methods fall into this situation, although his original design didn't have two invitational ranges to deal with so it probably works a little better under those constraints. Also note that while we say 1♠-1NT is "semi-forcing" we are really passing it with all balanced hands since balanced 14-16 is opening 1NT.

Our overall structure is not very different from what Straube did. One thing we liked about our natural methods though, was the ability to play in 2M a lot with a 5-2 fit. For example we could bid 1♠-1NT-2♦-2♠ showing about 9 to 12- and play 2♠ when opener has 9-13, but we could also bid 1♠-2♣-2♦-2♠ showing about 12+ to 14- and play 2♠ when opener has 9-11. Straube's method forces the "good invite" to pretty much always rebid 2NT when there's not an eight-card fit. Our relay approach gets to 2♠ on most (but not all) good invites which is nice. We're also able to distinguish all ranges for hands with a long minor (weak but not necessarily garbage vs. mildly invitational vs. strongly invitational) whereas I think his method was putting "invites" through the direct jump shift and didn't have a way to distinguish good from bad invites when holding a long suit like this.

I agree that we could use 1♠-2♦-2♥-3M, 3NT, higher for something and we have left it undefined for now. Seems not too critical though.

[straube]

awm, on 2016-November-14, 12:57, said:

Our overall structure is not very different from what Straube did. One thing we liked about our natural methods though, was the ability to play in 2M a lot with a 5-2 fit. For example we could bid 1♠-1NT-2♦-2♠ showing about 9 to 12- and play 2♠ when opener has 9-13, but we could also bid 1♠-2♣-2♦-2♠ showing about 12+ to 14- and play 2♠ when opener has 9-11. Straube's method forces the "good invite" to pretty much always rebid 2NT when there's not an eight-card fit. Our relay approach gets to 2♠ on most (but not all) good invites which is nice. We're also able to distinguish all ranges for hands with a long minor (weak but not necessarily garbage vs. mildly invitational vs. strongly invitational) whereas I think his method was putting "invites" through the direct jump shift and didn't have a way to distinguish good from bad invites when holding a long suit like this.

After 1S-1N, 2D

.....2H-is a marionette to 2S
..........2S-not very max
...............P-min, 2-fit
...............2N-misfit with enough strength to try to make this
...............3C-to play
...............3D-lightly invitational
..........2N-very max, asks what responder's rebid was going to be
.....2S-invite with 2S
.....2N-invite
.....3C-lightly invite, not as strong as 1S-3C
.....3D-heavy invite

Thinking 1S-2D is really better for 5+ hearts. It lets you show invitational 5H/5m and just clarifies other auctions.

[newroad]

Hi Adam,

I read your original post almost as soon as you made it, and had so many ideas flying around in my head that, combined with the fact it was slightly rhetoric in nature, meant I judged discretion was the better part of valour and temporarily kept quiet (unusual for me)

I'm not sure I fully understand your design criteria, or that I agree with your assertion that you need two relays with such a range - going back to the late 80's, the Antipodeans and in particular the Kiwis relayed* opposite a 9-14 range with shapes far more ill-defined than your 1M openings. I don't think the extra hcp that yours have (and that, only with shape) makes yours intrinsically more problematic. However, you play your system, so you are best placed to judge.

All the above said, given your belated switch to relays and your desire to stop in 2M where feasible, I wonder whether taking a different concept from some relay systems might be useful - where there is a "misfit relay" after the first shape showing bid. In this case, the first shape showing bid would be your opening. So ...

• Step One: Semi-forcing 1NT or equivalent
  
• Step Two: Relay, 2 or more in opener's major
  
• Step Three: Relay, 1 or fewer in opener's major

The relays could continue to taste, but simplistically, it allows

1♥ 1NT 2♣ being a minimum, after which 2♦=R and 2♥=to play

Analogous for 1♠ 2♣ 2♦. In the same vein, it allows

1♥ 2♣ 2♦ as a minimum, but two good (or at least, not bad) thing fall out from this

• 2♥ is not need to play, so can be used as R, and
  
• Opener gets to evaluate his hand opposite known shortage if responder breaks relay at this point

Once again, analogous for 1♠ 2♦ 2♥

Maybe slightly left field, maybe not - but I hope of interest.

