[rbforster]

Most relay precision systems these days have relay auctions that go like this:

1♣*-[GF response with a known suit]
relay-[more about shape]
relay-[more about shape]
relay-[exact shape]

This is all well and good, but a more complete system should assign other meanings to opener's non-relay bids, allowing opener to ask other questions (besides "tell me more about your shape") or to describe opener's hand instead. For example, in the system I play, once you get responder's exact shape, most of the relay breaks have an assigned meaning -

...
relay-[exact shape shown at 3♠ or less]
3N = to play
4♦ = signoff, relay to 4♥
(and then excluding the above)
step1 = asking for controls/strength (then denial cuebids to locate honors)
step2 = RKC in responder's long suit (and then Q ask and/or control asking bids in side suits)
step3-5 = RKC in other suits, ordered by responder's length
step6+ undefined

I've heard about people using a "reverse relay" to describe opener's shape instead of responder's at the first relay break. I've heard about people wanting to use relay breaks prior to exact shape to check for stoppers. Various other things have been suggested, and to the extent you define these bids as anything useful, your system is better than if you always just relay.

Does anyone play a systematic set of relay breaks, either on opener's first rebid or on any of his subsequent rebids prior to getting exact shape? Do you have any simple rules you use to give meanings to these non-relay bids?

[Echognome]

I played relay breaks as "natural, anticipating a problem." I kept this meaning for the first and second break. An example of breaking after the first positive:

Q xx AQxx AKQxxx

1♣ - 1♠ (or whatever spade positive you have)

I'd bid 2♣ here. If you relay out partner's shape and hear he is, e.g. 5=3=3=2, you will have no idea if you should be in 3NT, 4♠, 5♣, etc.

An example of breaking after the second shape bid of partner might be:

Kx AKxx xx AQJxx

1♣ - 1♠ (Again, whatever your spade positive is)
1NT - 2♣ (suppose this shows 4+ clubs)

Here I'd bid 3♣. You relayed in case partner had hearts (maybe in your system partner doesn't have hearts). However, now you anticipate a problem. Suppose partner shows a 5=1=3=4. Should we be in 3NT, 4♠, 5♣, etc?

So the idea is to keep things natural (so it's easy to remember), but you are definitely showing something about your hand and anticipating a problem in the auction. You can still investigate slam. Most of the time I'm breaking it's with the 16-18 variety hand.

[Mbodell]

Yes. I play relay breaks that show the key card auctions like you do when you have complete shape.

I play relay breaks earlier in the auction that show a minimum 1♣ opening and useless shortness in partners suit. Or, in a balanced auction, a balanced minimum. The unbalanced relay breaks are rare and basically have the 1♣ person show their shape in response, and since you already know their shortest suit (the suit that you showed) once you get their suit(s) you can jump right to pattern. We also, in theory, break after the second suit of a 2 suiter to show the minimum with 0 or 1 in the second suit and 3 or less in the first suit.

Generally one must check carefully that the shapely relay break will be able to pattern out at a low enough shape to be useful. Also, the relay breaker needs to have 4 controls since our eventual control asks will have the first step be 4 so with less than that you can't break. We've tried to start pausing for 10 seconds (sort of like a stop bid) before starting our relays in most situations so that if some one pauses to think about a relay break it isn't giving UI.

The balanced breaks are much more common, but still pretty rare. Often once the person breaks we just bid 3nt since usually there will be no slam to investigate so I don't get enough practice with these and may get them wrong (I'm just starting to work on my partner and my bidding notes, and relay breaks are something I'm looking forward to having documented since they come up so rarely). In our 1♣ system a 1♠ response shows a balanced gf positive or both red suits. Relay bid is 1nt. The 2♣ answer is the both red suit relay break showing both black suits and minimum hand strength. 2♦ shows a balanced hand with 4 or more hearts. 2♥ shows a balanced hand with 4 or more ♠ and fewer than 4 hearts. 2♠ shows some 4333. 2nt shows a 5 card suit (diamonds? if directly). 3♣ shows both minors 44. 3♦ shows 3235. 3♥ shows 2335. 3♠ shows 3325. After the cheap bids a bid of 2♠ shows spades, 2nt shows a 5 card suit, 3♣ shows diamonds, 3♦ and above shows submarine shortness (5332 if 2nt was bid, 4432 with clubs implied if not).

