[akhare]

Adding to the chorus here -- define a QP base (A=3, K=2, Q=1) for opener (suggest 9/10) and the positive responses (suggest 6 with some 5) and don't worry about strength until after you have resolved shape.

In most cases, you will know the combined strength below 3N (need in the ballpark of 20 QPs to consider slam and little less with shape).

[Zelandakh]

olien, on Sep 19 2010, 08:57 PM, said:

Hrothgar, our response structure doesn't allow for this. Our response structure is:

1D=negative
1H=4+S or 12+ balanced (if only 4S, then 3-suited or longer minor)
1S=5+H
1N=5+C (could be 5-5 minors)
2C=8-11 balanced (or 8+ if 33(43) or 4S-4D)
2D=6+D
2H=4+H 5+D
2S=1444, 04(54), 44(50)
2N+=5+D 4C

The simple solution to your problem is to include all the balanced hands with 4-5 spades in your 1H response and the balanced hands without 4 spades in your 2C bid. Within your original concept this might mean

1C - 1H - 1S - 2D - 2H
2S = 4 hearts, 4 spades
2N = 5S332
3C = 4 spades, 4 diamonds
3D = 4333
3H = 4=2=3=4
3S+ = 4=3=4=2

and
1C - 2C - 2D
2H = 2-3 spades, 2-3 hearts
2S = see below
2N = 5H332
3C = 4 hearts, 4 diamonds
3D = 3433
3H = 2=4=3=4
3S+ = 3=4=3=4

I left a gap at 2S because it spoils the symmetry of the structure (giving an overhead on memory). One option would be to use this bid as a 3-suiter without spades. You could, of course, also split the 'no major' balanced hands between 2H and 2S.

FWiiW I split my balanced hands up like this but in a way that always shows a major if I have one. Thus

1C - 1S (no major) - 1N - 2H = balanced with no major
1C - 2D = 4-5 hearts, 0-3 spades, balanced or 3-suited (then all rebids from 2NT up are balanced)
1C - 2H = 4-5 spades, 0-3 hearts, balanced
1C - 1H (spades) - 1S - 1N (hearts) - 2C - 2D = balanced or 3-suited (then 2S= 4432/4441)

Looking over your structure I would suggest a further tweak, namely switching round your 1S bid with your 1NT and 2C bids, ie

1C - 1S = balanced, or clubs
1C - 1N = hearts and clubs, or heart 1-suiter
1C - 2C = hearts and diamonds

This makes it more likely for the strong, unknown hand to be declarer in NT when that is most likely to be the final contract. It has no effect on anything else as these 3 bids form a closed system.

You could, of course, instead do a similar thing with your 2C and 2D/2H responses to give
1C - 2C = 6+ diamonds, or 5+ diamonds and 4 hearts
1C - 2D = various balanced hands (eg 4-5 hearts)
1C - 2H = various balanced hands (eg no major)

You do lose out on being able to show the hearts immediately. An option to show the hearts would be to then switch your 5D4H hands with your 5D4C hands. Thus
1C - 2C = 6+ diamonds, or 5+ diamonds and 4 clubs
1C - 2N+ = 5+ diamond and 4 hearts

So my 2 suggestions for tweaking your responses are

1D = neg
1H = 4+ spades, or some balanced hands (eg 4-5 spades)
1S = balanced, or 5+ clubs
1N = 6+ hearts, or 5+ hearts and 4+ clubs
2C = 5+ hearts and 4+ diamonds
2D = 6+ diamonds
2H = 5+ diamonds and 4 hearts
2S = 3-suiter with 0-1 hearts
2N+ = 5+ diamonds and 4 clubs

and

1D = neg
1H = 4+ spades, or some balanced hands (eg 4-5 spades)
1S = 5+ hearts
1N = 5+ clubs
2C = 6+ diamonds, or 5+ diamonds and 4 clubs
2D = various balanced hands (eg 4-5 hearts)
2H = various balanced hands (eg no major)
2S = 3-suiter with 0-1 hearts
2N+ = 5+ diamonds and 4 hearts

Finally, you can combine these 2 schemes as follows:-

1D = neg
1H = 4+ spades, or some balanced hands (eg 4-5 spades)
1S = 5+ clubs, or 6+ diamonds, or 5+ diamonds and 4 clubs
1N = 6+ hearts, or 5+ hearts and 4+ clubs
2C = 5+ hearts and 4+ diamonds
2D = various balanced hands (eg 4-5 hearts)
2H = various balanced hands (eg no major)
2S = 3-suiter with 0-1 hearts
2N+ = 5+ diamonds and 4 hearts

