denial cuebidding

[lowerline]

In Symmetric Relay (the system) one asks for the number of controls after the shape is known. In Moscito one asks for the number of AKQ-points. This has consequences on what you deny during denial cuebidding.

 

Anyone tried both approaches? Which do you prefer?

[Free]

You can combine them: step 1 = AKQ-ask (= SP), step 2 = control ask, step 3+ something else (RKC for example) Problem with combining them is that you need SP and controls enough, and sometimes hands just don't qualify to open or respond (semi)positive. For example, 4Q's and a K is GF opposite a strong 1♣ opener when using SP, but a negative when working with controls.

 

As for comparisson between the 2 methods, I think they're VERY similar but the control ask is probably slightly less accurate in some situations (when relayer has a singleton and missing 3 controls, where SP may give a better picture of the entire hand). When relayer is balanced it might be more efficient to play control asks, since you know about Aces and Kings a lot faster...

[DrTodd13]

One thing foobar and I recently added is that after the control showing bid, one step is the normal start of denial cue-bids for aces and kings. However, a two step starts denial cue-bidding for queens and skips the aces and kings. This would typically be useful where you are only missing 0 or 1 controls and therefore know what control is missing by looking at your hand. This just saves some space and let's you find out more about queens and jacks before you have to make a slam or grand slam decision.

[hrothgar]

Back in the day, MOSCITO used both forms of denial cue bidding.

 

After shape was shown, the relay asker had the option to bid

 

Step to ask for Controls (A = 2, K = 1), followed by DCB for controls

Step +1 to ask for slam points, followed by DCB for honors (A, K, Q)

 

More recently, the decision was made to replace the control ask with RKCB in the different suits. Its was felt that Keycard + control asking bids was more effective than the two tiered DCB scheme. Furthermore, there was more overlap between Keycard and the "Control" DCB style.

 

As I understand things, I THINK that a bunch of people are now playing arround with Exclusion Keycard Sequences rather than Roman...

[joshs]

In general, AKQ points are better when both players are relatively balanced. AK points (e.g. controls) works better when both are unbalanced. And in between there is not much to choose between.

 

It is possible, of course, to have two different asks. Or to have some sequences that say "bid on with extra values (when playing controls) or bid on with extras controls (when playing AKQ points). So the main issue is grand slam bidding, and for that, AKQ points are better. But none of this really matters much....

 

Josh

[awm]

I've tried both approaches and found the AKQ count to be a lot better than straight control asks. The main issue is that distinguishing between A+K and A+K+Q can be very difficult. Playing control-based relays, generally the first step is to show "a minimum response." The next relay response is "three controls." Then we have to scan through and locate the ace and the king, which often means wasting many more steps on DCB. By the time opener can figure out about the extra queen, we're often to high to sign off even at the five-level.

 

Of course, for some A+K isn't a positive, but then you will see basically the same issue with A+K+Q versus A+K+Q+Q.

 

Some kind of "two way" method has benefits, but it's nice to have limits on the expected strength of the hand and it's hard to install these on all three of high card points, AKQ points, and controls without being forced into fairly frequent awkward sequences on hands that are "worth a game force" but fail the criteria for a positive response on one of the three point measures.

[Echognome]

It's not so inefficient to play controls as you might think. We play that you count your control cards (#Aces and Kings) and let that equal X. Then you don't show your Q's until your Xth suit scan. This is usually plenty to piece together the missing honours. Also, you have a natural way of making the scan slower/faster in regards to the strength of relay responder.

[lowerline]

It's not so inefficient to play controls as you might think. We play that you count your control cards (#Aces and Kings) and let that equal X. Then you don't show your Q's until your Xth suit scan. This is usually plenty to piece together the missing honours. Also, you have a natural way of making the scan slower/faster in regards to the strength of relay responder.

I have read about this. It makes sense to focus on aces and kings first, but if relayee does not have a lot of controls, it also makes sense to show the queens asap...