We briefly experimented with bids after balanced breaks showing stoppers for nt but I didn't like that much.

[Echognome]

Oh. I didn't mention what we do when complete shape is know. We played almost the opposite of standard.

Shape resolved at 3♥ or lower, then:

Step 1 = Weak relay (bid next step unless you have 13+hcp and 4+controls)
Step 2 = Strong relay (tell me your controls)
Step 3 up to 4♥ = RKCB for 1st, 2nd, 3rd, 4th suits (Longest, Highest sorting)
3NT = To play (so steps skip the 3NT bid)
4M or 5m = To Play (note that we do not use the 4♦ termination)
4NT/5M/6m = Rare, but more specific invites (can't remember now what we used)

Shape resolved at 3♠ or higher, then:
Step 1 = Strong relay
Step 2 = Weak relay (if room available)
Others as above

[PrecisionL]

In Ultra Club: After two or three relays, a bid at the 3-level of responder's major or at the 4-level of responder's minor is Beta (for A & K Controls). CAB or Turbo often follow Beta, and also serious 3nt. Opener can break relay at any time to ask support for one of his suits with the following responses: 1st step = 0-1 cards, 2nd step = 2 cards, 3rd step = 3+ cards and 1-2 controls, zooming to total responder controls. Opener can also break relay to start Beta earlier without complete distribution if so desired.

Larry

[hrothgar]

The optimal use of a relay break is highly context specific. I recommend that you try to reframe the question and focus on a few key sequences that you find especially interesting.

Here are three examples that you might find interesting:

1. Playing MOSCITO, the auction 1♣ - 1♦ shows a game forcing hand and denies a 5440 shape or a solid 7+ card suit. The strong club opener has the option to bid 1♥ to ask for shape or bid 1♠+ to show shape. If the strong club opener breaks the relay and shows rather than asks he is limiting his hand.

The purpose of the reverse relay is to limit the strength of relay responder.

2. Playing MOSCITO, the auction

1♣ - 1♦
2♣

shows a minimum opening. Opener either has a single suited with Clubs or a two suiter with Spades and Clubs (Either way, he has clubs).

If responder bids 2♦ he asks for shape. If responder bids 2♥+ he is showing shape and explictly promises a singleton or void in Clubs.

The purpose of the reverse relay is to quickly identify misfitting hands and avoid bypassing 3NT

3. Playing MOSCITO, the auction

1♣ - 1♦
2♣ - 2♦
2♥

shows a single suited hand with Clubs

If the relay captain bids 2♠ he asks for shape. If the relay captain bid's 2NT+ he is transitioning to natural bidding and exploring for stoppers / half stoppers in side suits.

[Cascade]

At higher levels I like the puppet terminator concept.

After shape has been resolved:

4♦ is a puppet to 4♥ and then any bid by the relayer is a sign-off.

This allows the relay a multitude of relays which can be used for various purposes. The schemes I have seen are have been something like:

1st step relay for AKQ (3-2-1) controls
2nd step relay for AK (2-1) controls
3rd step RKCB in the longest highest suit
4th step RKCB in the next suit (in the longest highest ordering)
5th step RKCB in the next suit
6th step RKCB in the last suit

3NT and 4♦ are eliminated from the steps.

e.g. a relay that finished at 3♥ showing 4=1=6=2 would continue

3♠ asks for 3-2-1 controls
3NT sign-off
4♣ asks for 2-1 controls
4♦ puppet terminator
4♥ RKCB for diamonds
4♠ RKCB for spades
4NT RKCB for clubs
5♣ RKCB for hearts

I can see that some variation to this sort of scheme based on the level of suit length resolution may be useful.

