[straube]

Any updates on the IMprecision 2D? I was thinking of using the trick of including a GI hand along with the GF relay ask like you folks did for your major suit openings. I came up with the following but maybe you can find an improvement unless you consider it totally hopeless. Still, a relay would be nice to have.

2H-GI 5+S or any GF
.....2S-any shortness, also minimums with 4C
..........2N-GF
..........3C-5S/5C
..........3D-5S
..........3H-5S/5H
..........3S-6S
.....2N-4H
..........3C-GF
..........3D-5S
..........3H-5S/4H
..........3S-6S
.....3C-max with 4C
..........3D-GF
..........3H-5S/5H
..........3S-6S
.....3D-bal
..........3H-GF
..........3S-5S
.....3H-4S, high short
.....etc-4S, low short

2S-4+H, GI
.....2N-4S
..........3C-5H
.....3C-3H

2N-GI

3C-4S, 0-3H, GI

not sure what 3M should be

[steve2005]

What does 2♦ mean?

[awm]

We don't play relays there; never found a way to really make it work. We have also found a lot of advantages in being able to locate stoppers for 3NT.

For example, your structure does not seem able to distinguish min vs max semi-balanced, nor to really have room to relay out the semi-bal shapes.

I think Noble and Mike have a relay, but they have rearranged openings a lot and don't open 2♦ with a four-card major (I think 6♦-4♥ opens 2NT and 6♦-4♠ opens 1♦ for them).

[Kungsgeten]

steve2005, on 2017-November-12, 15:14, said:

What does 2♦ mean?

Limited with opening strenght (about 11-15) and 6+D. May be any 6-4 hand, but not 6-5.

I know some Swedish pairs play relays over this opening, but there the opening denies holding a 4+ major.

[straube]

I didn't know for sure whether you opened 1N or not with semibalanced 14-15. I suppose I would use 2D-2H, 3D-3H as a QP ask or strength inquiry. I don't feel like I'm giving up that much there but it's something.

If you don't mind my arguing, you use relays for your 2C and I guess you'd prefer them for 2D if there were room. Stoppers are useful and your structure gives you a lot of min/max information right away, finds major suit agreement frequently, but if opener has a balanced hand (particularly one without a major) a relay would be nice. Or if opener has a balanced hand with 4 spades and it starts 2D-2S, 3C (showing 3 spades I believe) that's not a great start.

So obviously you've explored relays here. Did you look at combining the GI 5+S hands with your 2H ask? Curious how you might rearrange things if you were determined to use relays.

[nullve]

awm, on 2017-November-12, 15:31, said:

I think Noble and Mike have a relay, but they have rearranged openings a lot and don't open 2♦ with a four-card major (I think 6♦-4♥ opens 2NT and 6♦-4♠ opens 1♦ for them).

Why don't you open 1♦ (and rebid 1♠ over 1♦-1♥) with 4S6D? (I'm looking at your IMprecision notes from July 2009.)

[awm]

We don't view our 1♦ opening as "natural except..." or as a garbage bin of hands that aren't easy to bid. Instead, we use it to show a hand which is (nearly always) one card from balanced. This definition is quite handy, especially in competitive auctions where we often have a guaranteed fit. Adding hands with long diamonds to the mix breaks a lot of our methods (two-suiters no longer have guaranteed fit, competitive transfers are no longer safe, etc). We also believe that we gain a LOT in four-handed bridge by opening 2♦ with a long diamond suit (gets in the opponents way, but provides enough strength that overcalling is quite risky and enough definition that we can often get to the right spot).

As far as relays vs. natural bidding, we do play natural and forcing 2M responses over our 2♣ which handle many of the hands where stoppers are an issue. But I agree it might be nice to have a relay over 2♦ if we are able to do so at a reasonable level. I'm just not convinced this is possible without removing some hand types. We're already a level higher than 2♣-2♦ (which is already losing a step in many auctions compared to standard symmetric). We don't really want to be stuck opening 1NT on all semi-balanced hands (sure we sometimes will do this, but making it automatic seems tactically bad) and we don't really want to turn our well-defined 1♦ opening into a garbage bin.