Regards, Newroad

* in fairness, some of them employed two relays

[Zelandakh]

As I mentioned, my method is not to put 2 invite ranges through 1NT. Indeed the weak invite range is not really considered an invite at all, it is dealt with by Opener showing extras in the same way as 17 opposite 9 is handled by Standard. You are thinking about this wrongly in my view by considering these as 2 invite ranges.

The choice of playing in 2M rather than 2NT with 5-2 in the major is less about structural design than choices. I play it similarly to straube but it it is no issue which way it works. Similarly, the decision of what to do with 3 card club support after a non-forcing 2♣ is easy - you pass without extras and invite if maximum. The whole basis of playing this way is to stop in the first playable spot without values for game and a 7 card fit at the 1 or 2 level counts as a playable spot.

One thing to note is that the specific 5(32)3 hand mentioned would be impossible for me though as my 1♠ opening promises an unbalanced hand. That in turn means that Opener's 2NT rebid typically promises an invite (max) with 3 card support for Responder's suit. Of course it is not such a big issue to incorporate balanced hands here but it adds a lot of pressure to the relays, which are already higher than the usual sym base for a 1♠ opening. Therefore I would suggest finding a way of reducing the hand types in your 1♠ opening if you want to take the relay route. That in turn might provide additional possibilities for your approach too.

In any case, just in case you have forgotten, my basic approach is:

1M
==
1M+1
... - 1M+2 = min without 4oM
... - 1M+3 = 4oM
... - 2M-1 = max, 4+ clubs, GF
... - 2M = max, one-suited, GF
... - 2M+1 and up = max, 4+ diamonds, GF
1NT (over 1♥) = 4+ spades, NF
2m = nat, NF
2♥ (over 1♠) = 5+ hearts, NF
2M = nat, NF
others = various raises, 4+M

Obviously you do not get to stop in 2M when Opener makes the invite, which would definitely be a negative by your criteria.

[awm]

straube, on 2016-November-14, 13:28, said:

After 1S-1N, 2D

.....2H-is a marionette to 2S
..........2S-not very max
...............P-min, 2-fit
...............2N-misfit with enough strength to try to make this
...............3C-to play
...............3D-lightly invitational
..........2N-very max, asks what responder's rebid was going to be
.....2S-invite with 2S
.....2N-invite
.....3C-lightly invite, not as strong as 1S-3C
.....3D-heavy invite

Thinking 1S-2D is really better for 5+ hearts. It lets you show invitational 5H/5m and just clarifies other auctions.

There are some pretty big problems here I think. First, by putting the "good invites" through 1NT, you can no longer pass the response with 12-13 balanced. So you wind up with some of the same issues as in 2/1, where the 2m rebid need not be a real suit, etc. Second, there are some sequences where you lose the invitational distinction -- one would be 1♠-1NT-2♥ where 2NT has to handle the whole invitational range without a fit. Another one is 1♠-2♦-2♥ where you have the same issue. There's also the classic 2/1 issue with 1♠-1NT-2♥ where 3m is something like 5-11 (since 1♠-3m was heavy invite).

One of our observations has been that playing 2M on a 5-2 is typically better (at both IMPs and MPs) than playing 3m on a 4-4 on the same hand; this is pretty counter-intuitive and not what we necessarily expected going in. It's part of the reason for our somewhat strange-seeming invitational method where we could miss a minor suit fit in some situations (although we shouldn't miss a ten card minor fit, which offers good chances at a light 5m game; opener's 5/5 with a minor typically does not pass after 1♠-2♦-2♥-2♠).

[straube]

awm, on 2016-November-15, 10:45, said:

There are some pretty big problems here I think. First, by putting the "good invites" through 1NT, you can no longer pass the response with 12-13 balanced. So you wind up with some of the same issues as in 2/1, where the 2m rebid need not be a real suit, etc.

fred, on 2011-March-02, 15:33, said:

I think 13 is about right, but of course not all balanced 12s and 13s are created equal.

