[rbforster]

straube, on 2012-May-06, 10:31, said:

I'm less concerned about getting too high. I'm more concerned about splitting opener's NT ranges. In one example, opener had 25 balanced. His systemic rebid was 2N, but partner had a 5/5 hand in the majors and slam was on. I'm sure continuations could be invented, but a 2N rebid as 20+ even GF is a problem.

Why don't you use one of your low level rebids by opener as GFR? It seems a waste to have limited responder to sound SP+ values if you don't have a way for opener to relay the SP hands. NT hands especially are good for this - why would you kill your slam methods by jumping to 2N just like the standard players?

[straube]

rbforster, on 2012-May-07, 05:37, said:

Why don't you use one of your low level rebids by opener as GFR? It seems a waste to have limited responder to sound SP+ values if you don't have a way for opener to relay the SP hands. NT hands especially are good for this - why would you kill your slam methods by jumping to 2N just like the standard players?

Yeah, the jump to 2N is unworkable. OTOH, reserving a low level bid for GFR kills the whole idea of being able to reverse relay opener's hand.

[hrothgar]

straube, on 2012-May-07, 08:41, said:

Reserving a low level bid for GFR kills the whole idea of being able to reverse relay opener's hand.

This might seem like a stupid question but:

The semi positive response has already limited strength and (typically) shown an unbalanced hand.
Just what do you hope to accomplish with an option to reverse relay?

Opposite a semi positive your absolute priorities are

1. Exploring strain in GF auctions (which means step = relay)
2. Getting out low in a misfit auction (which means that most other responses are typically natural, non forcing)

[straube]

hrothgar, on 2012-May-07, 09:03, said:

This might seem like a stupid question but:

The semi positive response has already limited strength and (typically) shown an unbalanced hand.
Just what do you hope to accomplish with an option to reverse relay?

Opposite a semi positive your absolute priorities are

1. Exploring strain in GF auctions (which means step = relay)
2. Getting out low in a misfit auction (which means that most other responses are typically natural, non forcing)

Well, the idea was that 1D would be either any semipositive or a GF balanced hand. Both of these facilitate opener making a natural rebid. After this rebid, responder can relay with the GF balanced and with certain of the semipositives (those having a fit for example).

I've looked at hands on BBO and it works better than might appear at first glance. The biggest problem is when opener has a balanced hand of 20+ points. I think that's probably the death knell of this idea, but perhaps there's a work around.

[hrothgar]

straube, on 2012-May-07, 10:03, said:

Well, the idea was that 1D would be either any semipositive or a GF balanced hand. Both of these facilitate opener making a natural rebid. After this rebid, responder can relay with the GF balanced and with certain of the semipositives (those having a fit for example).

I've looked at hands on BBO and it works better than might appear at first glance. The biggest problem is when opener has a balanced hand of 20+ points. I think that's probably the death knell of this idea, but perhaps there's a work around.

Guess we have a different in terminology

In general, when I use the expression "reverse relay" it means a scheme in which the exact same response scheme is being used.
For example, playing MOSCITO

1C - 1D 1C= strong, 1D = GF
1H - 1S+ 1H = relay, 1S+ = relay responses

1C - 1D 1C= strong, 1D = GF
1S+ 1S+ = reverse relay using the precise same response schedule

I'd describe your treatment as a relay break

[straube]

Yeah, sounds like we use the word differently. I use it when the strong unlimited hand relays pattern. For example, our basic relays are..

1C-
.....1D-GF, major or bal
..........1H-S or bal
...............1S-H/C or bal
....................1N-bal
.........................2C-bal, M
.........................2D-bal, no M
.........................etc-H/C
....................2C-S/C
....................2D-S
....................etc-S/D
..........1S-H
..........1N-H/S
..........2C-C/D
..........2D-C
.....1N-H/S
.....2C-C/D
.....2D-C
.....etc

So even here, we're reverse relaying (the way I use the term) but we don't use exactly the same structure to do so.

[awm]

I don't really understand this thread. There are two situations:

1. Not in competition. There isn't a huge difference between reasonable relay schemes. Perhaps you can show otherwise, but there are only so many ways to carve up space. The big gains will cone from well designed relay breaks. Anyway, the one type I'd be concerned about here is double negative because space is at a premium, and 1s is bad for those.

2. Competition. Here the goal is to have responder give useful shape info at first response as much as possible, as this will help more than a point range when 4th hand bids. There should be some bias to hands where competition is most common/dangerous. This last is the semi-positives; generally double negative responder won't have much to say, and a positive means opponents have less (so less likely to bid) and at least gets us to a forcing auction.

The suggestion here seems to lose on both counts...