[straube]

I tried putting the 4S/6D in 1D for awhile. There's room enough for relays after 1D-1H, 1S-2D to comfortably describe the 4S/6D in addition to the other hands that opener might have. I thought Adam's arguments against the 4S/6D inclusion were pretty convincing. I also didn't like the potential for such auctions as 1D-1H (1S) ? at which point you would want 2D as both a strong raise of hearts and natural (6D). Opening 1D artificial with 4S/6D probably doesn't respect the opponents enough.

I'm curious how IMprecision handles responder's big balanced hands without a major. Say opener has a 3325 and something like 18 hcps. No major to mention and I think 2N is only game interest but not sure.

Still liking my relay straw man. Any better ideas? I think it handles all of the GI major possibilities pretty well. Maybe 2D-3M can be stopper-asking/COG.

[awm]

2D-2N (and 2C-2N) includes GF raise when the priority is stoppers for 3nt. Also 2D-3C is natural GF and focused on major suits for 3nt.

Besides having to put all semi-balanced 13-15 through 1nt, you should consider the level of responders ask for RP. It looks like either you will be unable to distinguish 6- and 7-card diamonds (super important for slam bidding) or you will have a LOT of RP asks at 4C (also bad).

[straube]

I'm going to guess...

2N-
.....3C-medium
..........like for max
.....3D-min
..........like for max
.....3H-max, asks H stop
..........3S-asks S stop
.....3S-max, asks S stop
.....3N-max, both M stops

3C-natural, denies M
.....3D-denies club fit
..........like for fit
.....3H-club fit, asks H stop
.....3S-club fit, asks S stop
.....3N-club fit, both M stop

Regarding the 6 vs 7, was your concern for the semi-balanced? The unbalanced should work out. Here's for short hearts which is the hardest.

2D-2H, 2S-2N asks
.....3D-heart shortness
..........3H-asks
...............3S-3163
...............3N-(3)17(2)
...............etc-3073

But yeah, 2D-2H, 3D as semibalanced isn't a wonderful auction, but 10-12/13 semibalanced maybe aren't wonderful for slam.

I still have 2D-3M available for something. I'd thought maybe as a stopper ask. A hand that you might start 2D-2N. Or the equivalent of 2D (2M) 3M except when opp declines to overcall. Like it?

Curious what you think the best relay scheme would be.

[awm]

We tend to show stoppers rather than ask, so we bid the one we have rather than vice-versa. I doubt it matters all that much though. Notice that on the common 3163 in your example, your RP ask will be 4C. There is a big difference between this and 3S as an RP ask (which is what we get on the 2C-2D-2H-2S track). Similarly after 2D-2H-3D, if you want to distinguish shape (even 7D vs 6D) you will need another ask and RP ask will be 4C or above. At this level it's probably better to have set the suit and let opener evaluate how slammish his hand is, rather than have relayed out exact shape and try to look for specific cards. It's different from being 2-3 steps lower where there's more room for "science."

I'd thought of using something where 2D-2H is GF relay, so 2NT+ can be the one-suiters and you're on the same track as 2C. Then 2D-2H-2S = some four-card side suit and after a relay you have:

3C = hearts
3D = clubs
3H = spades high shortage (can zoom with void after next relay)
3S = spades low singleton
3N = spades low void

This is pretty good on the one-suiters, and while the 6421 hands have their RP ask at 4C that is true on normal track as well. The problems with this are:

1. The diamonds-clubs hand resolves pretty high
2. Since the relay is GF, it puts a lot more pressure on the other responses to 2D.
3. No way to distinguish 4072 vs 4063 (but maybe 4072 is a rare hand type).

[straube]

If the math is right...

2H-GI 5S or any GF
.....2S-4C or 4H or 1-suited lower short
..........2N-GF
...............3C-4C
...............3D-4H
...............etc-1-suited lower short
..........3C-5S/4H, GI
..........3D-5S
..........3H-5S/5H, GI
..........3S-6S GI
.....2N-higher short or bal lower fragment
.....3C-mid short (implies spade tolerance)
.....3D-3262 or 7222 (spade fit or tolerance)
.....etc-4S

btw, how would you relay a 6/4 hand?

[foobar]

straube, on 2017-November-15, 10:14, said:

If the math is right...

2H-GI 5S or any GF
.....2S-4C or 4H or 1-suited lower short
..........2N-GF
...............3C-4C
...............3D-4H
...............etc-1-suited lower short
..........3C-5S/4H, GI
..........3D-5S
..........3H-5S/5H, GI
..........3S-6S GI
.....2N-higher short or bal lower fragment
.....3C-mid short (implies spade tolerance)
.....3D-3262 or 7222 (spade fit or tolerance)
.....etc-4S

btw, how would you relay a 6/4 hand?