We also use weak (two-step) relays at low levels after which if partner shows a minimum with the first step we break to natural bidding. e.g. In an unsophisticated symmetric scheme

1♣ 1NT showing a balanced hand
2♦ weak relay

then after a 2♥ minimum response any new suit is natural.

[Gerben42]

At a very low level, 1♣ - 1♥♠NT, we use the bids 2♦ and up as role reversal with a minimum balanced hand by opener. In this case, OPENER describes.

In all other situations, break outs set trump and subsequently show a void in a suit, because hands with a void are the ones that do not work well with relays.

[hrothgar]

Gerben42, on Aug 7 2008, 04:37 PM, said:

At a very low level, 1♣ - 1♥♠NT, we use the bids 2♦ and up as role reversal with a minimum balanced hand by opener. In this case, OPENER describes.

In all other situations, break outs set trump and subsequently show a void in a suit, because hands with a void are the ones that do not work well with relays.

Hi Gerben

This treatment seems rather strange to me. Many people argue that relay structures work best when the balanced hand is asking and the unbalanced hand shows. Designing a reverse realy structure to permit the balanced hand to show seems a bit odd.

There are a couple examples inside MOSCITO in which balanced hand shows, but this is (normally) when

1. Both hands are known to be balanced
2. The strong club opener is VERY weak

The idea is that the first break (reversing the relay) shows a balanced hand and limits strength. In turn, this encourgaes a second relay break transitioning to natural bidding.

The following auction is protypical

1♣ - 1♦
1N

1♣ = Strong
1♦ = Artifical game force
1N = Minimum opening hand, balanced or 4441

at this point in time,

2♣ = relay
2♦+ = reverse relay (showing a balanced hand) and - once again - limiting strength

The idea behind an auction like

1♣ - 1♦
1N - 2♦

(Where 2♦ = a balanced hand with 4+ Hearts)

is that both hands know

1. That the other hand is limited
2. That the other hand is balanced

Its probaby a good idea to break relays and swtich to natural bidding so you can explore stoppers and the like and make an intelligent decision about 3NT.

[rbforster]

hrothgar, on Aug 6 2008, 04:17 PM, said:

The optimal use of a relay break is highly context specific.  I recommend that you try to reframe the question and focus on a few key sequences that you find especially interesting.

I'm not so concerned at the moment with optimal use of relay breaks just yet. I think the slam bidding rules I outlined in the OP are an example of why. Having a suite of slam bidding tools like denial cuebids and unambiguous keycard in each suit is already a huge improvement over the alternative (which is often keycard in only the last bid suit), in the same way having shape relays is a huge improvement over natural bidding for many slam-going hands. We can argue about whether denial cuebids should show stiff honors, scan for queens in the first pass, etc, but these are all small details compared to having the vast improvement you gain by having any decent relay framework for slam bidding.

In that vein, I was hoping to figure out what people thought were important and try to use a simple set of rules to describe the relay breaks at a given stage earlier in the auction. Even if more customized rules would be better, general rules will allow us to define lots more potentially useful sequences without a huge amount of research and memory overhead.

Let's take the example of stoppers for 3N. When should you worried about playing 3N? When it's going down, clearly . Likely cases for this situation are:

- you have no fit (or maybe a minor suit fit)
- you don't have extra strength
- you have a weak suit opposite responder's shortness

What does this suggest for your relay design and relay breaks?

First, you might want to take a "Majors First" approach towards showing suits so that opener/relayer can judge earlier whether to make a relay break to explore for stoppers (while there's still more space for exploration). Let me take the balanced relay as a good example (since there's a lot less space below 3N than with other games, so this wants to be more of a priority than opposite unbalanced hands more suitable to suit contracts).

