[straube]

So I compared three strong club structures for passed hand auctions (opener had 17+ and responder 10 or less)

Our present structure....
.....1D-GF, bal or possessing a major
.....1H-any semipositive (2-4 QPs)
.....1S-any negative (0-1 QPs)
.....1N-GF, both majors
.....2C-GF, both minors
.....2D-GF, clubs
.....2H+ GF, various with diamonds

A new structure...
.....1D-any semipositive (2-4 QPs)
.....1H-any negative (0-1 QPs)
.....1S-GF, bal or various minors
.....1N-GF, spades
.....2C-GF, majors
.....2D-GF, H/C
.....2H+ GF, various with hearts

Standard Symmetric with similar QP divisions...
.....1D-any 0-4 QPs
.....1H-GF, hearts
.....1S-GF spades
.....1N-GF, balanced
.....2C-GF, minors
.....2D-GF, clubs
.....2H+ GF, various with diamonds

Here were the breakdowns after 50 hands.....

Our structure
.....1D-20%
.....1H-48%
.....1S-26%
.....1N-4%
.....2C-2%

A new structure
.....1D-48%
.....1H-26%
.....1S-12%
.....1N-4%
.....2C-4%
.....2D-2%
.....2H+ 4%

Standard symmetric structure
.....1D-74%
.....1H-10%
.....1S-4%
.....1N-10%
.....2C-2%

I suspect that if I looked at more hands, the results might not be quite as dramatic. Still, I think this illustrates that the new structure does a much better job of utilizing space. Passed hands worth a positive response have much less to describe. They pretty much can only have 5, 6 or 7 QPs. They never need to super-accept (bid past 3N just to show extra QPs). Also, there is much less chance for slam opposite a passed hand.

If you contrast the new structure with a standard symmetric structure, you see that so much has to pass through the 1D (essentially 0-7 hcp) response. If you can split those responses between something like 0-4 and 5-7 it really ought to pay dividends in the subsequent rounds of bidding. In my other thread, I think I showed that it does. The basic idea is that responder doesn't have to use bids to separate 0-4 from 5-7 later on because he's already done that. Also, opener is empowered. Take an auction like 1C-1D, 1N-2D where responder has promised 5-7 hcps already...if opener now likes hearts, he can super-accept more easily.

So why worry about passed hand structures? I haven't done a tally, but I imagine that a high percentage (maybe 40%?) of the total 1C hands are opened in third or fourth seat. Enough to be concerned about.

[rbforster]

I imagine Adam's or other SP-first style strong club systems might work well in 3/4th. These are the ones were 1♦ is DN or GF, most cheap replies are transfer SPs, and the high responses are specialized GFs. It should be better in competition than your "new" structure which conveys no shape information on the vast majority of SP hands.

[straube]

rbforster, on 2012-December-22, 08:23, said:

I imagine Adam's or other SP-first style strong club systems might work well in 3/4th. These are the ones were 1♦ is DN or GF, most cheap replies are transfer SPs, and the high responses are specialized GFs. It should be better in competition than your "new" structure which conveys no shape information on the vast majority of SP hands.

Have you actually sat down and tried to memorize Adam's structure? I did and just about had it, but imo (and partner's, too) it was very difficult. Are many folks playing it aside from Adam and would they want to learn it for just 3rd/4th seat? I think you are right, though in that it would make a very good 3rd/4th seat structure. I did try it out and didn't like it for slam bidding because in 1st and 2nd seat there are too many 1C-1D, 1N auctions that give up on relays. I know that Adam feels differently about this point, but I remember two trial slam auctions where we really had to guess...and this despite having a very complicated NT structure at our disposal. As a side matter, I don't like that it seldom or never puts us in any kind of force. If RHO intervenes/preempts, there are no forcing passes. So I greatly respect Adam, but don't personally like his club structure though I think it's rather a marvel of engineering. I know he doesn't like SCREAM either, primarily because our semipositive gives no shape information.

