[straube]

I've seen ideas for aborting full pattern resolution for balanced hands in favor of QPs and rbforester has worked on structures that allow for asker to choose whether to relay for full shape first or only for the majors.

It's nice to know the full pattern. We've been able to choose 5m over 3N because we've learned of a big minor suit fit. However, there's a downside in disclosing RR's full pattern in that it may be more useful to the defense than to our side (on opening lead or when RR declares).

I always like to discover a major suit fit, not only because it may be the best game but because finding that fit means that I have a few more steps available for relaying than I would have for 3N. After determining whether we have that major suit fit, I'm usually starting to wonder if partner has extra values which might also justify moving past 3N.

Rbforester has a nice scheme but our structure includes the 5332s in balanced.

How about...

2D-no 4cd or 5cd major
........2S-2/3
........2N-3/3
........3C-3/2
2H-4 or 5 hearts
........2N-5 hearts
................3D-2 hearts
................3H-3 hearts
........3C-2 spades
........3D-3 spades
2S-4 spades, not 4 hearts
........3C-2 hearts
........3D-3 hearts
2N-5 spades
........3D-2 hearts
........3H-3 hearts
3C-4/4 majors

This scheme is based on the Moscito structure. After learning the major suit pattern we'd relay for QPs, then the minor suit pattern, and then DCB.

I feel it could be improved, especially because the 2D and 2S rebid is wasted.

The idea of partial relays holds particular interest to me because our RR only relays balanced shapes out when the relay captain is also balanced.

[olien]

Another possible solution, instead of showing major suit shape after denying a 4+ card major is to just show strength before even disclosing any distribution.

An idea might be after 2D denying a 4-card major:

2H strength ask
.......2S minimum
.......2N medium
.......3C+ maximum, and showing distribution


Now 2S, and 3C+ are symmetrical. Can divide into 3 ranges based on QPs. Maybe 5-6, 7-8, and 9+. Obviously, when responder is minimum, can't show FULL distribution below 3N, and it may behoove you to reverse 2NT and 3C+ immediately, and play 2N showing maximum as forcing to 4NT so 3NT can become a relay.

So, here's a possible structure:

2S minimum
.......2NT relay
.............3C 3334 or 5C(332) (3D relays then: 3H=3334, 3S+=5C-332)
.............3D 5D (3H relays for distribution)
.............3H 3343
.............3S 2344
.............3N 3244
2N maximum
.......3C relay
..............3D 3334 or 5C(332) (3H relays then 3S=3334, 3N+=5C(332))
..............3H 5D(332)
..............3S 3343
..............3N 2344
..............4C+ 3244
3C+ medium, and same steps as after 2S-2NT


This may require relay breaks to find a 5-3 major suit fit, but should work OK. To re-iterate, I just came up with this idea off of the top of my head. Also, if responder has not bid NT already (I'm not familiar with your initial response structure), this avoids wrong-siding the NT when responder is minimum. It may also be worthwhile to re-arrange the responses after 2N-3C, but I think keeping it in the same order may save a lot on memory strain

[hrothgar]

olien, on Sep 22 2010, 05:10 PM, said:

Another possible solution, instead of showing major suit shape after denying a 4+ card major is to just show strength before even disclosing any distribution.

A comment and a question:

The comment: If you are that desperate for information about range why not use a low level relay break as a range ask? This would allow you to preserve a simple symmetric framework for those occasions that you don't want to use the relay break.

The question:

Do you have any real experience playing a relay system? If not, you really might want to spend some time playing a well designed relay system before you trying to engineer your own.

Either Symmetric Relay or Numeric Relay provide a pretty easy entry point.

[olien]

I do have real experience playing relay systems. I've been playing them for 3-4 years now. That being said, I was just trying to help straube come up with a solution for what he is seeking. That doesn't mean I think it's good or bad, and I put in minimal effort to reply.

