[straube]

 Zelandakh, on 2017-October-10, 02:54, said:

What I am unclear on is how you deal with a weak 2542 or 1552 though. The latter in particular seems to be difficult to handle - perhaps 2♥ should include such 5♥-5♦ hands in addition to 6+♥?

You're welcome.

The 2542 is a particularly difficult hand. I seemed to have the best chance with a 2C response. I was looking at hands after an IMprecision-type 1C-1D, 1S start in which 1S was forcing, denied hearts and promised only 4 spades. In your case, you have a limited 1S bid so pass is an alternative. Supposing that you have a lighly invitational hand and want to raise partner's rebid, 2C is the most flexible. The worst outcome (and it is bad) is that partner passes you in a 4-2 fit when a 5-2 fit is available, but if he happens to support you with a 2N or 3C rebid and you have that light invitational hand, you can rebid 3H along the way to game.

The 5/5 red hand has the same problem as other players after 1S-1N, 2C-? Playing 1S-1N as GI+ I'd respond 2D promising only 5 the large part of the time.

If one's 5-cd heart suit will play opposite shortness and you don't mind a doubleton raise....

[awm]

As far as the weak hands go, it seems like you are ahead on responders 5-5 but behind on 1453 and 15(43); I’m not sure this is a good trade!

In particular, you are likely to do better by requiring opener to bid over 2C unless he has four. This avoids embarrassing combos like: 6223 opposite 2344, 5413 opposite 1534, and so forth. You can also then bid 2C with 1453 since you are pretty happy with any continuation (except I guess a raise).

It also seems like you will play some 4-3 and 5-3 minor fits instead of 5-2 major fits. At MP this is quite bad; at IMPs you might compensate the frequent “lose one” with the minor suit makes and the major doesn’t, but it seems a bit uncomfortable at times.

[nullve]

Another idea:

1M-[1M+1](= INV+ relay); ?:

1M+2 = "10-13", 5+M3-OM4+m OR "14-17", 1-suited
...1M+3 = GF relay
...other: don't know, but the fact that Opener has something resembling a Muiderberg-type hand if "10-13" might be useful
2M = "10-13", 1-suited
...P = allowed with 1- M!
...2M+1 = GF relay
...other = ?
other: same as before

[straube]

I ran 24 hands on BBO. Opener is 10-15 and 5-7 spades and responder 6-11 with 0-2 spades.

I bid hands according to both sets of rules and I pretended that neither hand would invite. I.e. the exercise is about finding the best part score at the lowest level.

The scheme I proposed has a bad result board 19 but good results on boards 1, 8, 12, 17, 20, 23 and possibly 2, 4, and 13. This is a small number of hands, but it bears out what I'd found concerning staying out of opener's way.



