Reverse relays

[rbforster]

In more detailed relay systems, there are often methods to "reverse the relays" and have the opening hand show rather than ask. For example, in TOSR after responder show spades:

 

1♣(16+) – 1♥(4+♠ GF)

 

opener can either relay with 1♠ to inquire or bid 1N+ to show aspects of his hand. Specifically TOSR shows minimum 3-suited hands with these relay breaks - 4441, 5440, and sometimes 5431.

 

1s: Normal relay.

1nt: RR, short spades.

2c: RR, short hearts.

2d: RR, short diamonds.

2h: RR, short clubs, zoom to next step.

 

I understand the logic of how showing short spade hands opposite partner's spade positive can be useful (such as determining whether his spade holding is suitable for NT or not). It's much less clear to me why you would want to show 3 suited hands with other shortnesses, especially if you might have a fit for partner's major.

 

What do you think about showing all hands with shortness in partner's suit, rather than just all 3-suiters? Or what about showing hands that have length in the suit partner bid (so he can declare), as opposed to the suit he showed? In the above case, it might make sense to describe a 1-suited heart opener since that's a likely strain when opener has long hearts and short spades, and responder is going to declare hearts in any case.

 

If you limit "shortness" to 0-1 cards, there are only 4 even moderately common single suited hand types (6331, 7(32)1, and 7330), and likewise only 4 3-suiters (4441, 5440's). Most 2-suiters remain, excluding only 5422. For example, what about something like this

 

1♣* - 1♥* (spades):

 

1S: relay

 

(all relay breaks show 0-1 spades):

 

1N: long hearts, long diamonds, or hearts & clubs

   2D long hearts or long diamonds

      2S long diamonds (3C-3S for which)

      2N+ long hearts

   2H+ TOSR relays for hearts and clubs

 

2C: reds (transferring to diamonds)

   2H+ TOSR relays for reds

 

2D: long clubs or 3-suited

   2S 3-suited (then 3C-3S show 4441/5440's)

   2N+ long clubs

 

2H+: TOSR relays for minors

 

Similar relays could be devised over different positive responses, depending in part on what was shown and how much space remained.

[hrothgar]

The first thing that you need to do is to figure out what you want your relay breaks to show:

 

Do you want your reverse relay to limit strength?

 

Do you want your reverse relay to show a misfit?

 

Do you want to use a relay break to transition to natural bidding, start asking/showing stoppers, what have you?

 

I suspect that you'll quickly discover is that you'll want the meaning of your relay break to change depending on

 

Whether or not a game force has been established

The level the auction has reached

How much information the relay asker has about relay responder's hand

 

Consider the following auction

 

1C - 1H

 

where 1H shows a unbalanced hand 4+ Spades and GF values

 

I prefer to use a relay break to show hands with 0-1 Spades (this could be three suited, two suited, or single suited). I like this style because I think its useful to try to identify a misfit as quickly as possible.

 

(BTW, this can be supported with a VERY simple symmetric response structure thats a lot easier than what you are suggestion)

 

From what you said, TOSR prefers a different style. The relay break shows ANY three suited hand.

[hrothgar]

For what its worth, here's the relay schedule that I describe

 

Let's assume that the auction starts

 

1C - 1H (where 1H shows a GF with 4+ Spades)

 

Your relay break shows 0-1 Spades. There are three remaining suits that you might have. Let's order them

 

A = Hearts

B = Diamonds

C = Clubs

 

Here's the initial round of breaks

 

1N = 4+ cards in suit A

2C = 4+ cards in suit B (0-3 in suit A)

2D = 3 suited hands with 0-1 Spades

2H = Single suited with suit C

 

After

 

1C - 1H

2C - 2D = relay

 

2H = Single suited with suit B

2S+ = Two suited with suits B and C

 

After

 

1C - 1H

1N - 2C = relay

 

2D = 2 suited with A+B

2H = single suited with A

2S+ = 2 suited with A+C

 

Notice that

 

Single suited hands always pivot arround a raise of RR's suit

Two suited hands always start resolving at raise of RR's suit +1

[Free]

I really love the scheme Richard explains. When it comes up, it's very useful. Showing any 3-suiter doesn't seem useful to me.

[Gerben42]

What I play with my relay partner is that you bid a suit that you set trumps, partner is forced to relay, and then you show a void.

 

Example:

 

1♣ - 1♥

2♦ (break - shows long ♦) - 2♥ (forced)

2NT (void in ♥)

[awm]

I've recently been trying a style where the non-step bids do not actually break the relay. So for example:

 

1♣ - 1♥ (4+♠ GF)

......