That, along with much of what I have said in this thread, assumes that your opening bids are not extremely light. If you routinely open balanced 10 counts or 5431ish 9 counts when you have a 5-card major, that complicates matters perhaps to the point that there are no great answers. I suppose that is part of the price you pay for opening very light, but maybe I am missing something as I have little experience playing a very light style.

I wouldn't lose any sleep over missing these 13 opposite 12 3NTs because:

- It is relatively rare that both opener and responder have complete maximums
- It is not as if 3NT is always going to make on these layouts
- All systems have their weaknesses
- Both opener and responder should be allowed to upgrade if they think their hand is worth more than its HCP total suggests
- The benefits of being able to play in 1NT and of 2m rebids delivering 4+ cards are easily enough to make up for the this (in my view)

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

Sure, this loses out on a fraction of hands. I once looked at 100 opening spade hands (a small sample, I know) and found zero misses of 25 pt games. Having spent a lot of time on the problem, I decided at that point that if I missed more than an occasional game, I could make such adjustments as the obvious inclusion of 13 pt 5S332s into 14-16 NT...which might distribute the problem more evenly.

There are lots of wins here, too. Often I'm at 1N when 2N or 3N isn't making. There's the certainty that 1M-1N, 2m is four cards and that helps responder a lot deciding on which partial and deciding on whether he can raise or not.

But I also figure that if Fred feels that the semiforcing NT is playable even at the cost of missing an occasional game, that I shouldn't worry because I'm actually better placed than he is at the start. For example, my NT denies 3 spades, denies 6 hearts, denies 5 hearts and GI, denies 6m and GI, and denies usually 4 hearts and GI. So, many more of my GI hands avoid the 1N semiforcing response compared to what he and his partners are doing. 1S-1N tends to be weaker and have more minor cards than say standard 2/1.

awm, on 2016-November-15, 10:45, said:

Second, there are some sequences where you lose the invitational distinction -- one would be 1♠-1NT-2♥ where 2NT has to handle the whole invitational range without a fit.

Well, I like to avoid 1S-1N, 2H without either extra distribution (6/4 or 5/5) or extra strength. The reason is that my 1N forcing can't have a spade fit and is much less likely to have hearts than most methods. So 1S-1N, 2H on Axxxx KJxx Kxx x has much less upside than for most folks. Opener is hoping for 2 spades or 4 hearts and to play a part score. There shouldn't be any game here, at least by system design. So not enough upside bidding again with this hand. We risk getting too high.

So I'd play 1S-1N, 2H-3m as to play and 1S-1N, 2H-2N would have to handle all invites including lightly invitational with a long minor. At least in that case, opener can pattern out if he wants to.

I also btw have after 1S-1N, 2S two separate raises of spades, 3H and 3S which can be light or strong raises to taste.

awm, on 2016-November-15, 10:45, said:

Another one is 1♠-2♦-2♥ where you have the same issue.

Assume you're referring to 1S-2D showing hearts. I just play opener's 2H rebid as not liking hearts and not having 6 spades or anything else worth showing. If responder has 6 hearts and less than invitation he can pass. With an invitation, he can bid again, including 1S-2D, 2H-2S to show two spades and 1S-2D, 2H-3S to invite with three.

awm, on 2016-November-15, 10:45, said:

One of our observations has been that playing 2M on a 5-2 is typically better (at both IMPs and MPs) than playing 3m on a 4-4 on the same hand; this is pretty counter-intuitive and not what we necessarily expected going in. It's part of the reason for our somewhat strange-seeming invitational method where we could miss a minor suit fit in some situations (although we shouldn't miss a ten card minor fit, which offers good chances at a light 5m game; opener's 5/5 with a minor typically does not pass after 1♠-2♦-2♥-2♠).

I can invite and play 2S in a 5-2 fit as well. But I wouldn't miss a 4m-5m fit (responder having 5-cd support) because I would know about opener's side minor.