1. K8653 A5 V AKJT97....T7 KQ63 A975 Q52

.....1S-2D, 3C..........1S-2C, P

2. KT8542 A7 J8 K87.......A Q9543 T954 942

.....1S-2H, ?...........1S-2C, 2S

3. AKQ54 QJ864 2 97.......72 752 AQ75 J532

.....1S-2C, 2H..........1S-2C, 2H

4. KQJT4 QT3 K KT54.......63 8 QJ875 AJ732

.....1S-?...............1S-2C, P

5. AKJT97 T QJ5 K32.......82 AK864 K62 J54

.....1S-2H, 2S..........1S-2C, 2S

6. KT852 QT6 AK62 7.......J9 J54 Q8543 KQ5

.....1S-2D, P...........1S-2D, P

7. AT654 A6 AK75 84.......7 754 964 AKQJ95

.....1S-2C, 2D-3C.......1S-2C, 2D-3C

8. AJT65 K Q65 K874.......9 J9853 K3 AJ632

.....1S-2H, 2S, 3C......1S-2C, P

9. KQT987 J93 4 A53.......J2 Q86 QT86 JT42

.....1S-2C, 2S..........1S-2C, 2S

10. K9752 KQ764 QT 6.......43 J932 AJ8653 8

.....1S-2D, 2H..........1S-2D, 2H

11. AJT976 V AQ63 K83......K2 AT632 52 QJ92

.....1S-2H, 2S..........1S-2C, 2D-2H, 2S

12. AKJ74 K T532 543.......T9 AJ97632 AQ4 V

.....1S-2H, 2S..........1S-2H

13. AT876 Q A4 QT643.......K5 J654 KJT92 82

.....1S-2D, ?...........1S-2D

14. AJT87 J974 AJ A5.......Q2 532 KT94 QJ84

.....1S-2C, 2H-2S.......1S-2C, 2H-2S

15. AKT543 K K54 763.......Q6 AJ9432 J86 T8

.....1S-2H, 2S..........1S-2H, 2S

16. AKQJ97 43 754 Q8.......32 KT975 8 KT974

.....1S-2H, 2S..........1S-2C, 2S

17. AT754 2 QJ5 KQ54.......K2 A76543 AT32 7

.....1S-2H, 2S..........1S-2H

18. KQJT72 AJ42 3 K5.......3 95 KQJT9864 T7

.....1S-2D, 2H-3D.......1S-2D, 2H-3D

19. AK852 T2 Q5 A532.......97 AQT53 A853 J3

.....1S-2H..............1S-2C

20. A9874 V 3 KQJ8542......J6 AKT83 JT98 96

.....1S-2H, 3C..........1S-2C

21. QT6532 AK86 K7 3.......V Q753 J9653 KQJ9

.....1S-2C, 2H..........1S-2C, 2H

22. AKT52 9754 K AT7.......74 JT AJT5 KQ942

.....1S-2C, 2H-2S.......1S-2C, 2H-2S

23. J86543 A J5 AK74.......7 743 AKT8762 53

.....1S-2D, 2S-3D?......1S-2D

24. A8762 J95 KQT9 2.......54 Q732 8642 A83

.....1S-2D..............1S-2C, 2D

[nullve]

 straube, on 2017-October-10, 19:53, said:

1. K8653 A5 void AKJ97

Only 12 cards.

[awm]

My view is that there are basically four weak hand types without fit:

1. (Semi-)balanced with doubleton in openers suit. While 2M rates to play fine, we want to check for a big fit somewhere else and also distinguish from the 3-fit hand that raises directly.
2. Three suiter short in openers suit. Includes 1(543) patterns as well as 1444 and 0(544). Basically wants to hear openers second suit and play there.
3. 5/5 hand, looking for the better fit usually in one of responders suits.
4. One suiter 6+ cards (and short opener suit). Pretty much knows where we want to play!

The standard approach of responding 1nt on weak hands is great on 1 and 2, and works okay on 4 (although sometimes you wind up at the three level especially holding a minor suit) but it’s pretty bad on 3.

The approach outlined by Zel seems a lot better on 3, but it seems worse on the other types. Straube’s approach seemed a bit better until he started bidding 2c on doubleton and having openers pass with four, which helps him a bit on type 4 hands but seems disastrous on some of the others.

I think I’d accept getting to the three level on many of the type 4 hands (and maybe get there fast via 1M-3x) in exchange for better results everywhere else. This seems pretty doable with a bit of tinkering with the standard approach — if 1nt is always less than inv you can use 1S-1N-2x-2 new suit as 5-card suit in a flexible hand and 1S-1N-2H-2N as both minors. Not sure how to do as well with natural NF bids, assuming I prefer to play 5-2 major fits over 4-3 minor fits.

[straube]

I would also recommend looking at 5D3C hands because they might do better responding 2C as well.

Well 2542 is a problem and I can easily construct other difficult combinations of hands.

Any suggestion for the 2542? Raise spades with constructive values and pass otherwise? I think that's where I'm at.

In any case, Zel very much likes his 1N GI+ response and I thought only to improve the handling of the less than GI hands. I.e. I like what I'm playing with 1N as semiforcing but I wouldn't claim it's better without studying it more.

Btw, I left unexplored what happens when responder has a light GI hand and I suspect that my suggestion would help a lot here, too. Just as an example, if you have 1S-3H as 6-cd less than GI then you can't have 1S-2H, 3H-4H. I think one can see at a glance that my way is less preemptory and tends to give both partners more chances to raise on average.