Maybe your balanced relay (1♣-1♠* for me) wants to show shape in a different order rather than length-first. With x44x, instead of showing 4♥, then 4♦, then 3-2 or 2-3 in the blacks, maybe you want to show 4♥ then # of spades (even if only 2 or 3), then minor lengths. This is because opener might have 5♠ and doesn't want to relay all the way to full shape just to see if you have 2 or 3 of them, at which point there's no way to tell if his small doubleton opposite your 3 card suit is going to hold up in NT.

Another reason to show majors first is that without extra strength opener is less likely to care about a minor suit fit since 3N is the most likely game anyway. An example of this principle might be how one shows 4333 hands. Rather than showing the 4 card suit and then the 4333 shape (or v.v), it might be better to have earlier steps show 4M333 (step1 - which major?, step2 - tell me about minor suit stoppers, step3 - show strength/controls) and another step to show 4m333 (step1 - strength/controls, step2 - tell me about stoppers, step3 - which minor?).

Of course if you determine you don't have the stoppers you want for 3N, the next order of business becomes identifying whether you have sufficient strength to play in a minor game (or not), and whether a 5-2 or 4-3 major fit might be playable.

These are just some ideas for how these other concerns might inform both your relay design process as well as your relay breaks.

[hrothgar]

Rob F, on Aug 7 2008, 09:01 PM, said:

hrothgar, on Aug 6 2008, 04:17 PM, said:

The optimal use of a relay break is highly context specific.  I recommend that you try to reframe the question and focus on a few key sequences that you find especially interesting.

I'm not so concerned at the moment with optimal use of relay breaks just yet. I think the slam bidding rules I outlined in the OP are an example of why. Having a suite of slam bidding tools like denial cuebids and unambiguous keycard in each suit is already a huge improvement over the alternative (which is often keycard in only the last bid suit), in the same way having shape relays is a huge improvement over natural bidding for many slam-going hands. We can argue about whether denial cuebids should show stiff honors, scan for queens in the first pass, etc, but these are all small details compared to having the vast improvement you gain by having any decent relay framework for slam bidding.

In that vein, I was hoping to figure out what people thought were important and try to use a simple set of rules to describe the relay breaks at a given stage earlier in the auction. Even if more customized rules would be better, general rules will allow us to define lots more potentially useful sequences without a huge amount of research and memory overhead.

Let's take the example of stoppers for 3N. When should you worried about playing 3N? When it's going down, clearly . Likely cases for this situation are:

- you have no fit (or maybe a minor suit fit)
- you don't have extra strength
- you has a weak suit opposite responder's shortness

What does this suggest for your relay design and relay breaks?

First, you might want to take a "Majors First" approach towards showing suits so that opener/relayer can judge earlier whether to make a relay break to explore for stoppers (while there's still more space for exploration). Let me take the balanced relay as a good example (since there's a lot less space below 3N than with other games, so this want be a priority).

Maybe your balanced relay (1♣-1♠* for me) wants to show shape in a different order rather than length-first. With x44x, instead of showing 4♥, then 4♦, then 3-2 or 2-3 in the blacks, maybe you want to show 4♥ then # of spades (even if only 2 or 3), then minor lengths. This is because opener might have 5♠ and doesn't want to relay all the way to full shape just to see if you have 2 or 3 of them, at which point there's no way to tell if his small doubleton opposite your 3 card suit is going to hold up in NT.

Another reason to show majors first is that without extra strength opener is less likely to care about a minor suit fit since 3N is the most likely game anyway. An example of this principle might be how one shows 4333 hands. Rather than showing the 4 card suit and then the 4333 shape (or v.v), it might be better to have earlier steps show 4M333 (step1 - which major?, step2 - tell me about stoppers, step3 - show strength/controls) and another step to show 4m333 (step1 - strength/controls, step2 - tell me about stoppers (or v.v.), step3 - which minor?).

Of course if you determine you don't have the stoppers you want for 3N, the next order of business becomes identifying whether you have sufficient strength to play in a minor game (or not), and whether a 5-2 or 4-3 major fit might be playable.

These are just some ideas for how these other concerns might inform both your relay design process as well as your relay breaks.