BTW, why not make any suggestions of your own instead of question or criticizing other people's ideas? Also, have you ever tried playing a structure like this, where range comes before shape? Or are you just running your mouth at other ideas that you haven't even tried yet?

[hrothgar]

olien, on Sep 22 2010, 05:39 PM, said:

BTW, why not make any suggestions of your own instead of question or criticizing other people's ideas? Also, have you ever tried playing a structure like this, where range comes before shape? Or are you just running your mouth at other ideas that you haven't even tried yet?

I've played a wide variety of different relay methods. Some of which, such as Ultimate Club resolve shape before range. I find these systems much more difficult to learn which is why I've been recommending starting with either symmetric or numeric. (There is a reason that Ultimate Club drifted off into oblivion and it had NOTHING to do with efficiency)

Three comments

1. I have posted a variety of different relay structures in these threads.

2. The basic approach that you are suggestion - adding layer after layer of special cases into your core relay structure - is fundamentally flawed. Keep the core structure as simple as possible. If you need to support some special case like a range ask use a relay break.

3. The systems that you are proposing don't show range before shape. They are a random mishmash of range and shape asks. Sometimes you're showing range, sometimes you're showing shape. You might be able to design something that's efficient, but I'd be shocked if you can come up with anything that humans can remember.

[akhare]

Edit: Deleted....

[straube]

I'm not sure this idea is a good one. I'm certainly concerned about efficiency. For example, by the time I learn QPs and full pattern am I at the same point or have I used up more room? It ought to be the same, but may not be if I haven't arranged things properly.

I see certain advantages for specific relay breaks. For instance, if I'm captain and holding a distributional hand (which I wouldn't in our case), I may want to have an ask for controls or just aces as opposed to QPs.

Generally speaking, I'm not fond of relay breaks because they cost space and are another thing to remember. I would prefer to know partner's specific major suit holdings (2/3 vs 3/3 vs 3/2) just in case opener has a 5 cd major rather than have to relay break or venture into the 4-level to find this out.

I'm still interested in gut reactions to this idea. I think most of us would vote against learning strength (QPs) before pattern, but would we rather learn QPs or minor suit holdings first? What if the condition was (as it is for us) that both hands are known to be balanced?

[hrothgar]

straube, on Sep 22 2010, 06:53 PM, said:

I think most of us would vote against learning strength (QPs) before pattern, but would we rather learn QPs or minor suit holdings first?  What if the condition was (as it is for us) that both hands are known to be balanced?

I'm not quite sure what you mean by "Pattern". (To me, pattern describes the complete shape of the hand, which includes minor suit holdings)

From my perspective, if you know that you have a balanced hand opposite a balanced hand

Your relay sequences should relay out the complete shape and then provide options to either

1. Ask for slam points
2. Set trump and start Keycard

Your relay breaks should be used to transition over to natural bidding.

If you have two balanced hands opposite one another one of your primary concerns has to be making a well informed decision regarding 3NT as a plausible contract. Natural bidding is going to be MUCH better at asking for/showing stoppers and the like.

From my perspective, this is MUCH more important than throwing in a low level range ask.

I'll note in passing: If you already know that you have a balanced hand opposite a balanced hand, you can use a relay break to incorporate a "free" range ask.

Assume the following auction

1♣ - 1♦
1♥ - 1N

where

1♣ = Strong
1♦ = GF
1♥ = Prefer's to ask rather than tell
1N = Balanced or 4441

At this point in time, the Strong club opener can either

Bid 2♣ as a relay and ask for shape OR
Bid 2♦+ and SHOW his shape using the exact same reponse structure that gets used over the 2♣ relay.

There are LOTS of easy ways that you can fold these types of breaks into your system that help avoid the need to show range during a primary relay sequence.

Now, lets assume that the auction goes

1♣ - 1♦
1♥ - 1N
2♦

where 2♦ limits the strong club opener's strength, promises 4-5 hearts, and denies 3=4=3=3 shape.

At this point in time, responder - who is now the relay captain - can either bid 2♥ as a relay if he wants to investigate slam OR
bid 2♠+ to transition to more natural bidding.