I think perhaps you're referring to Moscito as another 3rd/4th possibility? I hadn't considered that, but it would have just the same (high) number of 1S double negatives and none of these hands could land afterward in 2m. It only differentiates (immediately) the SPs with 5-major and those are the easiest to pick up later.

I'm going to guesstimate Moscito's 3rd/4th hand tallies...

1D-32%
1H-30%
1S-26%
etc-12%

reasonable anyway

[rbforster]

I haven't tried to play it, but I reviewed it enough to see that trying to have GF relays as well as constructive invitational auctions involved a lot of details and it would be quite hard to remember unless you wrote it . I know I have some very complicated relay systems in special situations, but they make sense to me and I don't forget them, but that's because I spent a long time thinking about all the issues for some specific sequence before settling on the one I like best.

Personally I just play the same methods in 3/4th. But for me, there are more GF hands - not just the 8-9hcp hands that didn't open, but decent balanced hands that we pass Vul (up to 12 hcp), and the club "openers" in the 10-15 range that I prefer to pass so I can have 2m preempts instead.

[straube]

Yeah, it's easier to remember if you wrote it and spent time thinking about why it had to be thus and so. Of course, it's a good idea to write notes with your partner's memory in mind. Will you grant me that this structure is at least a simple one? And an obvious improvement to 1C-1D (0-7)?

Separate note, but do you have an opinion on my thread on opening 2m preempts in third seat? I'd love to have P P 2C as a weak two.

[rbforster]

straube, on 2012-December-22, 10:49, said:

Separate note, but do you have an opinion on my thread on opening 2m preempts in third seat? I'd love to have P P 2C as a weak two.

I play 2m as weak to intermediate in 3/4. 2C is 8-15, 2D is 5-11. I like 2C as a preempt in 1/2 also .

[benlessard]

In my system its 1C = 15+

1D = SP or H or GF bal ( a lot of hands !)
1H S weak or GF
1S = no M weak or GF
1NT = 6 clubs W or GF
2C = 6D W or GF
2D = 6H W or GF
2H = 6S W or GF

IMO this is vastly superior to a system where you show the SP directly.

A big % of the sequence start with

1C-1D-1H-??

1S = inv no M or GF bal (15-16 vs 7-8 we play 1Nt)
1Nt = inv with 4/5 S (15-16 vs 7-8 we play 1Nt)
2C = D inv or H+D GF (allow us to stop low at 2D or to bid thin gagme based on running D)
2D = H inv or H+C GF -- this may lead you to 3-4 fit with 22 pts however
2H = S inv or H+S GF
2S = C+H GF

We split the NT range

1C--1D--1NT = 17-18
1C-1D--1H and refuse an inv = 15-16
1C--1D--1H and accept an inv =19-20

1C--1D--2NT = 21-22
1C--1D--2D = 22-23

The main drawback is that its not easy for declarer to show hearts sometimes.

[straube]

benlessard, on 2012-December-22, 16:56, said:

In my system its 1C = 15+

1D = SP or H or GF bal ( a lot of hands !)
1H S weak or GF
1S = no M weak or GF
1NT = 6 clubs W or GF
2C = 6D W or GF
2D = 6H W or GF
2H = 6S W or GF

IMO this is vastly superior to a system where you show the SP directly.

A big % of the sequence start with

1C-1D-1H-??

1S = inv no M or GF bal (15-16 vs 7-8 we play 1Nt)
1Nt = inv with 4/5 S (15-16 vs 7-8 we play 1Nt)
2C = D inv or H+D GF (allow us to stop low at 2D or to bid thin gagme based on running D)
2D = H inv or H+C GF -- this may lead you to 3-4 fit with 22 pts however
2H = S inv or H+S GF
2S = C+H GF

We split the NT range

1C--1D--1NT = 17-18
1C-1D--1H and refuse an inv = 15-16
1C--1D--1H and accept an inv =19-20

1C--1D--2NT = 21-22
1C--1D--2D = 22-23

The main drawback is that its not easy for declarer to show hearts sometimes.