I think each of these 3 schemes is a small improvement on the original, albeit that the benefits are not large enough to warrant picking any above something you are comfortable with. As you can probably tell from my earlier ramble I take all of the GF hands without a major and push them into a 1S response. If you were also to do this then you would free up the balanced hands within 1H to go back to being 12+ if desired. There is enough space to do this (so long as you move the 4 hearts and 5+ clubs hands elsewhere).
eg
1C - 1S - 1NT
2C = 6+ diamonds, or 5+ diamonds and 4 clubs
2D = 6+ clubs
2H = balanced
2S = 5+ diamonds and 5+ clubs
2N+ = 5+ clubs and 4 diamonds

If you want to stick with your 12+ balanced option then this would be my proposed solution (under the caveats of others that this may not be optimal since you cannot fully resolve the shape).

1D = neg
1H = 4+ spades, or 12+ balanced
1S = 0-3 hearts, 0-3 spades (if balanced then 9-11)
1N = 6+ hearts, or 5+ hearts and 4+ clubs
2C = hearts and diamonds (at least 5-4)
2D = 9-11 balanced with 4-5 hearts
2H = 9-11 balanced with 4-5 spades and 2-3 hearts
2S = 3-suiter with 0-1 hearts
2N+ = 5+ clubs and 4 hearts

Hopefully this final possibility is the kind of thing you were looking for. In reaching it we have lost the ability to promise 5+ hearts when we have both red suits. As that was a premise of your original concept it might be a step too far for you. I do not think there is a better way for you to get your split NT ranges though.

[dake50]

Do you expect top range to zoom above the responses for lower range if/when relayed?
That trades back space when slam/grand is most likely.
Or just hope good answers keep relayer continuing for slam/grand? Can the top range then ever decide to up a slam to grand with unshown values?
I have similar partitions to get asks simplified --at least for common shapes/strengths. There remain some that just persist as "by guess and by golly".

[olien]

Sorry, I didn't clarify very well.

1H=4+S, but may only have a longer minor (not 4S-5+H)
1S=5+H
etc...

Our balanced hands include 5S-(332) and 5m(332), and all 4432s and 4333s.

It looks like we're going to go with:

1H 4+S
1S 5+H
1N 5+C
2C balanced hand, <4S
2D 6+D
2H 5+D 4+H
2S 1444, 04(54), 44(50)
2N+ 5+D 4C


We will probably reorganize the responses to have fewer problems with wrong-siding. Rob F, I would like to see this structure of your that you're talking about if you don't mind sharing.

[straube]

just curious what olien is looking for now. Is keeping 1S as 5 hearts essential? How much space does he want to devote to the balanced hands.

[olien]

I'm flexible with whatever. At the moment, I'm playing a structure which divides all hands into strength immediately, and I'm beginning to dislike it very much. What does everybody think is the best structure? Complexity is not really an issue, but having something symmetric would be nice. I am familiar with TOSR, and most symmetric structures. I am not familiar with MOSCITO, but would like to know more if people think it's worthwhile. The part I am least familiar with are the continuations after the semi-positive responses.

The reason for 1S showing 5+H is for the following reason:

1C-1H// 1S-2C+=hands with 5+S
1C-1S// 1N-2C+=hands with 5+H

This way it is as symmetric as possible

But it's not really that important. However, I would prefer a structure where I can usually show my primary suit first, in case the auction becomes competitive. That's one of the reasons I'm not a fan of TOSR, but if the experience of most people is it's not that important to show primary suits first, then I'm game.

[olien]

Also, regarding QPs. We do use those. A positive response requires 5 QPs, and a 1C opening requires 10 QPs (but may be less with a very strong playing hand)

And yes, the top range strength is allowed to zoom. We don't just guess...we're not total amateurs

[straube]

You can try Mark Abraham's structure. The idea is to let responder be captain when he is balanced and opener is unbalanced and limited/minimum.

1D-bal, H, S, H/m, S/m
.....1H-relays
..........1S-bal, H/C
...............1N-relays
....................2C-H/C
....................other-bal
...............2C-S/C
...............2D-S
...............other-S/D
..........1N-H, H/D
..........2C-S/C
..........2D-S
..........other-S/D
.....1S-H, H/C, H/D
..........1N-relays
...............2C-H/C
...............2D-H
...............other-H/D
.....1N-H/S, H/S/m
.....2C-C/D
.....2D-C
.....etc.
1H-any semipositive
.....1S-GF
.....1N-nf
.....2L-nf
1S-double negative
1N-H/S, H/S/m
2C-C/D
2D-C
2H-C/D/M
2S-D-bal

[rbforster]

olien, on Sep 20 2010, 05:53 PM, said:

I would prefer a structure where I can usually show my primary suit first, in case the auction becomes competitive. 