 

I have a feeling that this approach is superior, but maybe it is my enthusiasm when learning something new... ;-)

[joshs]

I've tried both approaches and found the AKQ count to be a lot better than straight control asks. The main issue is that distinguishing between A+K and A+K+Q can be very difficult. Playing control-based relays, generally the first step is to show "a minimum response." The next relay response is "three controls." Then we have to scan through and locate the ace and the king, which often means wasting many more steps on DCB. By the time opener can figure out about the extra queen, we're often to high to sign off even at the five-level.

 

Of course, for some A+K isn't a positive, but then you will see basically the same issue with A+K+Q versus A+K+Q+Q.

 

Some kind of "two way" method has benefits, but it's nice to have limits on the expected strength of the hand and it's hard to install these on all three of high card points, AKQ points, and controls without being forced into fairly frequent awkward sequences on hands that are "worth a game force" but fail the criteria for a positive response on one of the three point measures.

Well I remain unconvinced, maybe you can give some examples?

 

Now I agree that if the only slam try in your arsenal was to relay for controls or for AKQ points, then AKQ points will be better. Just as if in standard you were allowed to only use rkc or a quantative slam try, quantatative would be better.

[awm]

Here's some examples. In each case, west is the dealer. I'll assume methods where 1♣ shows sixteen or more points, and a positive response is game forcing with 9+ points and 2+ controls. In each case responder's exact shape will resolve (playing symmetric or transfer-symmetric relays) with a 3♦ call. Let's suppose your options are control asks (3♥) or keycard in the suit of your choice (4♣+), as well as various "signoff" sequences that responder will not pull with any of the hands given.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx]266|100|Scoring: IMP[/hv]

 

Here you want to be in 4♠. If you start with 3♥ (control ask) partner will bid 3♠ (min); if you follow up with 4♣ (really control ask) then partner shows three and you can no longer stop at the four-level. Keycard will be 4♣ or above and will of course bypass 4♠.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx]266|100|Scoring: IMP[/hv]

 

Here the goal is to reach 6♣, which is excellent, especially with west declaring. This is relatively easy to reach as long as you don't sign off too early... but note that this hand will not be distinguished from the first hand until you're well past 4♠.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx]266|100|Scoring: IMP[/hv]

 

Here 7♠ has reasonable chances. Again this is quite reachable, but consider what happens if you start with 3♥. Partner bids 3♠ (minimum), then 4♥ (three controls), then 5♦ (a top spade, a top heart, no top diamond), then 5♠ (no spade queen)... and you are easily at the six level before you can distinguish this hand from the hand below. If you start with 4♣ (RKC in spade) then you get 4♠ (two no queen), then 4NT would ask for the heart king, 5♣ for the diamond king, 5♦ for the club king, 5♥ for the heart queen... and again you'll be at the six level before you know about the diamond queen.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx]266|100|Scoring: IMP[/hv]

 

Here no slam has real play, though 5♠ should almost always make. Note that this hand will make all the same bids as the hand above it until you are past 5♠!

 

-------

 

Admittedly these are made-up examples. Next time I practice relaying with someone I'll try to grab some more "random" hands. However, my feeling is that this issue arises fairly often. The main problem is when one of:

 

(1) You're not sure of the proper trump suit, because a 5-3 or 5-4 fit is likely to be "safest" but your only possible slam is in a 4-4 fit in another suit. If you relay for controls it is hard to find out about the important queens, but if you bid keycard you must commit to a trump suit too early.

 

(2) The key card for a making slam is a queen outside the trump suit, which covers a critical loser in declarer's hand.

 

(3) You really want a "quantitative" invite to slam in a suit, but the "max" partner needs to accept is not so strong that partner would pull a signoff attempt.

 

Another nice thing about AKQ points is that it reduces the frequency of game forcing "negative" responses. Assuming you require 2 controls for a positive, there are hands like K+Q+Q+Q that are quite good but have to start with a negative. You can still construct such hands if you require 4+ AKQ points for a positive, but there are fewer of them (i.e. K+Q+J+J+J+J or Q+Q+Q+J+J+J).