My impression is that this structure doesn't empower opener to do much...excepting sometimes when a fit has been found.

Say it goes 1C P 1D (3C). You don't have a forcing pass available. I assume double is takeout, but if you were in a GF, you could double for penalty (or takeout if you prefer that Pass-Double Inversion there) and pass with no suit you want to mention. Same thing for 1C P 1H (3C). It's great if opener has spades, but when he doesn't?

We lose fits on occasion (I admit it) but the way we're structured, we have forcing passes available. Say it goes 1C P 1S (3C) P P dbl....well that's our negative hand doubling with takeout shape. Perhaps responder has Kxxx xxx xxxxx x. Responder has freedom because he's already announced a bad hand. Say it goes 1C P 1H (2S) 3C P 3D P 3S P 3N....well in this auction 1H is a semipositive, 3C shows diamonds, 3D shows non-forward-going acceptance of diamonds, and 3N coughs up a spade stopper. Opener is empowered by knowing that partner has 5-7 to get into the auction using Rubensohl. So we lose fits, but I bet we gain fits, too.

I think by putting your semipositives in with your GF hands you handicap opener more than he would be otherwise. Put another way, opener probably can't use his higher bids (2C, 2H, 3S) as often as he could if 1D promised GF values.

I've made this point before, but I'm not a fan of 2-way bids because it's unusual for the higher bids to be useful opposite both meanings. This new structure I'm looking at uses a 2-way 1H rebid...

1C P 1D P 1H

is either natural or a GF relay. It works because partner only bids at 2C or higher when he is showing hearts...thus this information is welcome whether opener has hearts or just GF values. But I only use this 2-way bid after the auction appears to be uncontested because interference after this rebid (if it's very high) would use up the room opener left for responder. The two-way meaning would be lost.

[benlessard]

Quote

Same thing for 1C P 1H (3C). It's great if opener has spades, but when he doesn't?

We are in a great spot here. Opener just pass or bid 3S with 15-20 knowing that partner is weak or GF. X here show at least 20 pts and at least 2 clubs. Since we also open 1D with 4D 11-22 bidding 3D can be takeout with short clubs

A tougher case is
If 1C (P) 1H (3H)--?_______(1H show spades W or GF)

Responder will never be balanced if GF, so he can bid 3S with exactly 5S+ a minor or double with 4S+5m+ and at least 2H or bid 4m with 4S+5m and a stiff H.

Note however that my system is a lot closer to PC that to a regular strong clubs since when we open 1C we are much more likely to be bal than unbalanced.

15+bal are in 1C and many strong hands are in 1D (11-22) & 1M (11-14 or 18-21 unbalanced).

[straube]

I forgot that there was more to your 1C than 15+. Sounds like there is a lot of nuance then and it might be hard to compare response structures. Thanks for clarifying.

[Zelandakh]

How would the numbers for splitting the SPs look? For example

1♦ = DN or various GF
... - 1♥ = relay
... - ... - 1♠ = DN
... - ... - 1N = GF, 5+ clubs
... - ... - 2♣ = GF, 5+ diamonds
... - ... - 2♦ = GF, 4-5 hearts, bal or 3-suited with short spades
... - ... - 2♥ = GF, 4-5 spades, bal
... - ... - 2♠ = GF, 3-suited with short hearts
... - ... - 2NT = GF, 3-suited with short clubs
... - ... - 3♣+ = GF, 3-suited with short diamonds
1M = SP, 4+ suit
1N = SP, no 4M
2♣ = GF, 5 hearts, <4 spades, unbal
2♦ = GF, 5 spades, <4 hearts, unbal
2♥ = GF, 4 hearts, 5+ spades
2♠ = GF, 5+ hearts, 5+ spades
2N+ = GF, 5+ hearts, 4 spades

I do not think that playing 1♦ as split-range needs to be super-complex. You just need to adjust your preferred relay structure to the available space but reducing space for hands that should not be held after a Pass. Which hands those are depends on the opening structure.