[straube]

Just curious, but in your sequence 1C-1D(GF), 1H(prefers to ask)-1N (bal including 4441s), 2D (prefers to show) it seems like the strong club hand is cross-purposed, first wanting to be captain and then RR. Or is that 2D bid just a break into natural bidding?

I'd forgotten about passed hand situations. I'd definitely prefer knowing the pattern before QPs opposite PH and I wouldn't like having to think about different rules for PH vs UPH for an improvement (if it is one) of such marginal value.

Thanks for your comments.

[rbforster]

straube, on Sep 22 2010, 03:28 AM, said:

I always like to discover a major suit fit, not only because it may be the best game but because finding that fit means that I have a few more steps available for relaying than I would have for 3N.  After determining whether we have that major suit fit, I'm usually starting to wonder if partner has extra values which might also justify moving past 3N.

Rbforester has a nice scheme but our structure includes the 5332s in balanced.

How about...

2D-no 4cd or 5cd major
2H-4 or 5 hearts
2S-4 spades, not 4 hearts
2N-5 spades
3C-4/4 majors

This scheme is based on the Moscito structure.  After learning the major suit pattern we'd relay for QPs, then the minor suit pattern, and then DCB.

A few comments. My scheme shows 4432/4333 balanced hands with 2D+, so if you're trying to squeeze in 12 more shapes (5332s), you'll be short on space. That's fine if it's how you want it, but balanced hands are very common so it might make sense to have a 2C+ response scheme instead. Maybe 1C-1S(bal)-1N(relay)...

Also, if you're strapped for space, you might want to prioritize resolving the minor-heavy hands (short majors) first and let the major-heavy hands resolve higher. This will hopefully give you space to ask for the major stoppers below 3N (in the minor case). Then, in the major case, hopefully you find a major fit so you've got extra space up to 4M instead of just 3N to figure out strength etc.

[olien]

Well, at the risk of not receiving hrothgar's approval, I have a few things to say about my structure regarding relay breaks.

after 2D (balanced)-2H:
2S minimum
......2N relays (then aforementioned structure)
......3C 5+H (now 3D=no fit, 3H=fit...doesn't matter who declares since both M's bid)
......3D 5+S (now 3H=no fit, 3S=fit)
......3M can either play as shortness 5/4+ minors, or showing weakness (you choose one or the other)


Also, if playing high->low shortness showing, my structure provides few problems.

After 3D showing 5D(332), 3H relays, now 3S=2353, 3N=3253, 4C=3352, and you don't zoom past 4C so opener can sign off in 4M. It also works after 3C showing 3334 or (332)5, 3D relays then 3H=3334, 3S=2335, 3N=3235, 4C=3325.

Another thing to be said is if opener has a single suited M hand, and bids 3C/D showing H/S respectively, and responder denies a fit, can now bid 3M to set M for cue-bidding.

This may not be the easiest structure to remember, but doesn't seem too difficult. And anybody that's playing a relay structure shouldn't be worried about putting in a little effort. I love how hrothgar asks me to make a structure that a human can remember but then alludes to how much he likes ultimate club. And no, my structure is not a random mish-mash. Straube asked about a specific auction, and I made a suggestion. All of the responses are showing range (I just said minimum, medium, maximum because each partnership has it's own definitions). Also, 3C and higher are zooming, and the reason for making the order minimum-maximum-medium is so that responder doesn't zoom past 3N at any point UNLESS he is maximum or has given a good definition of his hand (in terms of shape/strength) and opener wants to know more.

[hrothgar]

straube, on Sep 22 2010, 08:00 PM, said:

Just curious, but in your sequence 1C-1D(GF), 1H(prefers to ask)-1N (bal including 4441s), 2D (prefers to show) it seems like the strong club hand is cross-purposed, first wanting to be captain and then RR. Or is that 2D bid just a break into natural bidding?

