[straube]

Our 2D opening is 10-15 with 6D and up to 4C (but no major). We'd like to play 2H as GF relay. Looking for suggestions. So far I've come up with...

2H-GF relay
.....2S-4C
.....etc-single-suited
2S-GI with one or both 5-cd majors
.....2N-minimum, no major
..........3C-and 5 clubs
..........3D-and diamond tolerance
.....3C-minimum, 3H (possibly 3S, too)
.....3D-minimum, 3S only
.....3H-maximum, no major
.....3S-maximum, 3H (possibly 3S, too)
.....3N-maximum, 3S only
2N-constructive or GI raise
3C-GI 5+ clubs, short diamonds
3D-weak raise
3M-FSJ?

Especially interested in advice about the continuations after 2S.

The 3C (short diamond) response is annoying because it's low frequency. I could use it for another diamond raise (separate the constructive from GI immediately) but it's kind of nice for responder to grab the NT whenever 3N is in the picture.

Not sure how to spend 3M. Have considered fit showing jumps, GI, or GF with 6. With the relay we don't really need GF with 6. With GI hands we can start off with 2S ask but we can lose a 6-2 fit when opener has a strong hand. OTOH, if we use it as GI and we have a diamond fit, then (for example) we might wind up in a 6-0 major suit fit when we have a 7-3 diamond fit available.

[whereagles]

Not suggesting any sequence in particular, but here are some ideas for responder:

- GF hands: full shape relays, followed by hcp/controls/denial cues. Good bids to start relays would be 2H, 2NT or 3C.
- INV hands use other bids.
- Bids of 2NT or 3C can be used as transfer with INV+ hands that want to bid naturally. E.g.

2D 2NT (relay to 3C)
3C 3D = inv to 3NT. Stoppers next.
.....3H = inv with 55 majors. NF.
.....3S = slam inv with 55 majors. Opener bids 3/4NT=misfit min/mas, 4m=slam accept in linked major, 4M=fit but not really into slam.
.....3NT+ = whatever you prefer.

[Kungsgeten]

I'm sure you've thought of this, but you could use:

2D--
2H = GF relay
2S = Natural, INV but non-forcing
2NT = Hearts, INV but non-forcing
3C = INV, at least some tolerance for diamonds
3D = Preemptive raise
3H = INV, 5-5 majors
3S = Transfer to 3NT?
3NT = To play

A problem is that you can not invite with long clubs.

[rbforster]

Another general thought would be not to use pure GF relays, but rather have an invitational or better relay using 2♥. If opener relayed again, it would be GF while other relay breaks would be natural. This could give meanings to more sequences and help handle more invitational and sign off hand types.

[straube]

rbforster, on 2014-June-18, 16:07, said:

Another general thought would be not to use pure GF relays, but rather have an invitational or better relay using 2♥. If opener relayed again, it would be GF while other relay breaks would be natural. This could give meanings to more sequences and help handle more invitational and sign off hand types.

I use that approach a lot. In fact I'd worked out continuations where 2H was GF or GI spades. This would be a simpler continuation and I haven't seen how we can profitably relay break after the obvious...

2S-4 clubs
etc-single-suited

continuations. If you see where we might, let me know.

[straube]

2H-GF relay
2S-5 spades or 5+ hearts
.....2N-2 or 3 hearts, not 3 spades
..........3C-5 hearts
..........3D-5 spades
..........3H-6 hearts
.....3C-3 spades, 2H
.....3D-3 spades, 0-1H, min
.....3H-3S, 3H, min
.....3S-0-2S, 0-1H, max
.....3N-3S, 0-1H, max
.....4H-3S/3H, max
2N-constructive or GI, f
3C-6S
3D-raise

[rbforster]

straube, on 2014-June-18, 16:25, said:

I haven't seen how we can profitably relay break after the obvious...

2S-4 clubs
etc-single-suited

continuations. If you see where we might, let me know.

No, I don't think you can use that approach. If the relay is GI+, you won't want to go past 3♦ without showing a max (or a fit for whatever GI hands are included). So instead of getting shape first and values second like most GFR schemes, you get a mix of shape and strength on the first ask, and then more shape on the second. It will end up being less symmetric but probably better if you're willing to put the work in to develop and remember it.