 

1NT = 0-1♠, but describer continues describing having lost a step. However, now a subsequent 3NT signoff is conditional on describer having spade cards (should be removed without), a subsequent control ask excludes the king of spades, and spades are removed as a possible suit for "RKC asks" unless describer shows six or more of them later on.

 

This seems easier than having a different relay structure to allow opener to describe (which is often space-inefficient anyway if the initial response is not so cheap, such as 1♣-2♣ showing diamonds).

[akhare]

IMO, the best use of reverse relays to limit opener's strength and it seems to work pretty well in Moscito scheme that uses 1♦ response to 1♣ as any GF.

 

For instance, the auction:

 

1♣ - 1♦ (GF) - 1N (RevR. balanced hand with 9-11 QPs) - 3N is much better than 1♣ - 1♦ - 1♥ - 1♠ - 1N - 2♣...3N.

 

Also, responsder can completely resolve RevR's shape if needed.

 

Does anyone have an opinion on whether the RevRelays should be optimized to maximize the chances of responder becoming declarer? The structure that pard and I came up with actually maximized the RevR's chances of become declarer.

 

The idea was that after opener limits the hand by reverse relaying, the chances of a slam become remote and therefore responder should relay once or twice to resolve relative suit length and place the final contract. This also allows responder to completely break out relays as well (say by using transfers over a 1N rebid).

 

For example after 1♣ - 1♦ (GF):

 

1♠: All unbalanced hands with with ♠

1N: All balanced hands

2♥: Single suited with ♥s

2♠: ♥+♦ (reversed)

2♠: ♥+♦ (LL) etc.

[PrecisionL]

In Ultra Club:

 

We don't break the relay unless opener is 5-5 or very strong 1-suited because responder could easily have another suit (even longer than the Major promised by the 1♥ or 1♠  transfer response to our strong opening 1♣ bid).

 

Very strong 1-suiters break the relay with an asking bid for support & Controls (if there is support).

 

5-5 hands ask for the 3-card fragment (if there is one) usually by bidding 2NT.

 

We have played this 3 years now and rarely find that we wish to break the relay.

[hrothgar]

I've recently been trying a style where the non-step bids do not actually break the relay. So for example:

 

1♣ - 1♥ (4+♠ GF)

......

 

1NT = 0-1♠, but describer continues describing having lost a step. However, now a subsequent 3NT signoff is conditional on describer having spade cards (should be removed without), a subsequent control ask excludes the king of spades, and spades are removed as a possible suit for "RKC asks" unless describer shows six or more of them later on.

 

This seems easier than having a different relay structure to allow opener to describe (which is often space-inefficient anyway if the initial response is not so cheap, such as 1♣-2♣ showing diamonds).

Hi Adam

 

In my original post, I commented that the optimal meaning of a relay break depends on a lot of different issues, including where you are in the auction.

 

I like the style of relay break that you are discussing. (As I recall, I first saw this style discussed here on the BBO forums a few years back. Sadly, I forget who initial introduced this concept)

 

However, I don't think that this style of relay break is a reasonable choice early in the auction. Consider an auction like 1C - 1S or the MOSCITO auction

 

1C - 1D

1H - 1S

 

You have all the time in the world to start excluding suits. I don't see the need to do this at this point in time. Surely, there must be more valuable information to convey...

[akhare]

However, I don't think that this style of relay break is a reasonable choice early in the auction. Consider an auction like 1C - 1S or the MOSCITO auction

 

1C - 1D

1H - 1S

 

You have all the time in the world to start excluding suits. I don't see the need to do this at this point in time. Surely, there must be more valuable information to convey...

What's your opinion on using reverse relays to limit opener's strength after 1♣ - 1♦ (any GF)?

[rbforster]

Thank you all for your descriptions, especially Richard for his symmetric reverse relays.

 

For those of you who play reverse relays, which ones do you tend to bid/use with some regularly?

 

I've anecdotally heard from a few relay players is that they tend not to use reverse relays or other relay breaks very often. This seems somewhat counter-intuitive in terms of efficient system design, since there are millions of bidding sequences that start 1♣-1♥-not 1♠ and it seems a shame not to use them. If they aren't getting used under your current definition of relay breaks, surely this is a sign you should try to find a more common/useful meaning that would put these sequences to use.

[Impact]

Rob,

 

The most common reverse relay I utilise is for opener with 15/16-18 after responder bids 1NT (10+ balanced) and opener shows his shape both to protect the likely 3NT declarer from revealing more and to evaluate shortage.