[Zelandakh]

I was going to go through some of the hands from straube but this seems more to the point somehow:

 awm, on 2017-October-11, 08:58, said:

My view is that there are basically four weak hand types without fit:

1. (Semi-)balanced with doubleton in openers suit. While 2M rates to play fine, we want to check for a big fit somewhere else and also distinguish from the 3-fit hand that raises directly.
2. Three suiter short in openers suit. Includes 1(543) patterns as well as 1444 and 0(544). Basically wants to hear openers second suit and play there.
3. 5/5 hand, looking for the better fit usually in one of responders suits.
4. One suiter 6+ cards (and short opener suit). Pretty much knows where we want to play!

If this is the criteria to cover, it seems fairly easy:-

2♣ = 3+ clubs (covers 1C, 2 and 4C)
2♦ = 4+ diamonds (covers 1D, 3DC and 4D)
2♥ = 5+ hearts (covers 1H, 3HD, 3HC and 4H)

This combines ideas some of the ideas from all of us, essentially moving the 1(54)3 hands down from 2♦ to 2♣ and keeping everything else the same. I need to have a think about the repercussions here but on the surface it seems to be an improvement, which would be great!

[awm]

 Zelandakh, on 2017-October-11, 11:06, said:

I was going to go through some of the hands from straube but this seems more to the point somehow:


If this is the criteria to cover, it seems fairly easy:-

2♣ = 3+ clubs (covers 1C, 2 and 4C)
2♦ = 4+ diamonds (covers 1D, 3DC and 4D)
2♥ = 5+ hearts (covers 1H, 3HD, 3HC and 4H)

This combines ideas some of the ideas from all of us, essentially moving the 1(54)3 hands down from 2♦ to 2♣ and keeping everything else the same. I need to have a think about the repercussions here but on the surface it seems to be an improvement, which would be great!

I do think this is an improvement, but there are still some annoying features:

If opener has 5224, there is no good rebid to 1♠-2♦.
If opener has 5134 and responder has 2443, you will play a 4-3 minor fit instead of a 5-2 spade fit. At MP this is definitely inferior.
If responder has a 6-card suit you will often end at the three-level, except on hands where you have a ten card fit (which might make a light game) where you stop at the two-level.

[yunling]

 straube, on 2017-October-11, 09:25, said:

I would also recommend looking at 5D3C hands because they might do better responding 2C as well.

Well 2542 is a problem and I can easily construct other difficult combinations of hands.

Any suggestion for the 2542? Raise spades with constructive values and pass otherwise? I think that's where I'm at.

If you are removing so many hands out of 2♦/2♥ then I think you should seriously consider to adopt the Auken-Welland approach of 2♦=doubleton support here. It is more important than showing a diamond one-suiter, imo.

[straube]

 awm, on 2017-October-11, 11:17, said:

If opener has 5224, there is no good rebid to 1♠-2♦.

True. 5314 is also awkward. I think opener needs to be able to treat a 2D response as a 6-cd suit even though it will occasionally be a 5-cd suit. With 5314 I think you need to be able to pass 2D. I like 1S-2D, 2S to be 6 spades (and not running) such that responder may raise spades with a 2-fit.

 awm, on 2017-October-11, 11:17, said:

If opener has 5134 and responder has 2443, you will play a 4-3 minor fit instead of a 5-2 spade fit. At MP this is definitely inferior.

I'm hoping he plays IMPs more I always design for IMPs myself.

Anyway, I'm glad that Zel thinks the discussion is useful. My amended rules would be

2S-could be 2542 if light invitation
2H-6H
2D-5+ diamonds, 3-card preference to clubs
2C-3+ clubs

I think this would do better with the 24 hands I looked at, but a fair test would require a new set of hands.

[straube]

 yunling, on 2017-October-11, 12:21, said:

If you are removing so many hands out of 2♦/2♥ then I think you should seriously consider to adopt the Auken-Welland approach of 2♦=doubleton support here. It is more important than showing a diamond one-suiter, imo.

Interesting. What are their other responses?

[Zelandakh]

 awm, on 2017-October-11, 11:17, said:

I do think this is an improvement, but there are still some annoying features:

If opener has 5224, there is no good rebid to 1♠-2♦.
If opener has 5134 and responder has 2443, you will play a 4-3 minor fit instead of a 5-2 spade fit. At MP this is definitely inferior.
If responder has a 6-card suit you will often end at the three-level, except on hands where you have a ten card fit (which might make a light game) where you stop at the two-level.

