BBO Discussion Forums: Comments on Strong Club System Design - Nashua Club - BBO Discussion Forums

I'm new to posting on this forum, but I'm not new to strong club systems, having both played and designed them since 1977. I'd like to get input and comments on a strong 1♣ response concept that I've designed and played sporadically over the past five years. I'm not sure how this works but I'll start the ball rolling with this post - and I'll add more of the design as we go.

I believe that this has benefits over both symmetrical relay (ease of memory, suit quality supplements) and standard xfer responses (better shape derivation).

The basis is a 15+ strong club within the ACBL GCC guidelines of a STRONG opening. The concept is a four-suit canapé, double-transfer system. All spade rebids by responder (effectively a 2nd xfer into notrump) have special meanings.

1♣ = 15+ any distribution

Ranges [HCP/Controls]: 0-6/0-1 weak, 7-9/0-2 invitational, 10-14/2-6 GF, 15+/4+ SI

Basic response structure (1-under xfer):
1♦ = 4+♥ GF or any weak hand (NOT one-suited, else jump pre-empt) - not sure whether to use 0-6 (my preference) or the Moscito range of 0-5 (input requested)
1♥ = 4+♠ GF
The 1R xfer structures are 100% symmetrical for GF auctions.
1♠ = balanced GF or 4+♦ GF (unbalanced)
1N = 4+♣ GF (unbalanced)

2♣ = balanced GI
2♦ = 4+♥ GI
2♥ = 4+♠ GI
2♠ = long minor, denies 4+M, GI
2N = minors (4+5+), unbalanced, GI

3♣ = any SI (this allows the GF bids to be capped at 14HCP/6 controls
3R = weak one-suiters

An immediate take away is that ALL GF hands respond at the 1level, and All invitational hands at the 2level. By having all GI bids at the 2level, it also allows unimpeded bidding over direct seat 1level interference. X is used to denote a GF bid that was space-prohibited by a natural (or passable) overcall with P (pass) being weak balanced. When the overcall is artificial and forcing, pass is a 2-way bid (weak or strong, forcing the opponents to further screw themselves)], and X is used to show very specific shapes (corollaries to what their bid shows). For example, over a suction overcall of 2D (hearts or spades and clubs), X would show diamonds (or hearts and diamonds).

Follow-ups to the 1level bids in the next post.

GF positive reponses are 1-under xfers after the 1♣ opening. All GF bids are at the 1-level.

So after 1♣, then
1♦ = 4♥/5+other or 6+♥
1♥ = 4♠/5+other or 55M or 6+♠
1♠ = balanced or 4+♦/5other or 6+♦
1N = 4+♣/5other or 6+♣
The xfer is essentially a relay.

Over 1♦ or 1♥, Opener ALWAYS accepts UNLESS:
1. They are balanced and have 3cards in the bid major and 2cards in the other major [bid 1N].
2. They are unbalanced and have the other major [or the other major and another] with 0-2 cards in responder’s major. Bid 1♠ over 1♦ and 2x [where x = ♥ + x] over 1♥.

After xfer accept, over both 1level sequences, we have 1-under xfer canapés: 1N=5+♣, 2♣=5+♦, 2oM-1=5oM, 2M-1=6M.
So for example: after 1♣-1♥-1♠ or 1♣-1♦-1♥, then:
1N = 4M/5+♣, and then 2♣=Gamma [sets trump for Epsilon bids], 2♦/2oM=Alpha [or Theta], 2M=Eta [sets trump for Epsilon bids], and 2N=Beta
2♣ = 4M/5+♦, and then 2♦=Gamma, 2oM/3♣=Alpha, 2M=Eta, and 2N=Beta
2oM-1 = 4M/5oM, and then 2oM=Gamma, 3m=Alpha, 2N=Beta
2M-1 = 6+M, and then 2M = Gamma67, 2N = Beta
2♠ = 4M441 [singleton in major not shown by responder], then 2N=Beta, 3m/M=Eta, 3oM=Zeta [asks for honor]