[hrothgar]

Here's some examples. In each case, west is the dealer. I'll assume methods where 1♣ shows sixteen or more points, and a positive response is game forcing with 9+ points and 2+ controls. In each case responder's exact shape will resolve (playing symmetric or transfer-symmetric relays) with a 3♦ call. Let's suppose your options are control asks (3♥) or keycard in the suit of your choice (4♣+), as well as various "signoff" sequences that responder will not pull with any of the hands given.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx]266|100|Scoring: IMP[/hv]

 

Here you want to be in 4♠. If you start with 3♥ (control ask) partner will bid 3♠ (min); if you follow up with 4♣ (really control ask) then partner shows three and you can no longer stop at the four-level. Keycard will be 4♣ or above and will of course bypass 4♠.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx]266|100|Scoring: IMP[/hv]

 

Here the goal is to reach 6♣, which is excellent, especially with west declaring. This is relatively easy to reach as long as you don't sign off too early... but note that this hand will not be distinguished from the first hand until you're well past 4♠.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx]266|100|Scoring: IMP[/hv]

 

Here 7♠ has reasonable chances. Again this is quite reachable, but consider what happens if you start with 3♥. Partner bids 3♠ (minimum), then 4♥ (three controls), then 5♦ (a top spade, a top heart, no top diamond), then 5♠ (no spade queen)... and you are easily at the six level before you can distinguish this hand from the hand below. If you start with 4♣ (RKC in spade) then you get 4♠ (two no queen), then 4NT would ask for the heart king, 5♣ for the diamond king, 5♦ for the club king, 5♥ for the heart queen... and again you'll be at the six level before you know about the diamond queen.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx]266|100|Scoring: IMP[/hv]

 

Here no slam has real play, though 5♠ should almost always make. Note that this hand will make all the same bids as the hand above it until you are past 5♠!

 

-------

 

Admittedly these are made-up examples. Next time I practice relaying with someone I'll try to grab some more "random" hands. However, my feeling is that this issue arises fairly often. The main problem is when one of:

 

(1) You're not sure of the proper trump suit, because a 5-3 or 5-4 fit is likely to be "safest" but your only possible slam is in a 4-4 fit in another suit. If you relay for controls it is hard to find out about the important queens, but if you bid keycard you must commit to a trump suit too early.

 

(2) The key card for a making slam is a queen outside the trump suit, which covers a critical loser in declarer's hand.

 

(3) You really want a "quantitative" invite to slam in a suit, but the "max" partner needs to accept is not so strong that partner would pull a signoff attempt.

 

Another nice thing about AKQ points is that it reduces the frequency of game forcing "negative" responses. Assuming you require 2 controls for a positive, there are hands like K+Q+Q+Q that are quite good but have to start with a negative. You can still construct such hands if you require 4+ AKQ points for a positive, but there are fewer of them (i.e. K+Q+J+J+J+J or Q+Q+Q+J+J+J).

Couple comments:

 

1. The sequences that AWM is suggesting don't conform to the denial cue bidding that I am used to. Most notable, the combination of a minimum and a slam point ask seems to burn a lot of bidding space. Most of the styles that I have seen unambiguously describe the number of controls/slam points held. For example, after a 3♥ ask, a 3♠ response doesn't promise a "minimum" hand, but rather 6 slam points. At this point in time, 3♠ by the relay asker launches directly into denial cue bidding.

 

2. I've always felt that this topic was particularly ammenable to monte carlo simulations. It should be possible to code a series of different bidding engines, each using a different approach towards placing controls. One engine might use RKCB, a second Control Asking Bids, a third denial cue bidding with AKQ slam points, a fourth DCB with controls, etc. Run a few millions hands and see what works...

[joshs]

Here's some examples. In each case, west is the dealer. I'll assume methods where 1♣ shows sixteen or more points, and a positive response is game forcing with 9+ points and 2+ controls. In each case responder's exact shape will resolve (playing symmetric or transfer-symmetric relays) with a 3♦ call. Let's suppose your options are control asks (3♥) or keycard in the suit of your choice (4♣+), as well as various "signoff" sequences that responder will not pull with any of the hands given.