As awm noted, it may be impossible to fit in everything without removing at least some hand type (probably one 6-4)? Regardless, the proof of the pudding is in the bidding , so if you want to try it out sometime...

[straube]

Tried it out with foobar last night and found a few changes. We noticed that there was nothing really precluding responder from bidding 2S, 2N, or 3C with the intent of bidding game later...as long as responder doesn't need to make a bid shy of game (which would be nf). So Foobar had a 4/7 hand and mentioned spades before bidding game in clubs. I also found I had more need of showing a GI with a long major (2D-3M) because if I start out say with 2S (showing hearts) and opener rebids 3N, I don't know whether opener has tolerance for the major or not. Lastly, I think I'm in favor of showing the full shape for the 6/4s. This gets us way too high for placing cards, but we always managed to get the QP information in time and I think we got all of the slam decisions right for those hands. Foobar argued that after 2D-2H, 2S-2N, 3D-3H, 3S that 3N still needed to be non-forcing where I was wanting it to ask full pattern, but maybe he's right. Plus his way doesn't require an exception. One problem we thought we had but didn't was that we had no bump raise of diamonds. I'd forgotten that 2D-2N handles invitational with diamond tolerance so we bid the whole series without even using that. 2D-2N also solves the stopper question for the majors and it doesn't need to deny GF values, so we can use it pretty much like Adam does. Not sure what an auction like 2D-2H, 2N-3S should show now but it seems to imply 6S/4H. With 4S/6H one might start off with a 2S response. Here's a revision.

2H-GI 5S or any GF
.....2S-4C or 4H or 1-suited lower short
..........2N-GF
...............3C-4C
...............3D-4H
...............etc-1-suited lower short
..........3C-5S/4H, GI
..........3D-5S
..........3H-5S/5H, GI
..........3S-6S GI
.....2N-higher short or bal lower fragment
.....3C-mid short (implies spade tolerance)
.....3D-3262 or 7222 (spade fit or tolerance)
.....etc-4S

2S-GI+ 4H+
.....2N-4S, f
..........3C-5H
.....3C-3H
.....3D-minimum
.....3H-minimum, fit
.....3S-max, stopper ask
.....3N-max, spade stopper

2N-GI+
.....3C-medium
.....3D-min
.....3H-stopper ask
.....3S-H stopper, asks spade stopper
.....3N-both major stoppers

3C-GI+ 4S, 0-3H

3D-bump

3H-GI 6+

3S-GI 6+

[straube]

I noticed that if I put the 7411 together with 6/4 higher short that the higher short is overloaded.

Say I have 3H and higher for a 6/4 pattern. If it goes...

3H-higher or even short
.....3S-relays
..........3N-4162
..........4C-4063
..........4D-4072
..........4H-4171, 6-7
..........4S-4171, 8
..........etc
3S-4261
3N-4360
4C-4270 6-7
4D-4270 8
etc

so the 3H bid has less room but one more hand type to relay. So one alternative is

3H-higher short
3S-4171
3N-4261
4C-4360
etc

which is symmetric but the 4171 resolves too low. So probably put the 4171 in with lower short

3H-higher short
.....3S
..........3N-4162
..........4C-4063
..........4D-4072, 6-7
3S-4261
3N-4360
4C-4270
4D-4171, 6-7

Arguably it may not make that much of a difference, at least as far as deciding on a diamond slam. All of these methods will get you the QP count and shape before you have to decide between 5D and 6D but none are likely to place cards in time.

[foobar]

straube, on 2017-November-16, 23:04, said:

I noticed that if I put the 7411 together with 6/4 higher short that the higher short is overloaded.

Say I have 3H and higher for a 6/4 pattern. If it goes...