A couple of things. Don't you lose some 5-3 heart fits when responder has GI strength and opener doesn't have enough strength to check-back? When you break relay...say 1S-2C, 2S-3D are you showing 5H/5D or can it be 4H/6D? If you separate 4 from 5 hearts immediately, the rebid is clear.

[nullve]

awm, on 2016-November-13, 14:14, said:

I think it's essential to have two different invites (one that's 11 to bad 12 and another that's good 12 to bad 14)

Is that because you play rule of 18 1M openings in conjunction with a 14-16 NT? The reason I ask is that I almost¹ play

* rule of 19 1M openings (also when 1-suited)
* 14-16 NT¹
* 1M-1N NF, denying support unless very weak (Opener will pass with 11-13, 5M(332)¹ and also with 10-12 and either 5M3-OM4m or 4S5H)
* artificial schemes over 1M-1♠/N (the one over 1♠-1N having a lot in common with your "modified Gazzilli" scheme)

precisely in an attempt (successful IMO) to solve the problems you describe, and one can obviously play something similar in a context with rule of 18 1M openings and 13-15 NT.

¹ In reality, my 1M openings are always unbalanced so it's irrelevant which 1N range I use.

[straube]

nullve, on 2016-November-15, 15:08, said:

Is that because you play rule of 18 1M openings in conjunction with a 14-16 NT?

I think that's the problem Fred was getting at created by a light opening style and strong (especially 15-17 but also 14-16) no trump openings.

I tallied another 100 hands, this time confining opener to 13 pt 5332 patterns. In those 100 hands, there were zero misses of 25 or 26 point games.

So I tallied a second set of 100 hands, again looking at 13 pt 5332 patterns and this time looking at how often 2D was used. For 2D promising either 6+ hearts or 5+ hearts and GI I found only four hands. For 12-13 with fewer than three spades and fewer than four hearts I found only one.

Out of 200 5332 hands, I would have missed one 25-26 pt game. It happened to be a 26 pt game

I've always thought that 2D as hearts was a bit under-utilized, but those invitational without a spade fit or four hearts are less so.

5332 can be a problem when it's lumped together with light unbalanced openings, but assigning 2D to GI with <3S and <4 hearts only caters to opener having a maximum 5332. It's a stop-gap for those max 5332s because 1N semiforcing handles other hands as well or better. It has the opportunity cost of showing hearts and it also has losses (getting to 2N when we could stay at 1N and missing a superior minor suit fit).

Anyone have an opinion on throwing the 13 5332s into a 14-16 NT? Just the ones worth an upgrade?

[Zelandakh]

straube, on 2016-November-17, 02:26, said:

Anyone have an opinion on throwing the 13 5332s into a 14-16 NT? Just the ones worth an upgrade?

I think 5332 is worth about half a point. If there are factors in the rest of the hand worth another half a point then not upgrading would strike me as the more eccentric option.

[straube]

awm, on 2016-November-13, 14:14, said:

Our sequences over 1♥ are similar but a little different:

1♥ - 1♠ = like a forcing notrump, can include a spade suit; in any case up to a bad 12 hcp, will not have three-card support for hearts
... 1NT = 3-4♠; responder can pass with 31(45), otherwise 2♣ shows both minors, 2♦ shows 4+♠ and asks about spade length, 2♥ is about 9-12 natural, 2♠ is to play
... 1NT-2♦-2♥ shows three spades, 2♠ shows four spades and not super-max, 2NT+ shows four spades and very maximum
... 2m = natural, 2♣ can be three if 2533
... 2♥ = 6+ hearts
1♥ - 1NT = GF (any) or 4+♠ and at least a good 12
... 2♣ = diamonds, or (semi)-balanced with hearts only
... 2♦ = clubs
... 2♥ = 6+ hearts only, side shortness, 11+ points and GF
... 2♠+ = 4+♠, GF because we must have a spade fit and some shape (or responder has an initial GF)
1♥ - 2♣ = good 12 to bad 14, without four spades or three hearts
... 2♦ = normal relay, and responder makes a natural bid; we use responder's 2♠ here to show 5+/5+ minors, whereas 2NT is generally 31(45)