[straube]

Zelandakh, on 2013-January-03, 03:03, said:

How would the numbers for splitting the SPs look? For example

1♦ = DN or various GF
... - 1♥ = relay
... - ... - 1♠ = DN
... - ... - 1N = GF, 5+ clubs
... - ... - 2♣ = GF, 5+ diamonds
... - ... - 2♦ = GF, 4-5 hearts, bal or 3-suited with short spades
... - ... - 2♥ = GF, 4-5 spades, bal
... - ... - 2♠ = GF, 3-suited with short hearts
... - ... - 2NT = GF, 3-suited with short clubs
... - ... - 3♣+ = GF, 3-suited with short diamonds
1M = SP, 4+ suit
1N = SP, no 4M
2♣ = GF, 5 hearts, <4 spades, unbal
2♦ = GF, 5 spades, <4 hearts, unbal
2♥ = GF, 4 hearts, 5+ spades
2♠ = GF, 5+ hearts, 5+ spades
2N+ = GF, 5+ hearts, 4 spades

I do not think that playing 1♦ as split-range needs to be super-complex. You just need to adjust your preferred relay structure to the available space but reducing space for hands that should not be held after a Pass. Which hands those are depends on the opening structure.

Hi Zelandakh. Thanks for your input. I just don't think this would work though.

1) After 1C-1D, it seems unlikely that opener would want to use any other bid but the relay bid of 1H. I mean, it would be difficult to assign useful meanings to other bids. Like suppose opener rebid 1N to show a very minimum balanced hand, but it's almost like starting from scratch. Information has been lost. I mean, you would probably want to invent a structure that would optimize against DN and GF various...or if you didn't optimize, then, well, you would lose something. This is really the trouble with 2-way bids. Unless they're constructed right, partner's higher (non-relay) bids are seldom useful to both kinds of hands pd may have. Like what would 1C-1D, 2D mean? How would you continue after? With our 1C-1D SP, I have useful meanings for all of opener's rebids.

2) 1C-1N, P a lot...but we could be missing a 7/4 club fit or something. Responder may not be balanced. Difficult to say what opener needs to do with say AKQxx xx x AKQxx or whatever. I mean is 1C-1N, 2C a relay bid?

3) 1C-1M, I suppose 2C is the relay bid. This gives up clubs and loses two steps for responder's heart hands. Neither a big deal but it's a loss.

Personally, I think 1C-1D as SP and 1C-1H as DN gives a huge upside to the 0-7 range at a cost of the GF hands. We're just far better off on those hands as opposed to the usual 1D as 0-7 folks. I don't feel the need to go (say) the Moscito route or this route where some of the SPs are immediately set up for relaying. We're far enough ahead already.

[Zelandakh]

straube, on 2013-January-03, 08:54, said:

1) After 1C-1D, it seems unlikely that opener would want to use any other bid but the relay bid of 1H. I mean, it would be difficult to assign useful meanings to other bids. Like suppose opener rebid 1N to show a very minimum balanced hand, but it's almost like starting from scratch.

I do not see this at all. Let us just take your example for a moment. If Responder bids 2♣ over 1NT then you can say immediately that they have the GF hand type. How? Well if they had a DN they would presumably either pass 1NT or transfer into their suit. Speaking of transfers, if all the GF hands with a 5 card major have been removed from 1♦ (I did not in my suggestion since it included GF 5332s) then you also know for certain that a 2♦ or 2♥ response is a DN. And so on. In other words, transfer and pass = DN; anything else = GF - easy.

And there are other ways of using the relay breaks. For example, you might use relay breaks to show the very strong (GF or close) unbalanced hands and relay with weaker unbalanced hands and all balanced hands. Or you can look at your 1♠ DN structure and simply remove all of the problem cases into relay breaks.

straube, on 2013-January-03, 08:54, said:

Information has been lost. I mean, you would probably want to invent a structure that would optimize against DN and GF various...or if you didn't optimize, then, well, you would lose something. This is really the trouble with 2-way bids. Unless they're constructed right, partner's higher (non-relay) bids are seldom useful to both kinds of hands pd may have.