[awm]

It seems like you could do something like:

2♥ = 5+♥ invite or GF relay
2♠ = 5+♠ invite
2NT = 4+♣ invite (5+♣ except 4414, NF, normally correct with 3♣)
3♣ = 2+♦ invite

2♦-2♥:
... 2♠ = 0-2♥, not semi-balanced
...... 2nt = relay
............ 3♣ = four clubs not 3♥
............ 3♦ = 3271 or 1273
............ 3♥+ = short hearts, 1363 etc
...... 3♣ = 5/5 hearts and clubs invite
...... 3♦ = 5(+?) hearts invite (without 5 clubs)
...... 3♥ = 6+ good hearts invite
... 2nt = 3262 or 2263 or 2272 (now 3♦/3♥ NF invites)
... 3♣ = short spades, 3♥ (as 3♥+)
... 3♦ = 2362 or 0364 (3♥ asks)
... 3♥ = 3361
... 3♠ = 2371
... 3nt = 3370

Of course, 2♥ is GF if opener has 3♥, but this is probably okay most of the time.

[straube]

Thanks. We played something similar but it didn't allow us to invite with and stop in 2S. What's your opinion of 2m-3M? Using it as GF natural seems redundant when having a relay. FSJs seems promising but I guess they are aways forcing? In an uncontested auction I think I'd like splinters best, but I'm not fond of telling the opponents they have a 9+ major suit fit.

[Siegmund]

I have experimented with 2H=spades, 2S artificial ask, 2NT clubs, 3C hearts, over a natural 2D denying 4CM.

It does have the merit of getting you to 2S often. The price is the asking bid starting one step higher, and the heart hands being very expensive. I probably ought to try 2NT=hearts, to give partner two steps rather than one before passing 3D, and live with 3C natural.

I just have a hard time making myself giving up so many sequences as I do with a 2-level NF response. Maybe there is a good method with 2S inv, but I always find myself wanting more auctions.

[kenrexford]

This 2♦ opening seems really bulky in its handling. I tried unwinding it mentally with a method I use for when 2♦ shows both minors, but there is simply too much variance on the outside. If any 6331 is possible, plus 6-4-3-0's with clubs, on top of 6322's, my own unwind approach breaks down.

This leads me to wonder why you have this opening in the first place? It is loved on its own merits, or is it a necessary part of another structure? If the former, I am worried about you. If the latter, I wonder if a reshuffling of the opening bid structure might avoid the problem.

[straube]

kenrexford, on 2014-June-19, 09:48, said:

This 2♦ opening seems really bulky in its handling. I tried unwinding it mentally with a method I use for when 2♦ shows both minors, but there is simply too much variance on the outside. If any 6331 is possible, plus 6-4-3-0's with clubs, on top of 6322's, my own unwind approach breaks down.

This leads me to wonder why you have this opening in the first place? It is loved on its own merits, or is it a necessary part of another structure? If the former, I am worried about you. If the latter, I wonder if a reshuffling of the opening bid structure might avoid the problem.

Compared to symmetric relays our 2D opening is actually under-loaded. Just to be sure you follow what we're doing, we've at most 4 clubs.

2D-2H GF relay
.....2S-4 clubs resolving at +0 for higher shortness but -1 (good) for lower shortness
.....etc-resolving single-suited at +0

So we have 7 hand patterns with 4 clubs and 13 single-suited (the 7321s don't differentiate the doubleton from the fragment) for a total of 20 hand patterns. What's your 2D opening and how many patterns does it have?

[kenrexford]

straube, on 2014-June-19, 10:22, said:

What's your 2D opening and how many patterns does it have?

My 2♦ is a minor two-suiter, intermediate. The normal expectation in the minors is 5-5, but 5♦/4♣/(3-1) majors is very common, 2-2-5-4 very rare, 2-2-4-5 never, (31)45 rare, and 6-5+ occasional.

The number of patterns is not the sole issue. The issue is more how well you can unwind. With at least 9 cards in the minors, and almost always unbalanced, you can essentially group patterns into those with spade fragments, those with heart fragments, and those with no fragments. This eases up the structure significantly, IMO.