 

[hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx

 

 

here you want to be in 4♠. if you start with 3♥ (control ask) partner will bid 3♠ (min); if you follow up with 4♣ (really control ask) then partner shows three and you can no longer stop at the four-level. keycard will be 4♣ or above and will of course bypass 4♠.

 

dealer: west vul: both scoring: imp

♠ kxx ♥ aqx ♦ akx ♣ ajxx ♠ axxxx ♥ x ♦ qxx ♣ kqxx

 

here the goal is to reach 6♣, which is excellent, especially with west declaring. this is relatively easy to reach as long as you don't sign off too early... but note that this hand will not be distinguished from the first hand until you're well past 4♠.

 

dealer: west vul: both scoring: imp

♠ aqxx ♥ kx ♦ akxxx ♣ ax ♠ kxxxx ♥ axx ♦ qxx ♣ xx

 

here 7♠ has reasonable chances. again this is quite reachable, but consider what happens if you start with 3♥. partner bids 3♠ (minimum), then 4♥ (three controls), then 5♦ (a top spade, a top heart, no top diamond), then 5♠ (no spade queen)... and you are easily at the six level before you can distinguish this hand from the hand below. if you start with 4♣ (rkc in spade) then you get 4♠ (two no queen), then 4nt would ask for the heart king, 5♣ for the diamond king, 5♦ for the club king, 5♥ for the heart queen... and again you'll be at the six level before you know about the diamond queen.

 

dealer: west vul: both scoring: imp

♠ aqxx ♥ kx ♦ akxxx ♣ ax ♠ kjxxx ♥ axx ♦ xxx ♣ jx

 

here no slam has real play, though 5♠ should almost always make. note that this hand will make all the same bids as the hand above it until you are past 5♠!

 

-------

 

admittedly these are made-up examples. next time i practice relaying with someone i'll try to grab some more "random" hands. however, my feeling is that this issue arises fairly often. the main problem is when one of:

 

(1) you're not sure of the proper trump suit, because a 5-3 or 5-4 fit is likely to be "safest" but your only possible slam is in a 4-4 fit in another suit. if you relay for controls it is hard to find out about the important queens, but if you bid keycard you must commit to a trump suit too early.

 

(2) the key card for a making slam is a queen outside the trump suit, which covers a critical loser in declarer's hand.

 

(3) you really want a "quantitative" invite to slam in a suit, but the "max" partner needs to accept is not so strong that partner would pull a signoff attempt.

 

another nice thing about akq points is that it reduces the frequency of game forcing "negative" responses. assuming you require 2 controls for a positive, there are hands like k+q+q+q that are quite good but have to start with a negative. you can still construct such hands if you require 4+ akq points for a positive, but there are fewer of them (i.e. k+q+j+j+j+j or q+q+q+j+j+j).

dealer: west vul: both scoring: imp

♠ kxx ♥ aqx ♦ akx ♣ ajxx ♠ ajxxx ♥ x ♦ jxx ♣ kxxx]266|100|Scoring: IMP[/hv] Well in my usual methods the auction goes after 3D: 3H(Are you max or min?)-3S(Min)-4S(I am still interested in slam, are you really min?)-P(Yes I am really min) [hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx

 

here you want to be in 4♠. if you start with 3♥ (control ask) partner will bid 3♠ (min); if you follow up with 4♣ (really control ask) then partner shows three and you can no longer stop at the four-level. keycard will be 4♣ or above and will of course bypass 4♠.

 

dealer: west vul: both scoring: imp

♠ kxx ♥ aqx ♦ akx ♣ ajxx ♠ axxxx ♥ x ♦ qxx ♣ kqxx

 

here the goal is to reach 6♣, which is excellent, especially with west declaring. this is relatively easy to reach as long as you don't sign off too early... but note that this hand will not be distinguished from the first hand until you're well past 4♠.

