[straube]

Atul and I have been practicing the IMprecision method for 1M openings. Has anyone else adopted it? I think it's a big advance for Precision relay partnerships. Posting because we have some questions about fit auctions, but first want to share about the non-fit auctions that we've spent most of our time on.

The heart opening has an extra step compared to the spade opening and I think 1H-2C, 2N is that extra step and it was designated as 6H. This is nice in that it puts us in a GF immediately and shows that sixth heart, but we've wondered if it is better to wait like we are required for our spade openings (1S-2D, 2H) and let responder show his hand. So 1H-2C, 2D in other words. Otherwise, it might go 1H-2C, 2N-3N and we miss a 7-1 (or good 6-1) heart fit. Maybe 1H-2C, 2N could be 5S/6H GF?

Another thing is after 1H-1N, 2m-2S to use a bid of the 4th suit as an artificial force. For example, I think 1H-1N, 2C is either diamonds or balanced, so after 1H-1N, 2C-2S, 3C would be that artificial GF. Maybe sometimes this works for 1S auctions. 1S-2C, 2H-2N, 3C?

We're wondering if there are any developments in the structures? What are good continuations for 2M-1 and 2N auctions. I think it's....

1H-2D either 3-cd LR or 4-cd LR with shortness

which is nice because if opener thinks too long and then rebids 2H the partnership can agree that the 4-cd LR with shortness must show the shortness. Am I right on that? I'm not sure about the laws.

But 1H-2N is either a balanced LR or any MR. We've no idea how to continue. In fact, we don't really know what to do after 1H-2D. Kokish?

It could go...

1H-2N
.....3C-what kind of raise do you have?
..........3D-good MR
..........3H-bad MR
..........4H-bal LR
.....3D-interested only in LR
.....3H-yuck

Is that what you're doing? Or hsgts? And then for 1S-2N there is extra room. How to use that?

[hrothgar]

Back in the weird old days, Klinger's POWER system used

1X - 2♦ as a GF relay
1X - 2♣ was game invitational

If the relay captain started with the GI relay, RR would make a 1 step response to show a minimum. With a hand strong enough to accept the invite, RR would respond as if the initial ask had been 2!D.

If I had an extra step lying around, I might consider something similar

[straube]

Klinger's Power System was for 5-cd majors? So then RR's relay steps are the same if RR is strong enough to enter a GF. That's nice on the noggin.

RDB is so much better than anything else I've seen that I think of it as having "solved" the limited 5-major opening. There's a bit of wiggle-room, like exactly how you define 2M and 2M-1 and 2N fit responses. But for example, compare it to KPS....

1H-1N as GI 5+S or any GF. That's two steps lower than KPS and includes a relay break hand (GI 5 spades) that takes pressure off elsewhere.

1H-2C as GI various loses to KPS because RDB needs two steps for minimum hands, one with and one without side spades. I think the loss is very small because a relay auction doesn't matter much when game is the probable limit of the hands.

1H-2D as LR is huge. Idk but suppose that KPS shows its 3-cd LR at the 3-level because showing fit is important for opener to evaluate his hand properly and 1H-2C, 2D-2H wouldn't show that fit. 1H-2D shows fit right away in case the auction gets competitive.

[awm]

straube, on 2019-September-28, 22:40, said:

The heart opening has an extra step compared to the spade opening and I think 1H-2C, 2N is that extra step and it was designated as 6H. This is nice in that it puts us in a GF immediately and shows that sixth heart, but we've wondered if it is better to wait like we are required for our spade openings (1S-2D, 2H) and let responder show his hand. So 1H-2C, 2D in other words. Otherwise, it might go 1H-2C, 2N-3N and we miss a 7-1 (or good 6-1) heart fit. Maybe 1H-2C, 2N could be 5S/6H GF?

After 1H-2C-2N, we use 3C as a shortness ask; this seems really useful and not something we can necessarily get after 1H-2C-2D; it also helps with decisions whether to play a 6-1 fit vs. 3NT I don't really see how starting with 1H-2C-2D-2N(or 3m)-3H is going to help you make better decisions in such situations. We have some other artificial follow-ups too (1H-2C-2N-3D=clubs, 3H=diams, 3S=5/5 minors). And we tend to open a lot of hands with 7H with 3NT or 4H.

straube, on 2019-September-28, 22:40, said:

Another thing is after 1H-1N, 2m-2S to use a bid of the 4th suit as an artificial force. For example, I think 1H-1N, 2C is either diamonds or balanced, so after 1H-1N, 2C-2S, 3C would be that artificial GF. Maybe sometimes this works for 1S auctions. 1S-2C, 2H-2N, 3C?