To be honest I made the last post just before leaving the office and was going through the numbers in my head a little on the 2♦ response. Of the 3 cases, obviously 4D has 6+ diamonds and 3DC is 5-5, so the only case with 4 diamonds and <3 clubs would be 2542, which can respond 2♥ given the safe haven of 2♠. So it seems that the 2♦ response can be 5+ diamonds. That means that your 5134 hand can choose between pass and 2♠. The former gains on case 4D, the latter will usually be best on 3DC and perhaps also 1D, at least at MP .

#2 is certainly true but I cannot see any alternative that would not involve playing at the 3 level a lot more often, which is worse.

I am not too worried about light 5m contracts - they are pretty rare and for many of them Opener will be able to make some sort of invite. Missing a light 4♥ contract is a bigger consideration but here I think the structure is ahead of standard rather than behind. Indeed this was one of the advantages that cropped up during the live playtesting we did a few years back.

[nullve]

 straube, on 2017-October-11, 12:35, said:

Interesting. What are their other responses?

Their CC from BB 2017:

http://www.ecatsbrid...lland-Auken.pdf

See also post #15 in this thread.

[Zelandakh]

 nullve, on 2017-October-11, 13:27, said:

Their CC from BB 2017:

The CC shows that the range of this 2♦ response is 9-13. This is necessary for them because they play their relay as GF. It seems to sit less well into the GI+ relay methods and my instinct is that Adam and straube's idea meshes better, making this not the right approach for this family of systems.

[straube]

What are your other (2N+) responses and how does your 1S-1N relay unwind?

I'm wondering if you could slot a few hand patterns (e.g. lightly invitational 2542) in with your 1N response.

So for example 1S-1N, 2m-2S would be 9-13 doubleton spade.

[yunling]

 straube, on 2017-October-11, 12:35, said:

Interesting. What are their other responses?

(from nullve's link)
They play 2♣=forcing NT and 2♥=good raise(which we do not need)
2♦shows 9-13 since their 1♠ is 10-20 and they need to split the range.

So reconstruct it like
2♣=4-5♣ or 2-3♣&0-1♠
...bid up the line,2NT=max 4♣
2♦=2♠ or 6+m, if 2♠ then not 4-5♣
...2♥=4+♥ any range
...2♠=no 4♥ not max
...2NT=no 4♥ max
2♥=6+♥

So I'll lose when you start with 1♠-2♦-P and also possibly 1♠-2♣-2♦-P but plays a lot more 5-2 major fits and will not passout a 10-card club fit at 2 level.

[straube]

Had the thought lately that Zelandakh is in a much better position than Adam and I with his 1S opening in that he doesn't have to deal with the 5332s. It's largely those patterns (maybe minimum 5422s and other minimums)that make a semiforcing NT response desirable...and then we have the problem of missing game when the 5332 is 13 hcps and responder 12 or 13.

So Zelandakh, you have a structure in which opener always has a comfortable rebid and the 1N response is pretty much going to be forcing whether it's GI+ or less than GF. I think your 1S opening is much more suited to 1N as less than GF and 2C as GF than what either Adam or I or doing. Curious why you're using 1N as GI+. If it's to save a step, it's pretty costly.

More complicated....

1N-forcing
.....2m-4+m
..........2D-6D
..........2H-Lebensohl
...............2S-p/c
....................weak actions
..........2S-2-fit, constructive
..........etc-invitational (e.g. you have two ways to raise opener's minor)
2C-GF (or GI with 4H/6m)
.....2D-4+C, 5440s or 5+D
.....2H-6S
.....2S-4D
.....2N-5422 or higher short
.....3C-4H, lower short
.....etc-5H
2D-5+H, GI OR 6+H
.....2H-minimum, no fit
.....2S-6S
2H-constructive or LR with 3
2S-bad raise
2N-LR+ with 4
3m-GI, not 4H
3H-GI
3S-mixed R

Less complicated....