Pretty much every bid that is not a limit bid is a two-way or some sort or another. The reason why natural works is that the various meanings have good homogeneity. It seems to me that DN or GF can also have good homogeneity providing you do it right. In exactly the same way as the weak NT or strong of Swedish Club does, for example.

straube, on 2013-January-03, 08:54, said:

Like what would 1C-1D, 2D mean? How would you continue after? With our 1C-1D SP, I have useful meanings for all of opener's rebids.

Well if 1♣ - 1♦; 2♦ showed a GF hand with long diamonds then I would think a natural approach would work. The DN hands will find the best game and stop; the GF hands will find the best strain and look for slam.

straube, on 2013-January-03, 08:54, said:

2) 1C-1N, P a lot...but we could be missing a 7/4 club fit or something. Responder may not be balanced. Difficult to say what opener needs to do with say AKQxx xx x AKQxx or whatever. I mean is 1C-1N, 2C a relay bid?

Relaying (at +1) would certainly be one possibility here. Then the 2♦♥♠ rebids would be natural and show a minimum hand. That said, I have never played SPs so I could not say that was an optimal arrangement. As for playing in 1NT rather than a 7-4 club fit with ~23 hcp, well I can think of worse problems.

straube, on 2013-January-03, 08:54, said:

3) 1C-1M, I suppose 2C is the relay bid. This gives up clubs and loses two steps for responder's heart hands. Neither a big deal but it's a loss.

It is also a potential win if fourth hand comes in. I think this was the original idea of SPs - get information across quickly on the most common hand type. If I switched to a SP method (and I probably should in truth) then it would definitely be an option that allowed Responder to show distribution with their points.

straube, on 2013-January-03, 08:54, said:

Personally, I think 1C-1D as SP and 1C-1H as DN gives a huge upside to the 0-7 range at a cost of the GF hands. We're just far better off on those hands as opposed to the usual 1D as 0-7 folks. I don't feel the need to go (say) the Moscito route or this route where some of the SPs are immediately set up for relaying. We're far enough ahead already.

In truth I can see the attraction. One of the methods I have played after a strong 2♣ opening was 2♦ positive; 2♥ DN; others SP. I play a variation of it in my own strong auction. This is essentially the same but switching the SPs with the GF hands, which is reasonable when the auction is not GF.

I also agree with you that relaying SPs is not a big deal. Where I disagree is that you are ahead on all of the 0-7 hands. You are ahead on SPs but may well be behind on some of the DN hands. The SP hands are common enough that this is probably a good trade-off but that is even more the case for a method where you have 3 (or more) responses set aside for them.

In other words, I cannot tell you what is best but I can say that if your aim is to maximise the most common hands while harming some of the less common ones (thus producing an overall positive effect) then I do not think you have found the optimal set-up yet.

[straube]

Hi Zelandakh,

1C-1D, 1N-2C most likely would be a transfer, just as 2D and 2H and everything else is a transfer. True, you could relay out opener's hand but this is usually the wrong hand being shown. There's really no reason that 1C-1D, 1N-2D, 2H-2N, 3C-3H can't announce a 2-4-2-5 shape or whatever. But you have to count steps and my feeling is that you are at least 1 step behind standard symmetric on account of partner choosing to rebid 1N rather than relay with 1H plus you were likely behind a step or part of a step because you've given so much more room to the SPs that you are not likely at +0 to begin with.

I think 1C-1D, 2D as potentially a GF bid (even opposite a DN) illustrates my concern about 2-way bids. That's just an unlikely hand for such a low spot. What I'm saying is that 2-way bids are seldom able to "branch" effectively.

Contrast this to my use of 1C-1D, 1H which is opener showing a heart hand or a GF relay hand. Responder is able to branch in a version of symmetric relay (i.e. branch to the max) because his higher up bids show hearts and his highest bids show five hearts...so opener will rarely be disappointed with any very high bid.