 

Otherwise, although I tinkered with all sorts of ideas, in GF auctions I have found that the best relay breaks are extreme single-suiters and 2 -suiters with a suit which might be solidified by finding a singleton honour. Hence relay break in such takes the form of an asking bid in that suit...(take your choice form a plethora of asking bids but Zeta was useful).

 

regards

[shevek]

We don't use reverse relay any more, though they were part of Kiwi club in early 80s.

Found it too complicated after say 1♣ - 1♠ (hearts)

 

We didn't want to put too much effort into rare artificial sequences. These days we break with ill-fitting hands, usually minimums. Auctions are F2NPR (forcing to 2NT, preference or repeat) so we can occasionally pull up short. This can happen when both have added length points, so 16+8 has become 15+7 with no fit.

 

Breaks other than raises show max of small doubleton in responder's suit:

1♣   1♠ (hearts)

1NT   GFR

2♣/♦/♠ F2NPR

2♥   balanced 4-cd raise

2NT+  various fit-jumps

 

Hands with shortage opposite partner's length are bad for relay. Also, good to raise with flat minimums to give responder a chance to show shortage.

 

We use a similar style opposite responder's balanced GF:

 

1♣   2♣   8+ balanced with a Major

2♦   GFR

2♥/♠/NT  natural, can stop short

3x  splinters

 

These splinters have been useful, to avoid 3NT, cool partner's heels and find slams on perfect fits.

[the hog]

We experimented with reverse relays to show different hand types.

First we played with the rr showing shortness in opener's suit. We found that wasn''t necessary that often.

The most useful rr we used was for opener to show a 15-17 hand type, iow a limited hand. This meant that responder was not obliged to flow on with the traditional base+2 AK points or basw +4 AKQ points as would be normal after a sign off. we thus avoided uncomfortable higher level contracts.

This was BY FAR the most useful method for using rrs. It was definitely NOT counter intuitive, nor was it rare or difficult to remember as some of the posters above have suggested.

[ulven]

I play reverse relays and have been for a decade now. Initial replies to 1C:

1H = balanced (include 5M332)

1S = 5+m

1NT = 5+S

2C = 3-suited

2D+ = 5+H

 

Only use rr after 1H:

1S = 5+m

1NT = 5+S

2C = range-ask w/bal/3-suited (no rr)

2D+ = 5+H

 

So, it's completely symmetric (but not using Kerr "symmetric relays") and only used when responder is balanced, which is where 80+% of the gain is.

 

Starting balanced hands a bit higher, we don't use relays there. Instead of showing exact shape, we choose to pinpoint ranges in 2-point intervals. It's perhaps a matter of opinion which is best (or not).

 

This works very well and is easy to use/remember.

[shevek]

We experimented with reverse relays to show different hand types.

First we played with the rr showing shortness in opener's suit. We found that wasn''t necessary that often.

The most useful rr we used was for opener to show a 15-17 hand type, iow a limited hand. This meant that responder was not obliged to flow on with the traditional base+2 AK points or basw +4 AKQ points as would be normal after a sign off. we thus avoided uncomfortable higher level contracts.

This was BY FAR the most useful method for using rrs. It was definitely NOT counter intuitive, nor was it rare or difficult to remember as some of the posters above have suggested.

Guess I don't understand this. It's been a long while since I played reverse relay.

So

1♣ - 1♠ (hearts, not spades)

 

1NT  relay

2♣  is this 15-17 and the shape that would respond 2♣?

 

I guess not since that means you miss out on shapes that would respond 1♥/♠/NT.

Not so useful so what is 2♣?

If it's a shape that would respond 1♥ (spades) you are 3 steps up. That can't be useful. If it's something else, I would forget.

 

Also, helpful to defence to have the 15+ hand known & the 9+ part known.

 

A minor point but responder should not autoressurect with base +4 Slam points.

Clearly not with say 4333 & ace-empties. Has to depend on shape & the amount of space relayer had.

A better solution is to give relayer a mild slam try.

We don't use end signal. Rather, after shape is out, step is SPs, +1 is kontrols, 4D is a mild slam try. Responder bids 4H with minimums & bids bad trump suits up the line with base +2 say.

[the hog]

Guess I don't understand this. It's been a long while since I played reverse relay.

So

1♣ - 1♠ (hearts, not spades)

 

1NT  relay

2♣  is this 15-17 and the shape that would respond 2♣?

 

I guess not since that means you miss out on shapes that would respond 1♥/♠/NT.

Not so useful so what is 2♣?

If it's a shape that would respond 1♥ (spades) you are 3 steps up. That can't be useful. If it's something else, I would forget.

 

Also, helpful to defence to have the 15+ hand known & the 9+ part known.

 

A minor point but responder should not autoressurect with base +4 Slam points.