I agree that generally speaking you want to show your suit early, just in case some joker crashes your auction. I too am a fan of symmetric relay, and designed my strong club system to be as good as I thought I could within the context of being fully symmetric. For example, this meant that I used 1C-1S for all balanced hands (despite it wrong-siding spades sometimes, although less so since 5332's are treated as 1-suited for me). I do put a high premium on right-siding contracts, so ideally I would try something like you describe with breaking out some 4-5♠ balanced hands and putting them into 1♥ but it got too complicated for my level of interest and broke enough symmetry to annoy me.

olien, on Sep 20 2010, 05:53 PM, said:

Our balanced hands include 5S-(332) and 5m(332), and all 4432s and 4333s.

The reason for 1S showing 5+H is for the following reason:

1C-1H// 1S-2C+=hands with 5+S
1C-1S// 1N-2C+=hands with 5+H

This way it is as symmetric as possible

If you're putting most 5332s into balanced, you'll need a lot of space for that. 2D+ isn't going to be enough, probably you'd want 2C+ (i.e. 1C-1S-1N-...). Also, I noticed you were resolving your 1-suited majors with 2C+ (1C-1H-1S-2C+, or 1C-1S-1N-2C+). I want to point out, in case you hadn't done the math, that 1-suited hands starting with 6+ suits are pretty rare in the space of all shapes - there are only 13 per suit of the usual ones. They are much less common than either 2-suited or balanced. Based on analytic calculations, I get for probabilities and shape counts:

51% balanced (including 5332=15%, 4441=3%); # shapes = 32
37% 2 suited (5/4+, includes 5440s); # shapes = 240
12% 1 suited (6-7 card suit); # shapes = 52
0.5% freak (8+); # shapes = 224

My point is that you won't need nearly that much space (showing with 2C+) to show the 1 suited hands as you are allocating to them, and almost certainly this means your balanced or 2-suiters are being squeezed for space and resolving too high. If you count up to 7 card suits and show the 7xx1's merged on the x's (in the typical way), you're looking at 13 shapes (3 each of 6322/6331/7xx1/7330, plus 7222), which is a very nice (Fibonacci) number to fit into a 2S+ zoom in TOSR for example and finishes with 7330 being shown at 3S (so you almost always get 2+ strength ranges at/below 3N, which matters less with a 7 card suit, but whatever).

olien, on Sep 20 2010, 05:53 PM, said:

Our balanced hands include 5S-(332) and 5m(332), and all 4432s and 4333s.

It looks like we're going to go with:

1H  4+S
1S  5+H
1N  5+C
2C  balanced hand, <4S
2D  6+D
2H  5+D 4+H
2S  1444, 04(54), 44(50)
2N+ 5+D 4C

We will probably reorganize the responses to have fewer problems with wrong-siding.    Rob F, I would like to see this structure of your that you're talking about if you don't mind sharing.

My basic structure for balanced hands, showing only 4432/4333, is as follows:

1C-1S-1N-2D+

2D 4H (then 2N shows 44 majors)
2H no 4cM
2S 42 majors (4S if direct)
2N 4M333
3C+ 43 majors (4S if direct)

Note that the above is designed to always right-side hearts, often diamonds, and not to worry about spades at all (since it starts 1C-1S). You may have different considerations. If you really want all my thoughts on this, take a look at my writeup here, under the title "RF TOSR Balanced Relays". In your case, should you decide to keep 5332's in your balanced section, you'll need to start lower and try something else. I have other balanced relay schemes for those cases, but none I worked on quite as much as the one above.

[straube]

I've wondered whether it would make more sense to unravel the major suit holdings of the balanced hands first, then ask QP strength, and then unravel the minor suit holdings.

So...

2-3, 2-4, 2-5, 3-2, 3-3, 3-4, 3-5, 4-2, 4-3, 4-4, 5-2, 5-3 =12 patterns vs the 32 or so total for balanced hands. You need 6 steps so say all balanced hands had to unravel at 2D up...that would mean that the highest bid needed would be 3D....and one would be zooming at 3D. Plenty of room to sort out strength. After the strength ask, RR would complete his minor suit pattern and then DCB.

Something like...

2D=unspecified 5M
.....2S=5H
..........3C-2S
..........3D-3S, 5 QPs
.....2N=5S/2H
.....3C=5S/3H, 5 QPs
2H=4S
.....2N=2H
.....3C=3H
.....3D=4H, 5 QPs
2S=4H
.....3C=2S
.....3D=3S
2N=2S/3H
3C=3S/3H
3D=3S/2H, 5 QPs

[akhare]

straube, on Sep 20 2010, 09:07 PM, said:

You can try Mark Abraham's structure. The idea is to let responder be captain when he is balanced and opener is unbalanced and limited/minimum.