Not so if 1C-1D is DN or various GF

I agree your structure wins when responser has a semipositive major and next hand comes in. In fact, it's a big win there. I just don't think you can afford to devote so many bids to the semipositives.

You are right in a sense that we lose out on 1H DN auctions. I mean, opener can't even show a major suit at the 1-level. OTOH, responder has more flexibility in a contested auction to compete. We also have 1C-1H, 1S where responder is free to show any 6-cd suit. So it has some wins, too.

I'll plug it here. We've been practicing it lately and like it a lot. What I like best about it is having a semipositive hand be able to box itself right away and then again. Say the bidding goes 1C-1D, 2C which is nf. Responder has already shown something. Further action shows he's at the high end of his bid. Or 1C-1D, 1H-1S, 2H where opener says he has 6 hearts and can't force game. First opener had the chance to invite with 3H or force game opposite a semipositive so he's kind of split his hand 3 ways...and responder can pass or raise or force game now.

[akhare]

straube, on 2013-January-04, 08:29, said:

1C-1D, 1N-2C most likely would be a transfer, just as 2D and 2H and everything else is a transfer.

...

I think 1C-1D, 2D as potentially a GF bid (even opposite a DN) illustrates my concern about 2-way bids. That's just an unlikely hand for such a low spot. What I'm saying is that 2-way bids are seldom able to "branch" effectively.

...

Contrast this to my use of 1C-1D, 1H which is opener showing a heart hand or a GF relay hand. Responder is able to branch in a version of symmetric relay (i.e. branch to the max) because his higher up bids show hearts and his highest bids show five hearts...so opener will rarely be disappointed with any very high bid.

Not so if 1C-1D is DN or various GF

...

FWIW, I think that playing 1C - 1D DN / some GF hands is very playable.

For example, consider what Adam does in IMPrecision over 1C - 1D (details omitted for brevity):

1H: Natural unbalanced with hearts or strong balanced hands
1S: Natural, unbalanced, forcing one round
1N: Minimum balanced NT
2C: Natural
2D: Natural
2H: Strong with 4+ H, 4c
2S: Minors with at least 5+5+
2N: 6+ in a minor, very strong
3C / 3D / 3H / 3S: GF

Note that as Zelandakh pointed out, bidding 1N with a minimum balanced hands doesn't really cost anything because one can always use 2C as the start of a relay Stayman sequence or something along those lines.

One valid critique about such methods is that they don't set up an immediate GF. However, this is more than balanced out by the equally valid concern that compressed SP responses don't convey any distribution information.

IMO, it ultimately comes down to a choice of what one deems more important (setting up an immediate GF or conveying distribution information).

[Zelandakh]

straube, on 2013-January-04, 08:29, said:

I think 1C-1D, 2D as potentially a GF bid (even opposite a DN) illustrates my concern about 2-way bids. That's just an unlikely hand for such a low spot.

You are probably right about this being a bad suggestion. But it could just as easily be natural and minimum where a DN would Pass and any other action is GF. The point is that you can always find ways of using the information from round one.

straube, on 2013-January-04, 08:29, said:

What I'm saying is that 2-way bids are seldom able to "branch" effectively.

This is far too general. Some multi-way bids branch very effectively and some not. Bids with a weak option, ideally with little shape, and various strong forms tend to work rather well. But that is also too big of a generalisation; best just to judge each case on its specifics.

[straube]

Zelandakh, on 2013-January-14, 05:07, said:

This is far too general. Some multi-way bids branch very effectively and some not. Bids with a weak option, ideally with little shape, and various strong forms tend to work rather well. But that is also too big of a generalisation; best just to judge each case on its specifics.

I think you're quibbling. I said seldom, not never. I use two-way bids in a few instances and each of these instances partner is able to branch effectively because 1) either he's confirming a big fit or 2) partner has a GF hand and wants to know shape information.