3H-higher or even short
.....3S-relays
..........3N-4162
..........4C-4063
..........4D-4072
..........4H-4171, 6-7
..........4S-4171, 8
..........etc
3S-4261
3N-4360
4C-4270 6-7
4D-4270 8
etc

so the 3H bid has less room but one more hand type to relay. So one alternative is

3H-higher short
3S-4171
3N-4261
4C-4360
etc

which is symmetric but the 4171 resolves too low. So probably put the 4171 in with lower short

3H-higher short
.....3S
..........3N-4162
..........4C-4063
..........4D-4072, 6-7
3S-4261
3N-4360
4C-4270
4D-4171, 6-7

Arguably it may not make that much of a difference, at least as far as deciding on a diamond slam. All of these methods will get you the QP count and shape before you have to decide between 5D and 6D but none are likely to place cards in time.

Think you can improve the mnemonics by modeling it after the IMP 2♣...2♦....3♣ and moving things down by two steps, i.e.:

2♦....2♥:

2S: 4♣ OR SS ♦ with high (♠) shortness OR SS ♦ bal with 2=2=7=2
.....3♣: 4♣, then 3♥+ as below
.....3♦: SS ♦ with short ♠
.....3♥: 2=2=7=2
.....3♠ / 3N: Vacant...can probably put something here
2N: SS ♦ with mid (♥) shortness OR SS ♦ bal with no singleton (implies 2-card ♠ tolerance)
.....3♦: SS ♦ with short hearts
.....3♥: 2=2=6=3
.....3♠: 2=3=6=2
.....3N: 3=2=6=2
3♣: 4♥, then 3♥+ as below
3♦: Single suited with ♦, short ♣ (6331, 7(32)1, 3370)
......3♠: 3136
......3N: (32)71
......4C: 3=3=7=0
3♥: 4♠, high (♥) short (add 4=1=7=1)
3♠: 4♠ low (♣) short, 4=2=6=1
3N: 4♠, low short 4=3=6=0
4C: 4♠=2=7=0

It does resolve some the voids above 3N, but they are relatively rare, and if relayer is still interested, we are likely looking at slam.

[straube]

I wanted to make them the same, too, but I realized that I used 4 steps for the balanced on the 2C structure but had only 3 to work with for the 2D structure.
So I decided to pool the 6m3m and 7m222 together as the first step. The higher fragments are shown right away. I have

2C-
.....2D-ask
..........2H-4D or single-suited higher or bal
...............2S-GF
....................2N-4D
....................3C-single-suited higher
....................3D-2236 or 2227
....................3H-2326
....................3S-3226, 6-7
....................3N-3226, 8-9
...............2N-4D
..........2S-4H
..........2N-mid short
..........3C-high short
..........3D-4S, high short
..........3H-4216
..........3S-4306
..........3N-4207
..........etc4117

2D-2H
.....2S-4C or high short or bal
..........2N-GF
...............3C-4C
...............3D-high short
...............3H-2263 or 2272
...............3S-2362
...............3N-3262
..........3C-4C, nf
.....2N-4H
.....3C-mid short
.....3D-low short
.....3H-high short
.....3S-4261
.....3N-4360
.....4C-4270
.....etc-4171

[foobar]

straube, on 2017-November-19, 17:12, said:

I wanted to make them the same, too, but I realized that I used 4 steps for the balanced on the 2C structure but had only 3 to work with for the 2D structure.
So I decided to pool the 6m3m and 7m222 together as the first step. The higher fragments are shown right away. I have

2C-
.....2D-ask
..........2H-4D or single-suited higher or bal
...............2S-GF
....................2N-4D
....................3C-single-suited higher
....................3D-2236 or 2227
....................3H-2326
....................3S-3226, 6-7
....................3N-3226, 8-9
...............2N-4D
..........2S-4H
..........2N-mid short
..........3C-high short
..........3D-4S, high short
..........3H-4216
..........3S-4306
..........3N-4207
..........etc4117

2D-2H
.....2S-4C or high short or bal
..........2N-GF
...............3C-4C
...............3D-high short
...............3H-2263 or 2272
...............3S-2362
...............3N-3262
..........3C-4C, nf
.....2N-4H
.....3C-mid short
.....3D-low short
.....3H-high short
.....3S-4261
.....3N-4360
.....4C-4270
.....etc-4171

Think that for 2♦ you a better off folding the two tail voids into 3♠ and making 3N as 4261. This flows naturally into 4♣ asking QPs for the more common shape, and then 3N is TP. Over 2♣, you can fold the voids into 3♥. which leaves you similarly placed with 3♠ for 4261 and now 4♣ makes the same QP ask. You can then use 3N for the 7411 or choose to show two ranges for 4261.