 

dealer: west vul: both scoring: imp

♠ aqxx ♥ kx ♦ akxxx ♣ ax ♠ kxxxx ♥ axx ♦ qxx ♣ xx

 

here 7♠ has reasonable chances. again this is quite reachable, but consider what happens if you start with 3♥. partner bids 3♠ (minimum), then 4♥ (three controls), then 5♦ (a top spade, a top heart, no top diamond), then 5♠ (no spade queen)... and you are easily at the six level before you can distinguish this hand from the hand below. if you start with 4♣ (rkc in spade) then you get 4♠ (two no queen), then 4nt would ask for the heart king, 5♣ for the diamond king, 5♦ for the club king, 5♥ for the heart queen... and again you'll be at the six level before you know about the diamond queen.

 

dealer: west vul: both scoring: imp

♠ aqxx ♥ kx ♦ akxxx ♣ ax ♠ kjxxx ♥ axx ♦ xxx ♣ jx

 

here no slam has real play, though 5♠ should almost always make. note that this hand will make all the same bids as the hand above it until you are past 5♠!

 

-------

 

admittedly these are made-up examples. next time i practice relaying with someone i'll try to grab some more "random" hands. however, my feeling is that this issue arises fairly often. the main problem is when one of:

 

(1) you're not sure of the proper trump suit, because a 5-3 or 5-4 fit is likely to be "safest" but your only possible slam is in a 4-4 fit in another suit. if you relay for controls it is hard to find out about the important queens, but if you bid keycard you must commit to a trump suit too early.

 

(2) the key card for a making slam is a queen outside the trump suit, which covers a critical loser in declarer's hand.

 

(3) you really want a "quantitative" invite to slam in a suit, but the "max" partner needs to accept is not so strong that partner would pull a signoff attempt.

 

another nice thing about akq points is that it reduces the frequency of game forcing "negative" responses. assuming you require 2 controls for a positive, there are hands like k+q+q+q that are quite good but have to start with a negative. you can still construct such hands if you require 4+ akq points for a positive, but there are fewer of them (i.e. k+q+j+j+j+j or q+q+q+j+j+j).

dealer: west vul: both scoring: imp

♠ kxx ♥ aqx ♦ akx ♣ ajxx ♠ ajxxx ♥ x ♦ jxx ♣ kxxx]266|100|Scoring: IMP[/hv] 3H(Are you max or min?)-3S(Min)-4S(I am still interested in slam, are you really min?)-5C(I have extras, but only 1 keycard)-6C-P (We can also bid 5N over 5C there to ask responder to bid the cheapist suit where he has an undisclosed honor card, and starts a choice of slam auction) [hv=d=w&v=b&w=skxxhaqxdakxcajxx&e=sajxxxhxdjxxckxxx

 

here you want to be in 4♠. if you start with 3♥ (control ask) partner will bid 3♠ (min); if you follow up with 4♣ (really control ask) then partner shows three and you can no longer stop at the four-level. keycard will be 4♣ or above and will of course bypass 4♠.

 

dealer: west vul: both scoring: imp

♠ kxx ♥ aqx ♦ akx ♣ ajxx ♠ axxxx ♥ x ♦ qxx ♣ kqxx

 

here the goal is to reach 6♣, which is excellent, especially with west declaring. this is relatively easy to reach as long as you don't sign off too early... but note that this hand will not be distinguished from the first hand until you're well past 4♠.

 

dealer: west vul: both scoring: imp

♠ aqxx ♥ kx ♦ akxxx ♣ ax ♠ kxxxx ♥ axx ♦ qxx ♣ xx

 

here 7♠ has reasonable chances. again this is quite reachable, but consider what happens if you start with 3♥. partner bids 3♠ (minimum), then 4♥ (three controls), then 5♦ (a top spade, a top heart, no top diamond), then 5♠ (no spade queen)... and you are easily at the six level before you can distinguish this hand from the hand below. if you start with 4♣ (rkc in spade) then you get 4♠ (two no queen), then 4nt would ask for the heart king, 5♣ for the diamond king, 5♦ for the club king, 5♥ for the heart queen... and again you'll be at the six level before you know about the diamond queen.