[straube]

awm, on 2012-May-07, 12:08, said:

I don't really understand this thread. There are two situations:

1. Not in competition. There isn't a huge difference between reasonable relay schemes. Perhaps you can show otherwise, but there are only so many ways to carve up space. The big gains will cone from well designed relay breaks. Anyway, the one type I'd be concerned about here is double negative because space is at a premium, and 1s is bad for those.

2. Competition. Here the goal is to have responder give useful shape info at first response as much as possible, as this will help more than a point range when 4th hand bids. There should be some bias to hands where competition is most common/dangerous. This last is the semi-positives; generally double negative responder won't have much to say, and a positive means opponents have less (so less likely to bid) and at least gets us to a forcing auction.

The suggestion here seems to lose on both counts...

Thanks for the feedback.

I don't think this design will pan out. Still looking at it.

As far as design goals, I think it's very useful to know immediately whether responder is broke or not before RHO intervenes. For instance, we play Rubensohl after our 1H semipositive response....

1C P 1H (2S)
3H shows 4 hearts and is GF

Opener may venture this because he knows responder has something.

1C P 1H (2S)
dbl says takeout shape and promises no extra.

These things mean we are better placed than after 1C-1D (0-7).

I don't mind so much our 1C-1S (DN) auctions. Again, we're better placed in competition than 1C-1D (0-7) auctions. In uncontested auctions, we've buried our major suit fits, but we've buried the opponents' fits as well and are likely to have only a small point majority on them.

What sucks for us is our 1C-1H uncontested auctions when we only are in part score/game range and opener has an unbalanced pattern.

There's a lot of upside, however, to 1C-1H as semipositive any. For one, opener only relays when he wants to...not has to. He can choose (instead) to rebid 1N (17-18) or Stayman or transfer, etc. Opener can try to sign off, invite, or GF relay. Opener is very well positioned with balanced hands. It's the unbalanced hands that are problematic.

So this thread was aimed to see if we could have opener start to describe his hand naturally at a lower level. The biggest snag so far that I've encountered is when opener has a big balanced hand.

[awm]

straube, on 2012-May-07, 13:12, said:

As far as design goals, I think it's very useful to know immediately whether responder is broke or not before RHO intervenes. For instance, we play Rubensohl after our 1H semipositive response....

1C P 1H (2S)
3H shows 4 hearts and is GF

Opener may venture this because he knows responder has something.

This very much differs from my experience. I have not found that there are many hands where I want to bid over 1C-Pass-?-2S if partner has any 5-7 but want to pass if he has any 0-4. Given the strong club opening the points are normally evenly divided (at worst) and it's generally good to compete for the partial if opener has shape or extras regardless of where responder lives in the point spectrum. It's not like 5-7 puts us in a forcing pass or anything, and the one trick difference between 6 points and 3 points is a one trick difference for opponents too... not to mention that it's often hard for opponents to penalize in these auctions even in the rare case that they have us dead to rights.

I also cannot imagine wanting to bid 3H here to show "four hearts" if the 1H bid is any 5-7. What do I do with five hearts? Haven't I just taken up even more of our bidding space? The vast majority of hands with "four hearts" I'd prefer to either pass or double for takeout... and this is true regardless of whether partner has 5-7 or 0-4.

[straube]

If you have 5 hearts then 1C P 1H (2S) 3D

I might want to show 4 hearts if I had

Kxx AQxx Ax AKQx

3H allows partner to bid 3S to deny a spade stopper...and then I rebid 3N.

If it goes 1C P 1D (2S) and partner can have 0-7, don't I need 2N to show an invitational NT? If so, I no longer have Rubensohl. So now my double has to handle the very strong hands as well as the takeout shape hands. The double is over-loaded.

I want to fight for the part score, too. I can do that with Rubensohl (opposite 5-7). I can also do that opposite 0-4. Our double (which now handles takeout as well as very big) will usually just be takeout because it's pretty hard to force game opposite 0-4. Also, it's really true that in many instances, opener may decline to fight for the part score opposite 0-4 if lacking extra values or shape.

Also, opener may pass 2S and the 0-4 hand can now act if he holds a maximum with shape...without fear that opener will take him for too many hcps.

[straube]

This discussion has got me thinking that maybe we would do better playing Rubensohl after 1C P 1S (2S) in which 1S is a DN.

This is consistent with what we're doing after 1C P 1H (2S).

Takeout doubles are takeout shape, etc.

1C P 1S (2S)
dbl P 3H P
3S

would show takeout shape, probably only three hearts and deny a stopper. This is just far ahead of

1C P 1D (2S)
dbl P 3H P
3S

which could be all sorts of hands.

1C P 1H/1S (2S)
3H

showing GF with 4 hearts is much more clear than a double that can be takeout or have other strong meanings.