Consider the sequence

1♣ - 1♦

1♣ is a strong club opening
1♦ shows a GF (and denies any 5440 shape or a solid 7+ card suit)

At this point in time, the strong club opener has the choice to ask or to show.

In general, the strong club opener will prefer to show with a minimum strength unbalanced hand.

The strong club opener will generally prefer to ask with a balanced hand, any max, or a 5440.

Lets assume that the auction starts

1♣ - 1♦
1♥ - 1NT

the strong club opener has a balanced hand (and therefore preferred to ask reather than show).

Responder has just shown a balanced hand, therefore, there is no advantage to having the balanced 1♣ opener ask. Therefore, the strong club opener might as well make a relay break, bid 2♦ or higher, and SHOW his shape using the exact same response scheme that responder would have used had the strong club opener relayed with 2♣.

This relay break, like the one available after 1♣ - 1♦, is primarily used to limit the strength of the strong club opener.

As I said before... If you use your relay breaks correctly you can do a really nice job limiting strength with the need to explicitly build QP shows into your relay scheme.

[hrothgar]

olien, on Sep 22 2010, 09:47 PM, said:

I love how hrothgar asks me to make a structure that a human can remember but then alludes to how much he likes ultimate club.

I believe that my comment about Ulitimate Club was

1. The relay structure is very efficient at conveying information
2. The structure is extremely complicate and, therefore, fell by the wayside

I was using Ultimate Club as an example of a flawed system whose complexity preventing from enjoying any widespread adoption.

[olien]

Hrothgar, I don't know what Straube's structure over 1C is. But you are assuming that he is playing 1C-1D as an artificial GF, which not everybody does *gasp* If he's not playing this, then maybe there's a reason for him wanting to re-order the way information is given from shape->strength, but from shape->strength->shape.

[hrothgar]

olien, on Sep 22 2010, 09:56 PM, said:

Hrothgar, I don't know what Straube's structure over 1C is.  But you are assuming that he is playing 1C-1D as an artificial GF, which not everybody does *gasp*  If he's not playing this, then maybe there's a reason for him wanting to re-order the way information is given from shape->strength, but from shape->strength->shape.

Straube noted that he had a relay sequence in which he knew that one balanced hand was asking about the shape of another balanced hand.

I provided an example where one balanced hand was asking about the shape of another balanced hand. I noted that this type of sequence provided the opportunity to employ a reverse relay.

True, I did not specifically state this could be generalized to Straube's auction where one balanced hand was asking about the shape of another balanced hand. However, I made the assumption that people participating in this thread have basic reading skills and are able to draw simple inferences.

In your case, it seems clear that said assumption was unwarranted.
I apologize if I have drawn attention to any particular learning disabilities.

[olien]

Hmm, well, 3C=5+H and 3D=5+S...Opener can still be balanced...Can then play 3M as showing weakness either in that major or the other major and looking for the best game. OMG...Opener doesn't have to be unbalanced to want to break the relay. This means that my breaks are completely useless.

Can even combine bids:
3C=whatever you want
3D=a 5-card major (now 3H=3S, 3S=3H, 3N=3/3 majors)
3M=weakness other major

But, what do I know...I've never, ever played bridge before, and don't know what I'm talking about because I apparently have a learning disability. Like I said:

"I put minimal effort into responding"

I was in class at the time (yes, believe it or not I do go to a university), and had to pay attention to the professor and take notes from what I was learning while I was posting...I didn't have time to read closely, and between the lines. I just read the general problem and replied. So hrothgar, I will quote a good friend of mine and tell you to bite me.

[hrothgar]

Quote

Hmm, well, 3C=5+H and 3D=5+S...Opener can still be balanced...Can then play 3M as showing weakness either in that major or the other major and looking for the best game. OMG...Opener doesn't have to be unbalanced to want to break the relay. This means that my breaks are completely useless.