Note that most weak hands can just pass 2S (it shows six). Other continuations for us are:

2N = shortness ask
3 of the "known" minor = weak 6/5 or more
3 of the other minor = 4th suit GF
3H = natural GF
3S = Hx or maybe a really bad balanced hand with three-card support, inviting game if the fit seems worth an upgrade
3NT = to play
4 of the other minor = splinter with 3-fit

straube, on 2019-September-28, 22:40, said:

1H-2D either 3-cd LR or 4-cd LR with shortness

which is nice because if opener thinks too long and then rebids 2H the partnership can agree that the 4-cd LR with shortness must show the shortness. Am I right on that? I'm not sure about the laws.

We are now putting all limit raises through 1H-2D (it's just "limit raise"). The 2NT bid is a pure mixed raise. We play natural / help suit tries over the mixed raise if it matters (nothing special here). We play Kokish over 1H-2D. If we start 1H-2D-2H, then responder can make a Kokish-style try of his own. This is mandatory with 4-fit and a singleton. While we understand our UI responsibilities and try to do the right thing if it comes up, we don't generally design our methods around "what if we break tempo" (rather, we design what we think are good methods and then try to bid in tempo... or if we can't bid in tempo, don't make the weakest possible call and put partner in a bind; note that a slow game try isn't really indicative about what the problem might've been in the same way a slow sign-off is).

[straube]

Thank you. I really like it. Also see 1H-2C, 2N as GF 5/6 is unnecessary.

I never realized that 1H-1N, 2m-2S was 6S. I assume with 5S/2H, responder breaks relay to 2H when he can.

For 1S-2C, 2D auctions, the relay breaks would resemble....
.....2S-5+H/2S
.....2N-5+H/0-1S
.....3C-5+H/5+C
.....3D-5+H/5+D
.....3H-good 6+ card suit ?

And assuming that...

1H-1N, 2C-2S, 3D would be 5H/6D and not 6H/5D? Or maybe it needs to be either way.

Do you have a way to look for a better fit in the OM after a 2M-1 start? We don't.

Cleared up a lot. I like the clean break between 2M-1 and the mixed raise. Just easier for continuations. Thanks again.

[foobar]

awm, on 2019-September-29, 15:10, said:

The 2NT bid is a pure mixed raise. We play natural / help suit tries over the mixed raise if it matters (nothing special here).

What is the range for the direct splinter bids over 1M? Also, does auction #2 below make sense in your methods?

1) 1M - 4M -> Classic wide ranging raise opposite limited hands, no slam interest
2) 1M - 2N - 3M (ostensibly sign-off) - 4M -> Distributional hand that was too strong to bid 1M - 4M directly and was mildly interested in slam?

[straube]

foobar, on 2019-September-30, 15:00, said:

What is the range for the direct splinter bids over 1M? Also, does auction #2 below make sense in your methods?

1) 1M - 4M -> Classic wide ranging raise opposite limited hands, no slam interest
2) 1M - 2N - 3M (ostensibly sign-off) - 4M -> Distributional hand that was too strong to bid 1M - 4M directly and was mildly interested in slam?

Think I remember that they use FSJs instead of splinters, but at the 4-level (or 1H-3S). Not sure. I think I switched that for splinters primarily because I understand what a splinter is supposed to show, but FSJs require some discussion. For example, how pure do they have to be? Do they require 4 trump? What range of strength? Like, I'm not sure how I would bid AQJxxx KJx x xxx after a 1H opening. Maybe they go 1H-3N (FSJ in spades?) and get to spades sometimes. That could partly answer my question about getting to the OM sometimes. Idk.

Well, he states 1M-2N is a mixed raise. Are you thinking of some sort of weak freak hand? LIke your partner opens 1S and you hold Kxxxx void Qxxxxx Ax or something? Interesting idea though. I picture that they relay a lot of hands that others are using Jacoby 2N on.

[foobar]

straube, on 2019-September-30, 20:17, said:

Think I remember that they use FSJs instead of splinters, but at the 4-level (or 1H-3S). Not sure. I think I switched that for splinters primarily because I understand what a splinter is supposed to show, but FSJs require some discussion. For example, how pure do they have to be? Do they require 4 trump? What range of strength? Like, I'm not sure how I would bid AQJxxx KJx x xxx after a 1H opening. Maybe they go 1H-3N (FSJ in spades?) and get to spades sometimes. That could partly answer my question about getting to the OM sometimes. Idk.

straube, on 2019-September-30, 20:17, said:

Well, he states 1M-2N is a mixed raise. Are you thinking of some sort of weak freak hand? LIke your partner opens 1S and you hold Kxxxx void Qxxxxx Ax or something? Interesting idea though. I picture that they relay a lot of hands that others are using Jacoby 2N on.