After 1♣-1♠/1N, opener will ALWAYS accept the transfer unless they are single suited in major [6+] without 3cards in the other major, or 2-suited in the majors [5+5+]. In these cases, opener bids 2♦ [majors] or 2M instead of accepting the transfer. Then 2M-1=5M, 3m-1=6+m, 2om-1=4+m/5+om
For example: after 1♣-1♠-1N or 1♣-1N-2♣, then:
2♦ = 4+m/5♥, and then 2♥=Eta5 [trump set for Epsilon bids], 2♠/3om=Alpha, 2N=Beta, 3m=Gamma45
2♥ = 4+m/5♠, and then 2♠=Eta5, 2N=Beta, 3om/3♥=Alpha, 3m=Gamma45
2♠ = 4m441 [both majors, singleton in minor not shown by responder], then 2N=Beta, 3m/M=Eta, 3om=Zeta
3om-1 = 4+m/5+om, and then 3om=Gamma56, 3M=stopper ask, 3m=Gamma45, 3N=to play, 4m=Beta

While this may look complex, it is in fact very easy - responder simply bids his/her hand naturally in a canapé style using the bid below his/her suits. Once the suit[s] are known, we invoke Precision asking bids to determine suit length and high honors [in suits that have a non-fixed length], specific honors [in suits that have a known fixed length], total controls, and specific controls. DOPI handles interference.

Handling balanced GF hands and weak responder's hands next...

I think playing 1♣-1♦;-1♥-1♠ as 0-6 negative is too wide a range

I agree with glen. Also you are way too high with your 2-level semipositives. I would recommend looking at awm's IMPrecision system.

This is not dissimilar to some of the ideas I have played around with. Some thoughts that might inspire you (or might just be noise):-
1. Your 1♠ response is potentially underloaded. You can fit in all GF hands without a 4 card major here if you want. For example, over a 1NT relay you might continue: 2♣ = 5+♦, 4♣ or ♦ 1-suiter; 2♦ = ♣ 1-suiter; 2♥ = balanced; 2♠ = 5+♦, 5+♣; 2NT+ = 5+♣4♦.
2. Freeing up 1NT enables you to sort out your semi-positives more effectively. It is also worth considering whether you have space within the 1♦ response for a balanced SP as that is usually not a difficult hand type to accommodate.
3. One option to avoid certain problems is to split off the 15-17 hands immediately with relay breaks. So the relay then shows 18+ and you have a natural auction with a weaker hand. Doing something along these lines after a 1♦ response means that the wide range stops being an issue.
4. My own experience is that this type of structure is easier to manage if you remove some 15-17 hands from the 1♣ opening, particularly the major-oriented ones. This has a low cost because the 1M openings can easily absorb the extra hands. Obviously that changes the character of the system though so it might not be the direction you want to go in.
5. As well as IMPrecision, Moscito is another obvious system to look at for inspiration. I think your system will be sound once you finish optimising it. I suspect you will find that the alpha, beta, gamma, etc asks will disappear before that point because you will have progressed on to a full relay system. That strikes me as the best way of continuing from the basic method.

Finally, good luck! As an experienced system designer you will know well that there is no substitute for seeing how the system performs in practice and making appropriate adjustments after locating the awkward hands. I think you have quite a way to go at the moment and that you will be able to find many improvements at the start of these auctions. There is a good system in there though if you put in enough time and effort.

glen, on 2015-September-21, 16:00, said:

Is 0-5 narrow enough? This is the range that the 1♣-1♠ Moscito negative uses. It seems to me that 6-9 is too wide a range for semi-positives, especially in view of the comments.

glen, on 2015-September-21, 16:00, said:

Not sure that this would help, but at least you get to show some shape too:

1♣---
1♦ = 4+ hearts GF or any weak without 4 spades
1♥ = 4+ spades, weak or GF

1♣-1♦; 1♥---
1♠ = GF
1NT+ = Weak without 4 spades

1♣-1♥; 1♠---
1NT = Weak with 4+ spades
2♣+ = GF

straube, on 2015-September-21, 16:29, said:

Thank you for the feedback. I've studied IMPrecision among others, and I like the general concept - the problem I specifically have is not its capabilities, but:
1) legality in the confines of the ACBL GCC [no Relay SYSTEMS allowed]. I believe that it qualifies as a relay system.
2) memory requirements - symmetric relay has very few adherents among my partnership group. They predominantly have played Precision with xfer positives and asking bids. Adding shape [as I have attempted to do] is a bonus, but it needs to be simple.