One-suit anchor bids, like yours and Precision/Neapolitan 2♣, even if the 4-card major is eliminated (thankfully), have a problem when Opener can actually have both three-card majors and where the total suit picture is somewhat difficult to unwind.

As a simple example, suppose you can ask for a fragment. If you start with a one-suit anchor bid, you can have two fragments. Hence, knowledge of one fragment leaves the other suits with a wild variance. For instance, suppose Opener can show three hearts. If he has one minor, he has 6-7 in that minor (usually), 3 in the fragment, 0-3 in the other major, and 0-4 in the other minor. That's a lot of variance.

Contrast this with a two-suit anchor. After showing a fragment, each minor is 4-5 in length (with 6430, one-suited), the known major is 3-card, and the other major is 0-1. Thus, each suit now has only a 1-card variance, as opposed to two of the three unknowns having a 3-card variance.

Thus, it is not necessarily the number of patterns initially shown. Rather, it is the rapidity of reducing that maximum number down to tight parameters.

A 1NT opening has wild initial variance (2-5, 2-5, 2-6, 2-6). After Stayman and a 2♦ rebid, the majors are rapidly resolved to 2-3 each, but the minors remain wildly unknown, still at 2-6. This is acceptable because of the major-centric focus of most sequences. This also illustrates another principle, IMO namely that the focus of the rapid resolution has a higher priority in some suits than in others.

A one-suited intermediate call in a minor has a lack of rapid resolution, but it also has that delay in the majors, which is difficult itself. Resolution of one major as 3 leaves the other at 0-3, which means unresolved and tough.

Slow resolution is fine if the other side instead does the pattern description. Hence, 1NT sequences are often geared toward Responder's pattern, because the 1NT holding is too difficult to unwind. The one-suit minor opening causes problems because of the self-preemption on Responder's efforts to resolve himself. This is what ends up going on in your structure and in others. Resolution of Opener's structure is difficult, as a result of which the pattern resolution switches to Responder, with a cost of space to accomplish this.

You then end up with pattern resolution being blended. Opener starts with a pattern that is difficult to resolve. Responder then makes a resolving bid himself, in a sense. Opener then has a blended pattern resolution scheme, which then converts to a blended resolution scheme by Responder. This ends up strained, lacking consistency, difficult to remember, and probably prone to interference.

This is why I firmly believe in intermediates that have rapid pattern resolution potential. These are much less prone to interference, so much so that I use penalty doubles after these types of openings. They enable very pure captaincy for Responder, with a one-side focus on pattern resolution. Memory is not a problem, and structure is consistent. Very few problem sequences exist, and very few problem patterns exist, if any.

[straube]

I think I understand what you're saying, but I'm not sure if the variance issue is as important as you feel it is (I agree that it is important though). I also don't know what the alternative structures would look like. Of course I'd be interested if you posted your system or linked to it.

Something that competes with the idea of variance is "What do you want to tell your partner with your first bid" and limited, long diamonds seems pretty useful. I looked at 30 hands for my opening and also for 5D/4+C openings (which seems fair as you've some rare 4D/5C openings) and found minor suit trump fits as follows...

#trump..................minors (best minor fit)........just diamonds (just diamond fit)
6...............................0.............................................0
7...............................7.............................................8
8...............................12...........................................8
9...............................9.............................................8
10.............................2.............................................3
11.............................0.............................................3

So very small sample but comparable results. OTOH, not all fits are equal and I'd rather be in a 6-1 diamond fit than a 5-2 diamond fit or a 4-3 club fit. Also, with the 2D opening for just diamonds we get to play 2D vs sometimes forced to 3C.

I'd argue that the 2D opening is safer, more "below our LAW level", and perhaps supports more major suit fit-finding than 2D as minors. I've posted a structure that can find all 5-3 and 6-2 major suit fits for when responder is invitational only before 3D has been passed. You've a point about memorization though.

So I know this isn't about which use for 2D is better and that you were making a point about my use having a lot of variance and I agree that it does. Appreciate you contributing.

[awm]

There are a few reasons I prefer the 2♦ 6+ opening over a "both minors" type opening.