Yeah, something along those lines. Basically, a hand that wants to commit to game, might be too weak for a regular splinter (depending on the range), potentially too strong for a direct 4M. Note that I am not entirely sure it's a good idea (since there might be a conundrum over 1M - 2N - 4M for example), but just curious about the treatment.

Regarding relaying, I would be wary about relaying with 4+ card support except in a strong balanced hand).

[awm]

foobar, on 2019-September-30, 15:00, said:

What is the range for the direct splinter bids over 1M? Also, does auction #2 below make sense in your methods?

1) 1M - 4M -> Classic wide ranging raise opposite limited hands, no slam interest
2) 1M - 2N - 3M (ostensibly sign-off) - 4M -> Distributional hand that was too strong to bid 1M - 4M directly and was mildly interested in slam?

We have switched to tiered splinters. While there isn't a precise point range, our agreement is:

1♠ - 3NT = any splinter; interested in slam only opposite a maximum with the right holding opposite the short suit
1♠ - 4♣♦♥ = splinter; interested in slam opposite 11-12 with the right holding opposite the short suit (or most maximums), but not 11-12 with the wrong holding opposite shortness

Auction #1 exists in our style, but we are less enthusiastic about using it than some strong club players. If the right (5431) max will offer play for slam we prefer the 1♠-2♦-2X-3♠ or 1♥-2♣-2X-3♥ sequence (which shows such a hand). This means the direct 1M-4M, when strong, normally will not have a three-small suit.

Auction #2 doesn't really exist in our style; we tend not to look for slam when only a max with extra shape will suit (i.e. a lot of hands make slam opposite Axxxxx - AKQxx xx but we're not going to take things slow to make allowances for that hand, whereas we would make allowances for Axxxx x AKQx xxx which is a max with more typical shape).

[Kungsgeten]

straube said:

Atul and I have been practicing the IMprecision method for 1M openings. Has anyone else adopted it? I think it's a big advance for Precision relay partnerships.

Are there updated IMPrecision notes available somewhere? The PDF notes I find after some "google-fu" is from bridgewithdan.com and from 2009. There the responses to 1M are "nothing special" IMO, except the continuations after 1H-1S. I've read about the double-barreled invites in a post on this forum somewhere, but I can't remember the details. Also I think there was different solutions discussed in that thread?

It would be easier to follow the discussion if there was some kind of reference to the structure mentioned

straube said:

What are good continuations for 2M-1 and 2N auctions. I think it's....

1H-2D either 3-cd LR or 4-cd LR with shortness

I'm currently playing 2M-1 as limit raise or mixed raise. Over this we play:

2M = To play vs a flat 3 card limit raise. Mixed raise also pass. Responder now uses Romex game trials (relay is short suit game trial, others is long suit game trial). Reponder's 3M rebid is a four card limit raise.
2NT = Max, some slam interest. 3C is min, 3D is extras without shortness, 3H/S/NT is extras and low/mid/high shortness.
New = Natural max, 5+ suit.
3M = To play vs mixed raise, limit raise bids game.
4M = To play.

Opener's most common rebids are 2M and 4M.

However I've been thinking about changing 2M-1 to limit+ raise (including GF hands that doesn't want to use relays). Here's my suggested structure over that:

2M = To play vs a flat 3 card limit raise. Romex game trials (which could also be mild slam tries) as above.
2NT = Anonymous game accept. Not interested in slam vs a limit raise.
...3C = Slam interest, shortness somewhere. 3D asks.
...3D = Slam interest, a minor side suit. 3H asks (3S clubs, 3NT diamonds).
...3M = Slam interest, followed by control bidding.
...3oM = Slam interest, natural.
...3NT = Suggestion to play.
...4M = To play.
3C = Max and shortness somewhere. 3D asks (low/mid/high/voids).
3D = Max and a minor side suit. 3H asks (3S clubs, 3NT diamonds).
3M = NF with 6(+) major.
3oM = Natural max.
3NT = Void oM, min.
4m = Void, min.

As you can see there's some symmetry in the artificial 3C and 3D bids.