Clearly not with say 4333 & ace-empties. Has to depend on shape & the amount of space relayer had.

A better solution is to give relayer a mild slam try.

We don't use end signal. Rather, after shape is out, step is SPs, +1 is kontrols, 4D is a mild slam try. Responder bids 4H with minimums & bids bad trump suits up the line with base +2 say.

Nick or Nicoletta

 

1C 1S

2C

would show S and a D suit and 15-17. You are showing the suit that responder BID, not showed, and shape

 

1C 1S

2S

S low shortage, (assuming 2S low sht, 2NT mid sht and 3D+ high sht)

 

1C 1S

3D

5233 shape exactly and 15-17, etc etc

 

1C 1H

1NT

 

Balanced 15-17

 

This is the original version played by Paul, and imo by far the best. Justin and I have experimented with natural relay breaks and found them a waste of time. Btw we disagree with your failure to use the end signal. We found this very useful.

[hrothgar]

However, I don't think that this style of relay break is a reasonable choice early in the auction.  Consider an auction like 1C - 1S or the MOSCITO auction

 

1C - 1D

1H - 1S

 

You have all the time in the world to start excluding suits.  I don't see the need to do this at this point in time.  Surely, there must be more valuable information to convey...

What's your opinion on using reverse relays to limit opener's strength after 1♣ - 1♦ (any GF)?

Always liked it...

[hrothgar]

One important point to consider about reverse relays:

 

The failure to use a reverse relay conveys information.

 

In most cases, this won't matter. The relay captain will be making almost all the decisions. However, on occasions the opponents might interfere in the middle of a relay auction.

 

It can be helpful to know what type of hands the relay asker (typically) won't have...

 

(BTW: Proper disclosure when the relay asker fails to use a relay break can be a bitch)

[the hog]

By the way, I once had pd reverse my reverse relay. He was just being a smartarse. I picked it up and we did get to the right contract.

[the hog]

sorry, duplicate post.

[shevek]

Nick or Nicoletta

 

1C 1S

2C

would show S and a D suit and 15-17. You are showing the suit that responder BID, not showed, and shape

 

1C 1S

2S

S low shortage, (assuming 2S low sht, 2NT mid sht and 3D+ high sht)

 

1C 1S

3D

5233 shape exactly and 15-17, etc etc

 

1C 1H

1NT

 

Balanced 15-17

 

This is the original version played by Paul, and imo by far the best. Justin and I have experimented with natural relay breaks and found them a waste of time. Btw we disagree with your failure to use the end signal. We found this very useful.

 

OKay I see. No thanks.

David Morgan likes end signal but like you with natural breaks, I found 4D a waste of time. I guess the appropriate hands didn't turn up.

 

Hands for natural breaks are quite common though

 

[hv=d=w&v=n&w=sxhakxxxdkqxckxxx&e=saxxxxhxxdaxxcxxx]266|100|Scoring: IMP[/hv]

 

We bid 1♣ - 1♥ - 1NT - pass!

 

Nick

[rbforster]

Looking at the above hands posted by Nick, I wouldn't be comfortable forcing to game with 5M332 and 2 aces opposite a 15+ club. Of course opposite most 15+ hands, two aces will be enough either because there's a fit with good controls and/or because partner has the extra strength I lack.

 

So one philosophy for relay breaks could be to show a misfitting minimum, to compensate for aggressive bidding by responder. Essentially this turns into a non-game-forcing relay system.

[shevek]

Looking at the above hands posted by Nick, I wouldn't be comfortable forcing to game with 5M332 and 2 aces opposite a 15+ club.  Of course opposite most 15+ hands, two aces will be enough either because there's a fit with good controls and/or because partner has the extra strength I lack.

 

So one philosophy for relay breaks could be to show a misfitting minimum, to compensate for aggressive bidding by responder.  Essentially this turns into a non-game-forcing relay system.

Old-fashioned 16+ for us. I guess West "added a point for a decent 5-card suit" which would be better with spades.

However, auctions starting 1♥ (or 1♦ showing 4+ hearts) tend to be awkward with this shape.

We play 1♣ - 1♦ - 2♣ as Aspro

[the hog]

Looking at the above hands posted by Nick, I wouldn't be comfortable forcing to game with 5M332 and 2 aces opposite a 15+ club. Of course opposite most 15+ hands, two aces will be enough either because there's a fit with good controls and/or because partner has the extra strength I lack.

 

So one philosophy for relay breaks could be to show a misfitting minimum, to compensate for aggressive bidding by responder. Essentially this turns into a non-game-forcing relay system.

No, I wouldn't gf on this either.