When I am designing relay structures I typically focus on ease of use and efficiency of the core relay structure. In theory, it might make sense to design a relay structure in to make it as consistent as possible with a set of relay breaks. However, my expectations is that any gains in efficiency would be outweighed by other issues.

Let's consider the following simple example

Lets assume that you're 1NT bid promises a balanced hand
(any 4432, 4332, or 5332 hand)

One pretty standard relay structure is

2C = relay, and then

2D = 4+ Hearts (not 4333)
2H = 4+ Spades (denies 4 Hearts)
2S = any 4333 hand
2N = 5332 with long clubs
3C = 2-3-4-4 / 3-2-4-4 shape
3D = 2-3-5-3 shape
3H = 3-2-5-3 shape
3S = 3-3-5-2 shape

In theory, we could decide something like the following: Relay breaks will be most useful when there is no major suit fit. We should be sure to clarify whether partner has a short major rather than a long major at as low a level as possible.

This in turn would lead to

2C = relay, and then

2D = 2-3 Hearts
2H = Precisely 2 Spades
2S = any 4333 hand
2N = 5332 with short Diamonds
3C = 4-4-2-3 / 4-4-3-2 shape
3D = 3-3-5-2 shape
3H = 3-5-3-2 shape
3S = 5-3-3-2 shape

This system does look to have some interesting propoerties. However, I suspect that whatever gains that you might enjoy from your relay breaks would be outweighed by

1. The failure to use transfers (bidding 2D to show 4+ Hearts and 2H to show 4+ Spades which may increase the likihood that the relay responder is declarer

2. Higher memory load (this relay scheme isn't consistent with the rest of the system)

[hrothgar]

For what its worth, here is the rule set that I tend to follow during GF auctions

The core principle to focus on is what new information was provided by relay responder’s last bid. (Did he show a suit for the first time? Did he show a two suited pattern with two known suits?)

If relay responder’s last bid showed a 4+ card suit, relay breaks show a singleton or void in that suit using a symmetric ABC relay

If relay responder’s last bid shows a two suited pattern (5-4), the relay break shows a two-suiter with the other two suits using a symmetric AB relay.

If relay responder’s last bid shows a balanced hand, the relay break shows a balanced minimum.

If relay responder’s last bid shows a single suited pattern, the relay break focuses on stoppers.

If relay responder last bid didn’t provide new information, the relay break focuses on stoppers.

[rbforster]

I agree transfers, especially to unbid majors, are an important goal of your relay system (the primary one in my system, Seismic club). In my case of 1♣-1♠* for balanced hands, after the 1N relay there are only red suits as yet unbid. Surely the 2♦ call wants to show 4+ hearts. Maybe 2♣ wants to deny a 4-5 card major (usually has diamonds)?

I'll have to think some more about it in detail, but I thought the balanced relay would be a good place to start for these stopper-based relay breaks since 1) balanced hands are most likely to play in NT, 2) several suits and usually NT have already been bid, making right-siding considerations less important, and 3) the shape relays for balanced hands are often pretty ad-hoc compared to 1- or 2-suited relays (e.g. TOSR) so changing them around isn't so much of an aesthetic loss.

[hrothgar]

Rob F, on Aug 7 2008, 09:56 PM, said:

I agree transfers, especially to unbid majors, are an important goal of your relay system (the primary one in my system, Seismic club). In my case of 1♣-1♠* for balanced hands, after the 1N relay there are only red suits as yet unbid. Surely the 2♦ call wants to show 4+ hearts. Maybe 2♣ wants to deny a 4-5 card major (usually has diamonds)?

I'll have to think some more about it in detail, but I thought the balanced relay would be a good place to start for these stopper-based relay breaks since 1) balanced hands are most likely to play in NT, 2) several suits and usually NT have already been bid, making right-siding considerations less important, and 3) the shape relays for balanced hands are often pretty ad-hoc compared to 1- or 2-suited relays (e.g. TOSR) so changing them around isn't so much of an aesthetic loss.