We currently play 1M-2NT as GF with 4 card support. However with this new structure we're thinking about using 2NT as "natural" GF to allow for more anonymous bidding instead of using relays. The idea is that 2NT shows:

1. A minimum balanced GF (about 12-15) with 2 or 3 card support. Responder may have four cards in the other major. Responder usually wants to play 3NT, or perhaps 4oM if we have a 4-4 fit, or 4M if opener have 6+ major. If responder has 3 card support he'll have 4333 or possibly 4432/5332 with a very good doubleton (AK or AQ).
2. A minimum GF with 3 card support and 4 cards in the other major. The difference here is that responder doesn't want to suggest 3NT, he just wants to find out if responder have 4 cards in the other major (playing a 4-4 fit instead of 5-3). Responder usually have 4333 or 4432, but could have 5m431 if there's no slam interest.

Over this 2NT response I'm thinking similar continuations as after the 2M-1 raise:

1M-2NT;
3C = Shortness somewhere. 3D asks. This is primarily a way to try to stay out of a bad 3NT.
3D = A minor side suit. 3H asks. Might have slam if there's a good fit, or perhaps we should play 4M or 5m instead of 3NT.
3M = Max with 6+ major, no shortness we want to show. Now 3NT is non-serious, others cue bidding.
3oM = Natural, could be minimum.
3NT = Not 4 cards in the other major. Nothing extra we want to show.
4new = Shortness with 6+ major.
4M = To play.

straube said:

because if opener thinks too long and then rebids 2H the partnership can agree that the 4-cd LR with shortness must show the shortness. Am I right on that? I'm not sure about the laws.

If partner thinks you have UI, but yes you can solve this by having written rules about when responder should bid again. We have an agreement that we pass with a flat three card limit raise. If we have shortness, a help suit, or four card support, responder will bid again. Then you still have UI, but now pass wasn't an option so there's "no problem". It may be problematic though if you have 5332 or perhaps 5431 unsuitable for a short/long game try, but choose to bid again.

[straube]

Kungsgeten, on 2019-October-02, 05:50, said:

I've read about the double-barreled invites in a post on this forum somewhere, but I can't remember the details. Also I think there was different solutions discussed in that thread?

It would be easier to follow the discussion if there was some kind of reference to the structure mentioned

https://www.bridgeba...rreled-invites/

[DinDIP]

Thanks for sharing the fruits of your and Sam’s thinking and testing, Adam. A few questions:

What does a 2♠ response to 1♥ show now? Is it a natural weak invite, with stronger invites going via 1NT? This would mean that 1♥-1♠-any-2♠ is always less than invitational.

I presume the rationale for including min GF hands with 3+M in the “good invite” bid is to minimise information leakage. Does this mean that responder rebids 3M with these hands where possible (e.g. 1♥-2♣-2♦-3♥) so that opener can sign off or show some feature (voids?) when slam is possible? Do you have any special agreement to show such hands when opener shows a shapely max (by rebidding 2♠ or higher)? Can responder distinguish between three- and four-card support?

I understand your decision to treat 1♥-1N-2♣-2♠ as showing 6S because responder can rebid 2♥ when he is 5-2-x-y. The only time you reach 2N when you’d prefer to play 2♠ is when responder is 5-1-x-y and opener is 2-5-3-3 or 2-5-2-4. (I’m presuming opener will rebid 3♣ when 55 if responder bids 2N.)

Was the decision to treat 1♥-1N-2♦-2♠ the same way to maintain consistency or because it was too hard to unravel things if responder could have 5 or 6 spades?

[straube]

I can answer the easier questions.

1H-2S is weak.
1H-1N, 2m-2S is strongly invitational.
1H-1S, 2L-2S is lightly invitational

2M-1 raises are GI with 3+ trump

Minimum GF hands with three trump don’t start with 2M-1. They start 2C for hearts and 2D for spades.

[awm]

DinDIP, on 2019-October-02, 22:28, said:

I presume the rationale for including min GF hands with 3+M in the “good invite” bid is to minimise information leakage. Does this mean that responder rebids 3M with these hands where possible (e.g. 1♥-2♣-2♦-3♥) so that opener can sign off or show some feature (voids?) when slam is possible? Do you have any special agreement to show such hands when opener shows a shapely max (by rebidding 2♠ or higher)? Can responder distinguish between three- and four-card support?
...
Was the decision to treat 1♥-1N-2♦-2♠ the same way to maintain consistency or because it was too hard to unravel things if responder could have 5 or 6 spades?

The sequences with min GF hands are mostly minimizing information leakage where we usually don’t have slam. Basically opener shows shortness over this if max and otherwise just bids 4M. Opener is primarily the one “showing” here so we don’t have a specific way to distinguish three vs four card support.