 

dealer: west vul: both scoring: imp

♠ aqxx ♥ kx ♦ akxxx ♣ ax ♠ kjxxx ♥ axx ♦ xxx ♣ jx

 

here no slam has real play, though 5♠ should almost always make. note that this hand will make all the same bids as the hand above it until you are past 5♠!

 

-------

 

admittedly these are made-up examples. next time i practice relaying with someone i'll try to grab some more "random" hands. however, my feeling is that this issue arises fairly often. the main problem is when one of:

 

(1) you're not sure of the proper trump suit, because a 5-3 or 5-4 fit is likely to be "safest" but your only possible slam is in a 4-4 fit in another suit. if you relay for controls it is hard to find out about the important queens, but if you bid keycard you must commit to a trump suit too early.

 

(2) the key card for a making slam is a queen outside the trump suit, which covers a critical loser in declarer's hand.

 

(3) you really want a "quantitative" invite to slam in a suit, but the "max" partner needs to accept is not so strong that partner would pull a signoff attempt.

 

another nice thing about akq points is that it reduces the frequency of game forcing "negative" responses. assuming you require 2 controls for a positive, there are hands like k+q+q+q that are quite good but have to start with a negative. you can still construct such hands if you require 4+ akq points for a positive, but there are fewer of them (i.e. k+q+j+j+j+j or q+q+q+j+j+j).

dealer: west vul: both scoring: imp

♠ kxx ♥ aqx ♦ akx ♣ ajxx ♠ ajxxx ♥ x ♦ jxx ♣ kxxx]266|100|Scoring: IMP[/hv]   Here we have 2 relatively balanced hands, and the question is 6 or 7, and responder is relatively weak (few features to show). This is the situation where AKQ methods are much much better than controls.   Josh

[Free]

I also don't understand why you ask for min/max. In the above hands, SP has a clear advantage over your methods:

- The first hand, responder has only a semi-positive hand opposite a moscito 1♣ opener (15+), so you'll probably stay out of slam for sure. The second hand is a GF, since the Queens count as SP.

- In the 3rd and 4th hand you see the same thing happening: min/max unnecessary wastes LOTS of bidding space, and the 4th hand again is only a semi-positive.

 

Without the min/max you win around 4 steps on average, which is almost an entire level! Imo it's quite useful to skip that question and immediatly start by asking exact controls. The scans will point out if partner is max or min, and you will even stay lower when you use a 'step+1' to skip 1 round of DCB's and immediatly start with scanning for Queens.

 

With the 'control'-method I describe, you can gain a lot of bidding space over SP+DCB since you can skip an entire scan. However, SP usually tend to show a difference in min/max a lot faster.

[joshs]

I also don't understand why you ask for min/max. In the above hands, SP has a clear advantage over your methods:

- The first hand, responder has only a semi-positive hand opposite a moscito 1♣ opener (15+), so you'll probably stay out of slam for sure. The second hand is a GF, since the Queens count as SP.

- In the 3rd and 4th hand you see the same thing happening: min/max unnecessary wastes LOTS of bidding space, and the 4th hand again is only a semi-positive.

 

Without the min/max you win around 4 steps on average, which is almost an entire level! Imo it's quite useful to skip that question and immediatly start by asking exact controls. The scans will point out if partner is max or min, and you will even stay lower when you use a 'step+1' to skip 1 round of DCB's and immediatly start with scanning for Queens.

 

With the 'control'-method I describe, you can gain a lot of bidding space over SP+DCB since you can skip an entire scan. However, SP usually tend to show a difference in min/max a lot faster.

Whether the last hand is a semi-positive or not depends on:

a. how light your 1C opening starts

b. how agressively you force to game over it

 

In Moscito (you sound like a moscito player) 1C starts at a 4432 15 count (Bergen 23), and this responding hand is just a good semi-positive.

 

In TOSR, 1C starts about the same point, but we force to game with a balanced 9 count opposite it.

 

In my non-transfer opening strong club system, 1C starts at a balanced 16 (we downgrade some 4333's but not many) or a 15 with a good 5 card suit or with shape (bergen 24's and good bergen 23's) and this hand is a minimum positive.