I agree with your sentiment on the high semi-positives. I felt I had to focus more on GF auctions, since this system is designed for IMP usage, where games and slams rule supreme. I'm less concerned about part score contracts. And "standard" Precision does not handle semi-positives particularly well, since it doesn't differentiate between 0-4, and 5-8 point hands (apart from over an intervening double).

What I'm trying to do is provide shape resolution for a specific group of common hands - no 8+ 1-suiters, no 65's, no 74's, just relatively common shapes. Symmetric relay spends an inordinate amount of time and mental energy on shapes that are rare. Furthermore, balanced hand resolution is, more often than not, unnecessary, especially in non slam auctions. Whether the hand is 5=3=3=2 or 5=3=2=3, or 4=4=3=2 or 4=4=2=3 is low on the scale of importance.

Certainly I'd like to improve the semi-positives - but I'd prefer not to negatively impact the GF hands. So my next question is - if I push more GF load on the 1♣-1suit structure and free up 1♣-1N for semi-balanced purposes, would this be enough?

Kungsgeten, on 2015-September-22, 07:36, said:

Unfortunately, this compromises my symmetry design (1N = ♣). The "only" balanced or semi-balanced weak sequence is 1♣-1♦; 1♥-
1♠. (other than when opener opts out of the relay, then pass is on the table).

kwiktrix, on 2015-September-22, 08:28, said:

Well there would still be symmetry, but yes 1NT would not show clubs. The GF hands would start at 2C instead of 1NT. You could play then for instance:

2C = (Semi)balanced
2D = Transfer
2H = Transfer
2S = Transfer to clubs
2N = Transfer to diamonds

The upside would be that the negative hands show some shape too, instead of just "any 0-6". For instance:

1C-1D; 1H---
1S = GF with 4+ hearts (1NT relays and then as above)
1NT = 4 hearts and a 5+ minor OR 6+ hearts, weak
2C = Balanced, weak (probably better to put this into 1NT, in case opener is balanced too)
2D = 5 hearts and 4 diamonds, weak
2H = 5 hearts and 4 clubs, weak

1C-1H; 1S-1NT (weak with 4+ spades);
2C = Take-out of spades, 0-2 spades non-forcing
2D = Strong, forcing
2H = Natural
2S = 3+ spades, non-forcing
2N = "Jacoby"

Kungsgeten, on 2015-September-23, 05:22, said:

Thanks again for the input. I really appreciate it.

However, I'm not sure that a 0-6 hand should be bidding at the 2-level opposite a 15 pointer when no fit has been established. In addition, we should have heard from the opps somewhere along the way, so these sequences are unlikely to play out at the table unless opener has some stuff. And if he has stuff and he is balanced, we want to escape to 1N [by opener]. This is why I like the 1C-1D-1H-1S sequence because it allows opener to play 1N even with a balanced 19-20. Note that responder is not allowed to bid 1D [weak] with 2 controls, so at best you have a 50% chance of getting an entry to dummy in a suit without the Ace in declarer's hand. This makes the 20 opposite 6 point hands unlikely to make 3N. And by having ALL suits already bid [so declarer is already established], responder's next bid over 1N is natural, including 2C. This allows the pair to find 2C when opener has 19 points and dummy 3 points with a 5card club suit.

My concern has always been the semi-positive 7-9 hands, and while the 1N response might be under-utilized [IMO], I can't create much more room without completely moving away from the fundamental concept of the system.

One last point - I'm not sure that all of the posters understand that if responder has a 5=3=1=4 GF hand [for example], he does NOT respond 1H, but 1N [4+c]. A 5card major can never be shown with a 1R response to 1C [unless 55 or better in majors], so with 5M the hand is either balanced [5M332 and hence a 1S response] or a 2suiter with exactly 4cards in the other major [xfer into the other major first], or 4+ cards in a minor [1S/2C response to 1C, then xfer into the 5cM]. In this way, all balanced [5332, 4333, 4432], 54, 55, 6+ suits are delineated before we reach 2N. The remaining suit distribution is left unresolved while we focus on QP and/or controls thereafter in the relevant long suit[s].