Here's my advice:

1. I think that you are jumping back and forth between two different topics. The first is how your shape showing relays are structured (what order do you show what). The second is what your relay breaks are showing.

I strongly recommend that you follow a stage designed process. Assume that your relay structure is fixed in stone. You're happy with it. You aren't going to change it. Focus on relay breaks in isolation.

2. Try to identify a few common cases. As I mentioned in an earlier post, I tend to break things up as follows

Relay responder has shown a suit for the first time
Relay responder has shown two known suits for the first time
Relay responder has shown a balanced hand for the first time
Relay responder has shown a single suited hand for the first time

[OleBerg]

Rob F, on Aug 5 2008, 10:39 PM, said:

...
relay-[exact shape shown at 3♠ or less]
3N = to play
4♦ = signoff, relay to 4♥
(and then excluding the above)
step1 = asking for controls/strength (then denial cuebids to locate honors)
step2 = RKC in responder's long suit (and then Q ask and/or control asking bids in side suits)
step3-5 = RKC in other suits, ordered by responder's length
step6+ undefined

My experience with relay-systems is somewhat limited, 2½ years of steady play, then some on and off play.

The idée with 4♦ appeals a lot to me, but I think I would prefer step 2-5 to be slaminvitational with the same rules for suits.

[Free]

OleBerg, on Aug 8 2008, 08:56 AM, said:

Rob F, on Aug 5 2008, 10:39 PM, said:

...
relay-[exact shape shown at 3♠ or less]
3N = to play
4♦ = signoff, relay to 4♥
(and then excluding the above)
step1 = asking for controls/strength (then denial cuebids to locate honors)
step2 = RKC in responder's long suit (and then Q ask and/or control asking bids in side suits)
step3-5 = RKC in other suits, ordered by responder's length
step6+ undefined

My experience with relay-systems is somewhat limited, 2½ years of steady play, then some on and off play.

The idée with 4♦ appeals a lot to me, but I think I would prefer step 2-5 to be slaminvitational with the same rules for suits.

If you have a slam invitational hand, you can ask partner's strength and start denial cuebids. Usually you'll know in time if slam will be possible or not.

Sometimes you just need to know keycards and what partner holds in 1 suit. In that case RKC is absolutely necessary!

Using steps 2-5 (or 3-6) as slam invite is imo quite useless. Partner showed his shape, so the only thing he can do is try to guess what you want to know and act accordingly. Position of honors may be the issue, but then you can ask pretty safely what he has in a certain suit. If you want an overall view of his honors, just start with an AKQ-ask or AK-ask and a first scan.

[hrothgar]

Here's another type of relay break that folks might find interesting:

Let's assume an auction like the following

1♦ - 1♥
2♣ - 2♦
2N -

1♦ = 4+ Hearts, 1♥ = relay, game invite or better
2♣ = Hearts and Clubs, 2♦ = relay (GF is established)
2N = 5+ Hearts and 4 Clubs, 0-1 Spades

At this point in time

3♣ = relay, asking for shape

3♦ = relay break. Responder will still show shape, however, he will exclude his longest suit (Hearts) when showing controls/denial cue bidding

3♥ = relay break. Responder will still show shape, however, he will exclude his second longest suit (Clubs) when showing controls/denial cue bidding

This style of relay break allows the relay captain to conveniently switch from standard control asking to exclusion sequences.

[Cascade]

hrothgar, on Aug 9 2008, 04:28 AM, said:

Here's another type of relay break that folks might find interesting:

Let's assume an auction like the following

1♦ - 1♥
2♣ - 2♦
2N -

1♦ = 4+ Hearts, 1♥ = relay, game invite or better
2♣ = Hearts and Clubs, 2♦ = relay (GF is established)
2N = 5+ Hearts and 4 Clubs, 0-1 Spades

At this point in time

3♣ = relay, asking for shape

3♦ = relay break. Responder will still show shape, however, he will exclude his longest suit (Hearts) when showing controls/denial cue bidding