1N-forcing
.....2m-4+m
..........2D-6D
..........2H-6H
..........2S-preference
2C-GF (or GI with 4H/6m)
.....2D-4+C, 5440s or 5+D
.....2H-6S
.....2S-4D
.....2N-5H
.....etc-4H
2D-5+H
2H-constructive or LR with 3
2S-bad raise
2N-LR+ with 4
3m-GI, not 4H
3H-GI
3S-mixed R

[straube]

 yunling, on 2017-October-11, 20:24, said:

(from nullve's link)
They play 2♣=forcing NT and 2♥=good raise(which we do not need)
2♦shows 9-13 since their 1♠ is 10-20 and they need to split the range.

So reconstruct it like
2♣=4-5♣ or 2-3♣&0-1♠
...bid up the line,2NT=max 4♣
2♦=2♠ or 6+m, if 2♠ then not 4-5♣
...2♥=4+♥ any range
...2♠=no 4♥ not max
...2NT=no 4♥ max
2♥=6+♥

So I'll lose when you start with 1♠-2♦-P and also possibly 1♠-2♣-2♦-P but plays a lot more 5-2 major fits and will not passout a 10-card club fit at 2 level.

So at worst 1S-2C is 1552?

Not sure I like the 2-way nature of the 2D response though it would frequently work out. Have you run it through the 24 hands? I'd be curious to see how it does.

[yunling]

 straube, on 2017-October-11, 20:56, said:

So at worst 1S-2C is 1552?

Not sure I like the 2-way nature of the 2D response though it would frequently work out. Have you run it through the 24 hands? I'd be curious to see how it does.

Yes, 1552 is forced to bid 2♣.

Compare to your method this wins on 19 and possibly 1/7/13
loses on 6/24 and possibly 23
Seems that it is good for thin games?

1. K8653 A5 V AKJT97....T7 KQ63 A975 Q52

1S-2D-3C-4C-5C ? not quite sure after 3C but 3C show max 5-5 so possibly a win

2. KT8542 A7 J8 K87.......A Q9543 T954 942

1S-2C-2S tie

3. AKQ54 QJ864 2 97.......72 752 AQ75 J532

1S-2D-2H-2S Responder should have correct to 2S in other methods as well so I'll consider it a tie

4. KQJT4 QT3 K KT54.......63 8 QJ875 AJ732

1S-2C-P tie

5. AKJT97 T QJ5 K32.......82 AK864 K62 J54

1S-2D-2S tie

6. KT852 QT6 AK62 7.......J9 J54 Q8543 KQ5

1S-2D-2S lose

7. AT654 A6 AK75 84.......7 754 964 AKQJ95

1S-2D-2N-3N possible win

8. AJT65 K Q65 K874.......9 J9853 K3 AJ632

1S-2C-P tie

9. KQT987 J93 4 A53.......J2 Q86 QT86 JT42

1S-2C-2S tie

10. K9752 KQ764 QT 6.......43 J932 AJ8653 8

1S-2D-2H tie

11. AJT976 V AQ63 K83......K2 AT632 52 QJ92

1S-2C-2D-2H-2S tie

12. AKJ74 K T532 543.......T9 AJ97632 AQ4 V

1S-2H tie

13. AT876 Q A4 QT643.......K5 J654 KJT92 82

1S-2D-2S possible win, especially at MPs

14. AJT87 J974 AJ A5.......Q2 532 KT94 QJ84

1S-2C-2H-2S tie

15. AKT543 K K54 763.......Q6 AJ9432 J86 T8

1S-2H-2S tie

16. AKQJ97 43 754 Q8.......32 KT975 8 KT974

1S-2C-2S tie

17. AT754 2 QJ5 KQ54.......K2 A76543 AT32 7

1S-2H tie

18. KQJT72 AJ42 3 K5.......3 95 KQJT9864 T7

1S-2D-2H-3D tie

19. AK852 T2 Q5 A532.......97 AQT53 A853 J3

1S-2D-2S win

20. A9874 V 3 KQJ8542......J6 AKT83 JT98 96

1S-2D-3C no I'm not opening 1S here

21. QT6532 AK86 K7 3.......V Q753 J9653 KQJ9

1S-2C-2H tie

22. AKT52 9754 K AT7.......74 JT AJT5 KQ942

1S-2C-2H-2S

23. J86543 A J5 AK74.......7 743 AKT8762 53

1S-2D-2S-3D 3D on 7-2 fit, small loss but usually not matter

24. A8762 J95 KQT9 2.......54 Q732 8642 A83

1S-2D-2S lose