Wondering about

1H-1S artificial GF or natural 6+
.....1N-bal 11-13
..........2C-puppet to 2D for pass or invite
..........2D-GF relay? (maybe too much room)
.....2C-4C
..........2S-GF relay +1 step
.....2D-4D
..........2S-GF relay +1 step
.....2H-6H
..........2S-GF relay +0 mostly
.....2S-4S
..........2N-GF relay +1 step
.....2N-5S
..........3C-GF relay

1H-1N-6-11, not 4S, not 3H

1H-2C-12-13, not 4S, not 3H

1H-2D-LR

1H-2H-raise

1H-2S-6S 5-10

After 1H-1S opener pretends that responder has spades, raising in competition to as high as 3S or using doubles at higher levels to suggest 4S and allowing for pass when responder has a GF without spades.

It's also possible for something like 1H-1S, 2S-3S to allow for pass when responder has an invitational hand and opener is very minimum.

Is the 1S overloaded? It's handling constructive, light and heavy invitational spades along with all GF here.

[awm]

straube, on 2017-October-26, 13:31, said:

Wondering about

1H-1S artificial GF or natural 6+
.....1N-bal 11-13
..........2C-puppet to 2D for pass or invite
..........2D-GF relay? (maybe too much room)
.....2C-4C
..........2S-GF relay +1 step
.....2D-4D
..........2S-GF relay +1 step
.....2H-6H
..........2S-GF relay +0 mostly
.....2S-4S
..........2N-GF relay +1 step
.....2N-5S
..........3C-GF relay

1H-1N-6-11, not 4S, not 3H

1H-2C-12-13, not 4S, not 3H

1H-2D-LR

1H-2H-raise

1H-2S-6S 5-10

After 1H-1S opener pretends that responder has spades, raising in competition to as high as 3S or using doubles at higher levels to suggest 4S and allowing for pass when responder has a GF without spades.

It's also possible for something like 1H-1S, 2S-3S to allow for pass when responder has an invitational hand and opener is very minimum.

Is the 1S overloaded? It's handling constructive, light and heavy invitational spades along with all GF here.

Seems like you have lost the following:

1. Ability to distinguish two invites when responder has spades.
2. Ability to stop in 2S with one-suited spade invites (except when opener rebids 1NT).
3. Ability to find 5-3 spade fits when opener's second suit is a minor.

The only possible advantage I see here is competitive auctions, but while you're ahead on hands with 4-5♠ in competition you've lost a LOT of ground when you have a GF relay in competition.

[straube]

awm, on 2017-October-26, 13:44, said:

Seems like you have lost the following:

1. Ability to distinguish two invites when responder has spades.
2. Ability to stop in 2S with one-suited spade invites (except when opener rebids 1NT).
3. Ability to find 5-3 spade fits when opener's second suit is a minor.

The only possible advantage I see here is competitive auctions, but while you're ahead on hands with 4-5♠ in competition you've lost a LOT of ground when you have a GF relay in competition.

Ok. I agree all of that. It looks as if you rebid 1N with 35(41)? Is 1H-1N still GF or possibly 4+S if GI? Or does it promise 5 if GI?

[awm]

straube, on 2017-October-26, 14:42, said:

Ok. I agree all of that. It looks as if you rebid 1N with 35(41)? Is 1H-1N still GF or possibly 4+S if GI? Or does it promise 5 if GI?