As for 1H-1N-2D-2S and similar, showing six gets us the 6-1 fits and allows for much better follow ups when we have game values (since we don’t need 2nt as a scramble on 15xy). We can play the occasional 5-2 fit in 2nt (I suspect this is only a small loss whereas the 6-1 fits have a stronger 2s preference).

[DinDIP]

Thanks Adam and Dave. Decided to do some testing. Have a couple of questions of detail and some of theory/practice.

Big picture
How does this structure -- especially the step three responses showing a bunch of strong invite hands -- fare when the opponents compete? I’m reminded of Jeff Rubens’ argument when he advocated Rubensohl that it was better to show shape and clarify strength later than group a number of hands with very different shapes but the same strength.

Would it be better to play jump shifts as strong invites and bunch all the weak jump shifts?

Or play a structure like Dave was experimenting with previously over 1♠ and extend it to 1♥ so:
step one = usually at best a weak invite; could have strong invite if not one-suited (show by rebidding the other major)
step two = relay, either GF or strong invite with the other major
step three = the other major, either 6+ any strength below strong invite or 5+ weak invite

(One option would be to play two-below transfers to cater for weak and strong invites. For example, after 1♥
2♠ = clubs, weak invite or strong invite
2N = diamonds, weak invite or strong invite
3♣ = mixed raise

Over 1♠, there isn't the same room but you could play
2N = diamonds, weak invite or strong invite
3♣ = strong invite
3♦ = mixed raise)

Detail
1. 1♥-1♠-1N-2♣-2♦-?
Does responder have any way to enquire about opener’s strength and/or shape? I’m guessing that 2N = a weak invite with exactly 4S. Do 3m = 4S and 6m, weak invite? Or are the latter so rare that you use 3m as some sort of trial bid with 5+S (which, on first glance, looks to have MUCH greater frequency but leaves responder stuck when he is 46)?

Is the same true if opener rebids 2♥, showing 3S and 6H? Or are 3m ambiguous as to which major responder

2. I assume that if the opponents interfere in a relay auction you treat the auction as GF unless relayer can break to two of the other major. Or do you have some other rule?

[straube]

There were a few things I didn't like about my structure.

1)I didn't like missing games after a semiforcing NT. They were infrequent, but it's embarrassing to be in 1N with 26 hcps.

2) I didn't like my continuations after 1S-2D (showing hearts). 2D required a very wide range because I needed 1S-1N, 2m-2H to be artificial.

3) I didn't like getting to 3m with 4/4 fits when a 5-2 major suit fit was available. Adam pointed this out in the previous thread.

I think it works better to separate responses into three strengths. Opener's rebids are structured to be more conservative the weaker the responding hand.

[nullve]

straube, on 2019-September-29, 08:00, said:

RDB is so much better than anything else I've seen that I think of it as having "solved" the limited 5-major opening.

Aren't constructive weak 2M openings also part of the "solution" in an IMPrecision context? If so, are you able to open anything on, say, KQTxxx xxx Qxx x? (I think most bridge players would hate not being able to do that.)

[straube]

Yes, the upper range of 2M is higher. I think Adam addressed the range somewhere. It’s not a lot different from standard, maybe 6-11 but the heavy invitational hands (good 12 to bad 14) are mostly designed to avoid game opposite light openers. For example , I think Adam and Sieong open as light as rule of 18 with possibly as few as 4 queen points. Adam has listed example minimums if you can find them.

The way I think of it, I’m not opening a 5 or 6 point 2M with mostly quacks outside the suit, but I’d open that example hand where the suit has the strength. On the high end, I’d open most 6/4 11s at the 1-level.

Yeah, “solved” is pretty strong wordage but it’s how I think of it. And before Adam posted this method, I’d come to think that the range of 2M should be slightly higher such that a rebid of 2M after a forcing NT showed something more than that sixth card. In other words, it required something I thought was necessary anyway.

What do you think of the method?

[nullve]

straube, on 2019-October-05, 05:41, said:

What do you think of the method?

I see some problems with it, but then I may not fully understand it.

Anyway, how would you bid the following hands

1) W: Axxxx QTxx Kxx x, E: x Axx QJxxx Axxx
2) W: Axxxx QTx Kxxx x, E: x Axxxx QJx AJxx
3) W: AQxxx KQTx Kxx x, E: xx Axxx QJx Jxxx

using Relays with Double-Barreled Invites?

My guess, based on what I've read:

1) 1♠-1N; 2♥-2N; P
2) 1♠-2♣; 2♥-2N; P
3) 1♠-P

[straube]

The first two look right. The last one would go 1S-1N, 2H because responder has 8 hcp support with doubleton support. I guess you could downgrade it.