Can even combine bids:
3C=whatever you want
3D=a 5-card major (now 3H=3S, 3S=3H, 3N=3/3 majors)
3M=weakness other major

But, what do I know...I've never, ever played bridge before, and don't know what I'm talking about because I apparently have a learning disability. Like I said:

"I put minimal effort into responding"

I was in class at the time (yes, believe it or not I do go to a university), and had to pay attention to the professor and take notes from what I was learning while I was posting...I didn't have time to read closely, and between the lines. I just read the general problem and replied. So hrothgar, I will quote a good friend of mine and tell you to bite me.

I guess we should just sit back and be grateful that you can spare the time to post.

Maybe, at some point in time in the distant future, we'll be lucky enough to see you both think and post.

[straube]

hrothgar, on Sep 22 2010, 01:51 PM, said:

straube, on Sep 22 2010, 08:00 PM, said:

Just curious, but in your sequence 1C-1D(GF), 1H(prefers to ask)-1N (bal including 4441s), 2D (prefers to show) it seems like the strong club hand is cross-purposed, first wanting to be captain and then RR.  Or is that 2D bid just a break into natural bidding?

Consider the sequence

1♣ - 1♦

1♣ is a strong club opening
1♦ shows a GF (and denies any 5440 shape or a solid 7+ card suit)

At this point in time, the strong club opener has the choice to ask or to show.

In general, the strong club opener will prefer to show with a minimum strength unbalanced hand.

The strong club opener will generally prefer to ask with a balanced hand, any max, or a 5440.

Lets assume that the auction starts

1♣ - 1♦
1♥ - 1NT

the strong club opener has a balanced hand (and therefore preferred to ask reather than show).

Responder has just shown a balanced hand, therefore, there is no advantage to having the balanced 1♣ opener ask. Therefore, the strong club opener might as well make a relay break, bid 2♦ or higher, and SHOW his shape using the exact same response scheme that responder would have used had the strong club opener relayed with 2♣.

This relay break, like the one available after 1♣ - 1♦, is primarily used to limit the strength of the strong club opener.

As I said before... If you use your relay breaks correctly you can do a really nice job limiting strength with the need to explicitly build QP shows into your relay scheme.

It seems like one could have a more helpful continuation after your 1C-1D, 1N sequence. 2C should ask, but the 2D on up bids could be used to show something other than balanced hand patterns. More often than not, with 2 balanced hands one might as well have the stronger hand do the asking.

In Abraham's structure we get to a similar point...

1C-1D, 1H-1S shows that responder is either balanced or has H/C.

opener's rebids are...
1N-relay
.....2C-H/C
.....other-bal
2C-S/C
2D-S
2H+ S/D

[hrothgar]

straube, on Sep 23 2010, 12:35 AM, said:

It seems like one could have a more helpful continuation after your 1C-1D, 1N sequence. 2C should ask, but the 2D on up bids could be used to show something other than balanced hand patterns. More often than not, with 2 balanced hands one might as well have the stronger hand do the asking.

Weren't you the one complaining that you don't want to use relay breaks because they're too difficult to remember?

The system that I am describing might not be the world's most efficient way to utilize space. However, preserving parallelism has significant advantages with respect to memory load.

Regardless, you're welcome to do whatever you damn well please...

[straube]

I like the ability to reverse relay and had never really thought of them as relay breaks, but maybe they are.

I was just trying to point out an advantage I thought I saw in Abraham's method over Moscito and I didn't mean to offend you.

I'd never really liked Moscito's 1C-1D, 1N because it used up so much room. I'd thought to try to preserve 2D+ continuations so that responder could even then pattern out some of his hands, but he can show far less than after 1C-1D, 1H.

After 1C-1D, 1H-1N then I don't understand which of opener's hands he would want to show. If opener had a minimum balanced hand, wouldn't he have rebid 1N immediately? 1C-1D, 1N?

If I remember Moscito correctly, opener can't show the 3-suited hands and has to bid 1C-1D, 1H. If now 1C-1D, 1H-1N and responder is showing a balanced hand, then perhaps

2C-relays
other-5440s