It is kinda symmetric:

1♥ - 1NT = 5+♠ good invite, or any GF
1♥ - 2♣ = good invite, less than five spades (or min GF with 3+♥)
... 2♦ = not four spades, if max excludes some hands
... 2♥ = min, four spades
... 2♠ = max, four spades
... 2N = max 6+♥
... 3m = max 5/5 or better
... 3♥ = short spades max

1♥-1♠ = at most a bad invite, forcing one round, any number of spades, 0-2♥
... 1NT = 3-4♠
... 2m = NAT, 0-2♠
... 2♥ = NAT, 11-15
... 2♠ = some 6m
... 2N = 5♠/6♥

1♥-1♠-1NT
... Pass = 3145 or similar, less than INV
... 2♣ = check-back, 4+♠, normally at least constructive
... 2♦ = 3154 or similar
... 2♥ = 8-12, not four spades, two hearts
... 2♠ = to play
... 2NT = 11-bad 12, normally 3145

1♥-1♠-1NT-2♣
... 2♦ = 5♥-3♠ (can bid on over 2M if 14-15 hcp)
... 2♥ = 6♥-3♠, less than 14-15 hcp
... 2♠ = 4♠, less than 14-15 hcp
... 2NT = 14-15 hcp 4♠ and a void
... 3m = 14-15 hcp 4♠ singleton showing
... 3♥ = 3♠-6♥ 14-15 hcp
... 3♠ = 4522 14-15 hcp

[straube]

I'm used to showing the spades, but I think you get a lot more by partitioning responder's strengths as you've done. Thanks for sharing it.

[straube]

I've replaced my 1H notes with this structure (already replaced 1S notes). I've been looking at hands, too.

I really like how this structure gives both partners more freedom to describe their hands. For example, after 1H-1S, 2C a return to 2H is approximately 8-11 with a doubleton, but responder also has the option to rebid 2N with 11ish. So he'd do this with a misfit and it may not be a wonderful contract (maybe the partnership has 20 hcps or so), but it could be the best available and on a good day opener can raise to 3.

For awhile I was hung up on opener not being able to rebid 2S (which would be the case if the 1S response were natural). In particular, 1H-1S*, 1N**-P could leave the partnership in NT when either a 4/4 spade fit or 6-1 heart fit was available. But this only happens when responder has like 6-7 hcps and specifically 4-1 in the majors. I looked at 1H-1S, 1N-2D being that particular hand but that was a big loser. As written, presumably 3154 works much better.

It really makes sense that a 1S response promising at most a bad 12 would not empower opener to rebid very high. So 1H-1S* and opener has usually 1N, 2C, 2D, and 2H rebids while 1H-1S natural opener has in addition one more rebid (2S)so we seem to be behind here, but if we play 1H-1N as 6-13 this bid is absolutely terrible for letting opener do anything. As has been discussed many times, it's problematic when opener doesn't have a 4-cd minor (such as a maximum 5332) but also problematic if opener has 4 spades and a good hand. 13-15 45(31) has to either rebid a 3-cd minor suit (so responder can no longer trust a 2m rebid to promise 4) or rebid 2H or pass. But with this structure it goes 1H-1S*, 1N and if responder passes he doesn't have an 11 ct and he doesn't have an 8 ct with a doubleton heart. I used to slot 14-15 4522 in with my opening 1N because I had no bid after 1H-1N but I won't need to do that any more.

So 1H-1N* as GF relay (else strongly invitational with 5S) compared to 1H-1N as semiforcing lets opener make use of all of the 2S+ rebids. 1H-2C* also gives opener a lot of rebid freedom. Really good trades.

Seems like a lot of people would want to use this. It's better than 1S natural or Kaplan Inversion. It requires a slightly higher end for a weak two. It works better with relays but probably those who don't use relays could use it, too; they would have to ensure that you can't rebid higher than 2D after 1H-1N unless you have either 6H or 4S. Should be doable. Just 2C for bal/other and 2D for diamonds if nothing else.

Other folks looking at using this?

[nullve]

straube, on 2017-November-06, 11:36, said:

but if we play 1H-1N as 6-13 this bid is absolutely terrible for letting opener do anything. As has been discussed many times, it's problematic when opener doesn't have a 4-cd minor (such as a maximum 5332) but also problematic if opener has 4 spades and a good hand. 13-15 45(31) has to either rebid a 3-cd minor suit (so responder can no longer trust a 2m rebid to promise 4) or rebid 2H or pass.

These problems exist if you open 1♥ with 10-13, 5H(332) and insist on NAT rebids over 1♥-1N(6-13).