I think you should tally 100 (or more) hands for responses to 1C and post the results. That's a basic check that I wish everyone would present when posting novel systems. Your results will likely illustrate that your 1C-1L responses are under-utilized and 1C-2L responses are over-utilized. In addition, your 1C-1D, 1H and 1C-1H, 1S continuations are likely over-utilized. I think your system doesn't branch properly (in accordance with Fibonacci expectations) and is not workable.

Aside from what tabulation will show, one can see that responder's 2-level responses interfere greatly with opener's ability to describe his own pattern (as well as the rest of responder's pattern). I think you are concerned about the semi-positives already. Also, your 0-5 pt hand responding auctions don't allow for fit-finding before the 2-level. IMP considerations notwithstanding, your really hurting with non-GF (0-7) responding hands which are the majority.

You can calculate the frequency of responses using dealer.

This sample script (which will need to be tweaked):

============
north_1C = hcp(north) >= 15

south_1D_resp = (hcp(south) <= 6 and not shape(south, any 6xxx + any 7xxx + any 8xxx))
or (hcp(south) >= 9 and hearts(south) >= 4 and hearts(south) > spades(south))

south_1H_resp = (hcp(south) >= 9 and spades(south) >= 4 and not south_1D_resp)

condition north_1C

produce 1000000

action
frequency "1D" (south_1D_resp , 0, 1),
frequency "1H" (south_1H_resp , 0, 1),
============

deals 1 million hands and calculates the frequency:

Frequency 1D:
0 561490
1 438510
Frequency 1H:
0 845409
1 154591
Generated 7053676 hands
Produced 1000000 hands
Initial random seed 1443050232
Time needed 7.016 sec


Which shows the 1D response occurring 43.8% of the time and the the 1H response 15.5%. Change your definition of weak to 0-5 and 1D now occurs 35.6% of the time, which seems better.

You can look at my system here. This is a vert old version but the basic structure is good.

http://fulldisclosur...tem=s2ftdw1w63o

With 15 to 22 with primary D I strongly suggest to open 1D not 1C.

I also suggest that your transfers are weak or GF. Its the semi-positives hand that are isolated in 1C-1D.

One the reasons is that after a X fer that is weak or GF opener is more likely to make a simple acceptance.

1C---1H (S weak or GF)

here opener is more likely to bid 1S or 1NT. If 1H is 6+ than opener will often jump to 2nt or bid 2S.

The other reason is that SP hand can often compete later ex

1C--(P)---1D---(2H)
P----(P)--2S

its much easier to bid 2S as a semi positive than a 0-5 with good spades.

straube, on 2015-September-23, 12:18, said:

Thank you for your comments as well. Good idea to determine frequency distribution of responses. I use DEAL and I can run a frequency analysis on the response structure using millions of deals if necessary. Hopefully, I can get this done over the weekend.

Quote

I'm not sure that this is relevant in my case - after all this is a "natural" system [albeit using xfers], not an artificial relay, where maximizing space is the objective.

Quote

True to some extent - opener can "break out" of the xfer to show his shape if required. However, consider the differences between this and Precision. Give opener a 5=3=3=2 hand and 16-17 points - in standard Precision the auction goes 1C-1D-1S-1N [don't know whether responder has 8points or 5points, but he has 2spades].

With my proposed system:
If responder has 8points, after 1C-1S-1N-2C, opener knows responder is 7-9 and balanced, and can bid 2S with 16-, which will be corrected to 2N/3N, or a 3H bid with 16+/17, which will be corrected to 3N.
If responder has 5points, after 1C-1D-1H-1S, opener knows responder has 0-6 and needs to bail in the lowest sensible contract [even with 17points], and bids 1N - which can be passed or corrected to a 5card suit in partner's hand.

I believe that this is superior auction and result in both instances.

Quote

Mainly because of their importance in reaching game contracts when opener has 16 or 17 points.

Quote

This is certainly a concern, but I point out that there are many relay Precision designs, where the auction goes 1C-1D [negative, 0-8], 1H [relay], 1S [super-negative, 0-4]. My auction is exactly the same [albeit with up to 6 points].

I also want to add that the semi-positive bids are better in x-fer and as 2 ways bids.