1D-bal, H, S, H/m, S/m
.....1H-relays
..........1S-bal, H/C
...............1N-relays
....................2C-H/C
....................other-bal
...............2C-S/C
...............2D-S
...............other-S/D
..........1N-H, H/D
..........2C-S/C
..........2D-S
..........other-S/D
.....1S-H, H/C, H/D
..........1N-relays
...............2C-H/C
...............2D-H
...............other-H/D
.....1N-H/S, H/S/m
.....2C-C/D
.....2D-C
.....etc.
1H-any semipositive
.....1S-GF
.....1N-nf
.....2L-nf
1S-double negative
1N-H/S, H/S/m
2C-C/D
2D-C
2H-C/D/M
2S-D-bal

I generally agree with this structure, but since the SPs are being compressed into one single bid, why not flip it and use 1♦=SP and 1♥+ as positives?

Of course, the positive steps are +1, but this scheme is much better at landing in part score after 1♣ - 1♦ (not to mention being able to use the same relays over 1♣ - 1♦ - 1♥):

1D: Any SP
1H: Spades, S+m / Bal -> 1N: S+C, 2C: S / S+D, 2D+ = bal
1S: Junk
1N: H / H+m -> 2D = H+C, 2H = H, 2S+ = H+D
2C: D / minors -> 2H = diamonds, 2S+ = minors
2D: Three suited
2H: Single suited clubs
2S+: Majors reversed, etc.

[Free]

Not again, we already had a lengthy discussion why it's a bad idea to put all semipositives into 1 response...

[straube]

Free, on Sep 21 2010, 02:40 AM, said:

Not again, we already had a lengthy discussion why it's a bad idea to put all semipositives into 1 response...

Which thread was that?

[Zelandakh]

olien, on Sep 20 2010, 10:53 PM, said:

I'm flexible with whatever. At the moment, I'm playing a structure which divides all hands into strength immediately, and I'm beginning to dislike it very much. What does everybody think is the best structure? Complexity is not really an issue, but having something symmetric would be nice. I am familiar with TOSR, and most symmetric structures. I am not familiar with MOSCITO, but would like to know more if people think it's worthwhile. The part I am least familiar with are the continuations after the semi-positive responses.

The reason for 1S showing 5+H is for the following reason:

1C-1H// 1S-2C+=hands with 5+S
1C-1S// 1N-2C+=hands with 5+H

This way it is as symmetric as possible

But it's not really that important. However, I would prefer a structure where I can usually show my primary suit first, in case the auction becomes competitive. That's one of the reasons I'm not a fan of TOSR, but if the experience of most people is it's not that important to show primary suits first, then I'm game.

I like to be able to show my major suit immediately. I do not think it is so important to be able to show a minor. I am also a firm believer in right-siding since this is one of the key advantages you have in using a 1-way information system from the weak hand to the strong. So I designed my responses to make it so that Respionder always shows a major if they have one, and so that any hand without a major makes Opener declarer in the high likelihood og playing in 3NT. Full system that matches symmetric for most auctions:-

1D = any non-GF
1H = 4+ spades, unbalanced unless 44(32), not 3-suited with short hearts
1S = any GF hand without a 4 card major
1N = heart 1-suiter, or hearts and clubs
2C = hearts and diamonds
2D = balanced with 4-5 hearts and 2-3 spades, or 3-suited with 0-1 spades
2H = balanced with 4-5 spades
2S = 3-suited with 0-1 hearts
2N+ = anything you want them to be (eg could be used for pre-empts, or for stronger balanced hands, or for freak hands that are otherwise difficult to handle)

Notice how similar the response structures get when you start optimising them for right-siding even when they start out quite different in nature. The balanced hands are almost symmetric because they all start at 2NT. They cannot be fully symmetric since the balanced without major shapes do not exactly match the balanced with major shapes. All of the 1-suiters, 1.5-suiters and 2-suiters are fully symmetric. This is actually quite close to what you are already playing in some ways, but nonetheless further away than what I suggested in the previous post.

[akhare]

straube, on Sep 21 2010, 09:17 AM, said:

Free, on Sep 21 2010, 02:40 AM, said:

Not again, we already had a lengthy discussion why it's a bad idea to put all semipositives into 1 response...

Which thread was that?

Free and Richard outlined their objections to using a single bid for SP responses here:

http://forums.bridgebase.com/index.php?sho...ic=37213&st=75#