 

In what adam et all play, 1C starts at a balanced 17 or an unbalanced 16, and they would force to game without the club J.

 

Anyway, there are advatages and disadvantages to a direct run on to SP. Since they start at 4 for a positive when 1C is of the heavier variety, you are using up many steps to show the sp's for mediocre hands. For instance this 11 count: xx Axx Qxx Axxxx has 7 sp's so when opener with AQ KQxxx xxx KQJ relays with 3H the 4D relay response precludes playing 4H so you are past your safety level.

 

The optimization problem is: Determine a method that can determine if slam is a good bet at or below 4C as often as possible (and 3N or below is even better if 3N is a possible contract). And if you can't figure that out in time, have some way where you can stop at 4M or 5m if partner is minimum for previous bidding and not if he is max. Only rarely do you want to have to play 5M or 5N. (I have played 5N twice this year on 16's opposite 16's without a fit)

 

 

And yes, these hands are easier in the most recent version of moscito, since opener will limit himself (it doesn't matter who makes the quantitative bid, but many auctions need a quantitative bid...).... The problem's with moscito occur in the part score auctions.

 

Josh

[lowerline]

[hrothgar]

From these examples one might conclude that Moscito is better than Symmetric Relay in bidding slams! And I thought the real benefit of Moscito was when you don't open 1♣...

As alway, I'd like to make an argument in favor of more precision:

 

MOSCITO is a bidding system. MOSCITO can be contrasted with Precision, Viking Club, Acol, yada, yada, yada..

 

Symmetric Relay is relay structure. Symmetric Relay can be contrasted with nummeric, Ice relay, Albarran, ....

 

Denial cue bidding is an auction termination mechanism. Denial Cue Bidding can be contrasted with RKCB, CABS, etc...

 

To some extent, these structures can be mixed and matched. In theory, there's nothing to prevent me from playing MOSCITO using Ice Relay. (It would confuse people, but it wouldn't impact the core of the system).

 

For what its worth, I don't think that these hands demonstract that MOSCITO is "better" than a Symmetric Relay based precision style. Any bididng system has its cracks... Hands that fall into them won't necessarily be handled well. I don't find it surprising that hands which are intended to demonstrate problems for one scheme are handled easily by a different bidding systems which uses different demaractions between positives/semi-positives/neagatives.

[joshs]

Here's some examples. In each case, west is the dealer. I'll assume methods where 1♣ shows sixteen or more points, and a positive response is game forcing with 9+ points and 2+ controls. In each case responder's exact shape will resolve (playing symmetric or transfer-symmetric relays) with a 3♦ call. Let's suppose your options are control asks (3♥) or keycard in the suit of your choice (4♣+), as well as various "signoff" sequences that responder will not pull with any of the hands given.

 

<!-- EASTWEST begin --><table border='1'> <tr> <td> <table> <tr> <td>Dealer:</td> <td> West </td> </tr> <tr> <td>Vul:</td> <td> Both </td> </tr> <tr> <td>Scoring:</td> <td> IMP </td> </tr> </table> </td> <td> <table> <tr> <th> <table> <tr> <th class='spades'>♠</th> <td> Kxx </td> </tr> <tr> <th class='hearts'>♥</th> <td> AQx </td> </tr> <tr> <th class='diamonds'>♦</th> <td> AKx </td> </tr> <tr> <th class='clubs'>♣</th> <td> AJxx </td> </tr> </table> </th> <th> <table> <tr> <th class='spades'>♠</th> <td> AJxxx </td> </tr> <tr> <th class='hearts'>♥</th> <td> x </td> </tr> <tr> <th class='diamonds'>♦</th> <td> Jxx </td> </tr> <tr> <th class='clubs'>♣</th> <td> Kxxx </td> </tr> </table> </th> </tr> </table> </td> <td>  </td> </tr> </table><!-- EASTWEST end -->

 

Here you want to be in 4♠. If you start with 3♥ (control ask) partner will bid 3♠ (min); if you follow up with 4♣ (really control ask) then partner shows three and you can no longer stop at the four-level. Keycard will be 4♣ or above and will of course bypass 4♠.