1C-1D-1H (1D is any 6-8 or 4 hearts) 1H show 3 15-20

1S = sp without 4M (can be 6C 6-8)
1NT = 4or5S 6-8
2C = 6D 6-8 or H+D GF
2D = H inv or H+C GF
2H = 6S inv or H+S GF
rest is C+H GF

In short its not enough that you transfer on the 2nd round its important that you x-fer on the second round too, if your responder second round bids are not forcing you lose a lot of bidding space and wrongside the contract. Using a 2nd round 1NT as not forcing is normal but 2m and higher should be x-fer.

My system aslo allow got multiple NT ranges

1C-1D-1NT = 17-18 bal without 4H
1C-1D-1H-1S= 15-16 not always balanced however
1C-1D-2H = 18-20 with 4H (bal or 4H+clubs) (if responder got 6-8 we are GF if hes got 4H we have a fit)

1C-1D-1M-bid 2NT= 19-20 (denies a 5 card suit)
1C-1D-2D = 23-24 (bal or any hand that is near GF)
1C-1D-2NT = 21-22 (bal denies 4H but may have 5 card suit)

For sure having multiple 2 pts range NT is memory intensive but there is a lot less guessing, if partner show 17-18 you rarely need invitationnal bids.

benlessard, on 2015-September-23, 21:35, said:

Interesting idea. Might make more sense to include clubs as well in the 1D opening, since 1D-1M/1N-3suit are available bids. Bears further review.

Quote

While I understand your logic, I simply can't see this working. Using the frequency of 15 vs 16 vs 17 point hands for example, where 15point hands are twice as likely as 17point ones, we are putting in a structure that is not likely to see the light of day - the opps will simply be in the auction anyway - and we have now added X and pass to our arsenal. Weak hands [0-6] are simply not worth putting aside bidding space. I'm surprised at all of the attention 0-5/6 point hands are receiving in this thread.

Quote

This would make sense if opener was not unlimited. For example: 1C-1D-1H-2D-2H-pass with 6hearts and 0 points. Unfortunately, opener can still have 24 points. Now, if we add some sort of "super-acceptance" by opener, so that a simple acceptance is limiting, this makes sense, but then it removes the possibility of asking bids [very important in the system].

Quote

Currently 1H cannot be 6+, it can only be 10-14. If you make it either 0-5 or 10-14, as you suggest, then we need a method of differentiation. This would force opener to super-accept with a big hand, losing the benefits of asking bids when responder actually has the 10-14 hand and we are probing for slam.

Quote

If 1C start at 15 than I assume you play a weak NT so 1D always show 4diamonds right ? (or maybe 4135?) once 1D is natural-ish 15-22 with primary D are better under 1D than under a strong clubs. It also free up 1C-1Y-2D to be artificial strong hand rather than 15-20 with diamonds.

Quote

Shape before strenght is the key to a good bidding system. Strong hands are very easy to bid its the weak hand and the 6-8 that are tough. If the strogn club can be agressive (like 15 bal)I would rather play 1C-1D as artifical GF and all my other bid not GF than the other way around. But both these methods are bad its much better to show shape as much as possible. GF hand can always keep bidding on anyway. Losign the forcing pass is annoying but showign shape is better anyway.

1C-??

1NT to 2H are 6 cards xfer that are weak or GF.

So 2H can be

Kxxxxx
xx
xxx
xx

or

AQxxxx
AJx
xxx
x

I can tell you that these are our biggest imps winners.
I can tell you that if you have the 2nd hand opener will bid 2S like 99% of the times and then the strong response can continue to describe (showing singleton clubs)

I strongly believe that 2C and higher responses should always show 6 or at least a good 5 card suit since they are the hand that lead to competitive auctions. Using 2C and higher for balanced hand or 3 suiters or for hands with 4 card suits is imo awful. If i have a balanced GF I welcome overcalls i dont want to prevent them.

What are you 1NT opening what are your 1D opening ?

Quote

Its easy with 0-5 you pass and with 10+ responder keep bidding.

Even if responder is weak opener will have 90% of his hands in the 15-20 range.

1C--1H--??

1S= at least 3S 15-20 (responder will pass with 0-5 and keep bidding in trasnfer with 9+)
1NT 15-20 without 3S
2C = 15-20 with 6 clubs
2D = can be artificial (for us its near GF so its 23-24 bal or 9 tricks hands)
2H = 15-20 with H (for us its 15-17)
2S = S raise around 21 pts.
2NT= 21-22 without 4S

90% of opener rebids are going to be 1S/1NT/2C/2H/2S none of those bid need to be forcing.