3♥ = relay break. Responder will still show shape, however, he will exclude his second longest suit (Clubs) when showing controls/denial cue bidding

This style of relay break allows the relay captain to conveniently switch from standard control asking to exclusion sequences.

i like the idea

[Mbodell]

OleBerg, on Aug 7 2008, 11:56 PM, said:

Rob F, on Aug 5 2008, 10:39 PM, said:

...
relay-[exact shape shown at 3♠ or less]
3N = to play
4♦ = signoff, relay to 4♥
(and then excluding the above)
step1 = asking for controls/strength (then denial cuebids to locate honors)
step2 = RKC in responder's long suit (and then Q ask and/or control asking bids in side suits)
step3-5 = RKC in other suits, ordered by responder's length
step6+ undefined

My experience with relay-systems is somewhat limited, 2½ years of steady play, then some on and off play.

The idée with 4♦ appeals a lot to me, but I think I would prefer step 2-5 to be slaminvitational with the same rules for suits.

Yeah we use the 4♦ transfer signoff as well when we can't fit all our steps into 4♦ and less. When we know the location of 12 cards then our relay structure, skipping 3nt and 4♦ if necessary, is:

step 1: relay out full shape
step 2: key card for suit 1 (longest first, then highest, if we don't know about a card put it in the highest suit it can be in, so if we know the person has a majors 4333 assume it is in spades)
step 3: key card for suit 2
step 4: key card for suit 3
step 5: key card for suit 4

If we assume the last bid was 2♠ say showing that 4333 majors hand then this would mean these steps were 2nt through 3♠.

If we know exact shape then step 1 is control ask and steps 2-5 are key card in suits as above.

Once we know the control answer below 4♦ though we use the 4♦ sign off to let us start a variety of scanning cue bidding sequences:

step 1: viking cues where denying a suit shows 0 or 2 of the top cards, queens on second pass.
step 2: wimpy cues where denying a suit shows 0 of the top 2, skipping it shows 1+, queens included on second pass.
step 3: super viking where denying a suit shows 0 or 2 of the K and Q, jacks on second pass.

I like being able to have a number of different ways to ask things, and the opportunity for these breaks do come up more frequent than the earlier low level shapely breaks.

With respect to the relay breaks saying skip a suit for control/denial cues, how does that effect your control asks? For us we have gf positives need 2 controls. And our cheapest answer to control ask shows 2 controls. If we said skip the ♦ suit do you now need the cheapest suit answer to the control ask to start at 0? Or do you play a range on that, in which case the denial cues might get more tricky.

[Zelandakh]

Relay breaks should be used to find out things that are otherwise difficult in your system imho. For example, some relay systems have difficulty in determining fragment stoppers, and if this is the case then some breaks should be assigned to this. For slam purposes it can often be useful to find out about a specific side suit's controls before number of controls - so a way to revert to 'natural' slamming is good if you use a spiral scan/denial cue method within the relays. It is even more important using the 4-suit RKCB idea too - it's always a good idea to know the 5 level i safe before you get there!

A simple concept that you might like:-

Non-jump relay breaks in a fragment suit are stopper asks; think of it like 4th suit forcing for natural systems.
Non-jump relay breaks in an 'unfocused' suit are a stop ask if one of the fragment suits is unavailable for a relay break.
Jump relay breaks below 3NT are natural with slam interest, and start natural slamming in your preferred style.
A relay break of 2NT is a natural slam invite that sometimes saves going to 4NT.

The above is not perfect from a theoretic perspective, but is simple and effective. An more complex and better alternative would be to order the suits by
fragment > unknown suit > real suit > splinter
and use relay breaks to refer to each suit in order.

Basically if you are designing a new system, play-test it using only relays and see which hands give you the most difficulties, and then design relay breaks that address that issue first and foremost. Do you reach 3NT on Qx opposite xxx? Or bid slams with an AK missing in a suit? Or ever end up in 5m when 3NT was better? There's no way you can pass all the information required using only relays, so whatever is missing fill it in with the breaks.