1. Safety. While you are almost as likely to have a seven-card fit opposite both minors, that fit might be in clubs and a level higher.
2. Related to the first, the odds that you want to play 2nt are pretty small when responder has an invite, so you can keep that call forcing and bail to 3♦ when no game is available. This is less comfortable if you could have (say) 4423 opposite 1354.
3. Frequency. This only really comes into play if you reserve 2♦ for 5/5 shape, in which case you are way less frequent than the natural 6+♦ meaning. If you allow 5/4 the frequency is comparable (if you allow also 4/5, higher) but this of course creates increased safety issues.
4. Perhaps more important than any of the above is the effect on the 1♦ opening. The main point is that opposite a 6+ card diamond suit, you can often make 3nt on light values by running the suit. You also can have a ten-card fit without much length in responder's hand, which can allow you to make 5♦ on less than 24 hcp (you usually need ten-card fit for this in a minor, because game is one level higher). This means it's quite important to communicate the six-card suit early. In comparison a 1354 hand is very like a balanced hand (you will typically need at least 24 hcp for 3nt opposite a responder with no special shape, and you won't have a ten-card fit unless responder has five-plus in the suit) and even a 1255 hand is not so different in these respects.

[kenrexford]

One aspect of the both minors option you also need to add in is the common ability to strike with a penalty double. With one suit openings, responder less frequently has a penalty double in my experience, and opener's contributions to defense are more tentative. There's a lot of subtle reasons why. As a simple example, the club suit could be the opponents refuge, rougher to find if Opener shows clubs.

[straube]

If I open 2D with 10-15 3-suited short diamonds, I'm using double as penalty also. This doesn't mean that this opening is good (and I don't think it is). I think your opening is an overbid and wouldn't bid aggressively against it lest I take you off the hook and put myself on it.

[relknes]

I personally like symmetry in my systems, and would consider extending whatever system you use over a weak 2. If that is a "First step response is an artificial game force" then great, the continuations shouldn't be hard to adjust. If you use Ogust over a weak 2, then you might consider using something of a similar spirit where 2N is the artificial ask, unless you are really attached to the 2♥ as an artificial GF. It might look like:

2♦-2N:

3♣ = minimum with 4 clubs (or minimum with a weak suit, if you prefer)
3♦ = minimum without 4 clubs (or minimum with a good suit, if you prefer)
3♥ = max with heart stopper, but no spade stopper
3♠ = max with spade stopper, but no heart stopper
3N = max with both majors stopped

Or something... Anyways, just a thought, since your bid is essentially a weak 2 + about 5 points.

[PrecisionL]

relknes, on 2014-June-22, 13:45, said:

I personally like symmetry in my systems, and would consider extending whatever system you use over a weak 2. If that is a "First step response is an artificial game force" then great, the continuations shouldn't be hard to adjust. If you use Ogust over a weak 2, then you might consider using something of a similar spirit where 2N is the artificial ask, unless you are really attached to the 2♥ as an artificial GF. It might look like:

2♦-2N:

3♣ = minimum with 4 clubs (or minimum with a weak suit, if you prefer)
3♦ = minimum without 4 clubs (or minimum with a good suit, if you prefer)
3♥ = max with heart stopper, but no spade stopper
3♠ = max with spade stopper, but no heart stopper
3N = max with both majors stopped

Or something... Anyways, just a thought, since your bid is essentially a weak 2 + about 5 points.

Very similar to what we play in Ultra / C3:

2♦ Opening = 10-14 hcp and 6♦, sometimes 5♦ and 4♣.

2NT Reply = G.I. Values or better and responses similar to the above post.

[straube]

We're kind of settling on

2H-GF relay
2S-GI with a 5+ major
.....2N-min, no 3-cd major
..........3C-6S
.....3C-min, 3H, could have 3S
.....3D-min, 3S, not 3H
.....3H-max, no 3-cd major
..........3S-6S
.....3S-max, 3H, could have 3S
.....3N-max, 3S
2N-9+ with fit
.....3C-game try
3C-GI, 6H
3D-weak
3M-FSJ

Lots of good ideas volunteered here. Thanks all. awm's proposal was better than this, but this has a simple relay so a little less to remember