If responder happen to have a GF hands than 99% of opener response are going to be the non forcing one.

kwiktrix, on 2015-September-22, 08:28, said:

I don't think there's enough symmetry in your design, because although you play

1C-1D; 1H-1N = 4H5+C, GF
1C-1H; 1S-1N = 4S5+C, GF
1C-1D; 1H-2C = 4H5+D, GF
1C-1H; 1S-2C = 4S5+D, GF

you also play

1C-1S; 1N-2N = 4+D5+C, GF
1C-1N; 2C-3C = 5+D4+C, GF

So you're actually putting more hands into e.g. 1C-1N; 2C-3C than into 1C-1D; 1H-2C even though you're a whole level higher! To rectify this, I think you should give up the idea that 1N is always a transfer to clubs.

I also think you should be able to stop in 1N with something like 15-17 bal. opposite 6-8 bal., so I suggest the following (while trying to interfere with your design goals as little as possible):

1C-?:
1D = 0-5, any / 6-8, 4-S4-H (or possibly 5M(332)), unsuitable for 1C-2C+ / 9+, "H"
...1H
......1S = 0-5, any
......1N = 6-8, 4-S4-H (or possibly 5M(332)), unsuitable for 1C-2C+ [Opener will pass with 15-17 bal.]
......2C+ = 9+, "H". Specifically:
......2C = 4H5+D
......2D = 6+ H
......2H = 5S4H
......2S = 4H5+C [think of 2S as a (two-under) transfer to C, not as a (one-under) transfer to N]
......2N = 1444 [think of 2N as a (two-under) transfer to D, not as a (one-under) transfer to C]
1H = 9+, "S" / 12+ bal. [edit]
...1S
......1N = 12+ bal. [puts an upper limit on 1C-1S; 1N-2C]
......2C = 4S5+D
......2D = 4S5H [but what do you do with 5S5H?]
......2H = 6+ S
......2S = 4S5+C [think of 2S as a (two-under) transfer to C, not as a (one-under) transfer to N]
......2N = 4144 [think of 2N as a (two-under) transfer to D, not as a (one-under) transfer to C]
1S = 9+, "D" or 9-11 bal.
...1N
......2C = 9-11 bal. [necessary to put an upper limit on 2C here IMO]
......2D = 5H4+D
......2H = 5S4+D
......2S = 5+D4+C [think of 2S as a (two-under) transfer to C, not as a (one-under) transfer to N]
......2N = 6+ D [think of 2N as a (two-under) transfer to D, not as a (one-under) transfer to C]
......3C = 4441 [think of 3C as a (two-under) transfer to H, not as a (one-under) transfer to D]
1N = 9+, "C"
...2C
......2D = 5H4+C
......2H = 5S4+C
......2S = 6+ C [think of 2S as a (two-under) transfer to C, not as a (one-under) transfer to N]
......2N = 5+D4+C [think of 2N as a (two-under) transfer to D, not as a (one-under) transfer to C]
......3C = 4414 [think of 3C as a (two-under) transfer to H, not as a (one-under) transfer to D]
2C+ = inspired by decent 1N defence (such as Multilandy), but with 6-8 and possibly allowing 5M(332).*

*Alternatively, and conforming to your scheme:

2C = 6-8, 5M(332) [would have responded 1D with (4333), (4432) or 5m(332)]
2D = 6-8, 44(41)/(50) or 4H6+m or 5+ H, unbal. [would have responded 1D with 1444 or 3-S4H5m and 2C with 5H(332)]
2H = 6-8, 4S6+m or 5+ S, unbal. [would have responded 1D with 4144 or 4S3-H5m and 2C with 5S(332)]
2S = 6-8, 6+m4Om or 7+ m, one-suited [would have responded 1D with 6m(322) or 6m(331), not wanting to preempt p, who is often minimum with a 5c major]
2N = 6-8, 5+D5+C [would have responded 1D with 22(54) or (31)(54) and 2S with 6+m4Om, allowing p to ask for the longer minor]
3m = 6-8, 7+ m, one-suited [my suggestion, allows 2S to be 6-8, 6+m4Om unambigously]