 

<!-- EASTWEST begin --><table border='1'> <tr> <td> <table> <tr> <td>Dealer:</td> <td> West </td> </tr> <tr> <td>Vul:</td> <td> Both </td> </tr> <tr> <td>Scoring:</td> <td> IMP </td> </tr> </table> </td> <td> <table> <tr> <th> <table> <tr> <th class='spades'>♠</th> <td> Kxx </td> </tr> <tr> <th class='hearts'>♥</th> <td> AQx </td> </tr> <tr> <th class='diamonds'>♦</th> <td> AKx </td> </tr> <tr> <th class='clubs'>♣</th> <td> AJxx </td> </tr> </table> </th> <th> <table> <tr> <th class='spades'>♠</th> <td> Axxxx </td> </tr> <tr> <th class='hearts'>♥</th> <td> x </td> </tr> <tr> <th class='diamonds'>♦</th> <td> Qxx </td> </tr> <tr> <th class='clubs'>♣</th> <td> KQxx </td> </tr> </table> </th> </tr> </table> </td> <td>  </td> </tr> </table><!-- EASTWEST end -->

 

Here the goal is to reach 6♣, which is excellent, especially with west declaring. This is relatively easy to reach as long as you don't sign off too early... but note that this hand will not be distinguished from the first hand until you're well past 4♠.

 

<!-- EASTWEST begin --><table border='1'> <tr> <td> <table> <tr> <td>Dealer:</td> <td> West </td> </tr> <tr> <td>Vul:</td> <td> Both </td> </tr> <tr> <td>Scoring:</td> <td> IMP </td> </tr> </table> </td> <td> <table> <tr> <th> <table> <tr> <th class='spades'>♠</th> <td> AQxx </td> </tr> <tr> <th class='hearts'>♥</th> <td> Kx </td> </tr> <tr> <th class='diamonds'>♦</th> <td> AKxxx </td> </tr> <tr> <th class='clubs'>♣</th> <td> Ax </td> </tr> </table> </th> <th> <table> <tr> <th class='spades'>♠</th> <td> Kxxxx </td> </tr> <tr> <th class='hearts'>♥</th> <td> Axx </td> </tr> <tr> <th class='diamonds'>♦</th> <td> Qxx </td> </tr> <tr> <th class='clubs'>♣</th> <td> xx </td> </tr> </table> </th> </tr> </table> </td> <td>  </td> </tr> </table><!-- EASTWEST end -->

 

Here 7♠ has reasonable chances. Again this is quite reachable, but consider what happens if you start with 3♥. Partner bids 3♠ (minimum), then 4♥ (three controls), then 5♦ (a top spade, a top heart, no top diamond), then 5♠ (no spade queen)... and you are easily at the six level before you can distinguish this hand from the hand below. If you start with 4♣ (RKC in spade) then you get 4♠ (two no queen), then 4NT would ask for the heart king, 5♣ for the diamond king, 5♦ for the club king, 5♥ for the heart queen... and again you'll be at the six level before you know about the diamond queen.

 

<!-- EASTWEST begin --><table border='1'> <tr> <td> <table> <tr> <td>Dealer:</td> <td> West </td> </tr> <tr> <td>Vul:</td> <td> Both </td> </tr> <tr> <td>Scoring:</td> <td> IMP </td> </tr> </table> </td> <td> <table> <tr> <th> <table> <tr> <th class='spades'>♠</th> <td> AQxx </td> </tr> <tr> <th class='hearts'>♥</th> <td> Kx </td> </tr> <tr> <th class='diamonds'>♦</th> <td> AKxxx </td> </tr> <tr> <th class='clubs'>♣</th> <td> Ax </td> </tr> </table> </th> <th> <table> <tr> <th class='spades'>♠</th> <td> KJxxx </td> </tr> <tr> <th class='hearts'>♥</th> <td> Axx </td> </tr> <tr> <th class='diamonds'>♦</th> <td> xxx </td> </tr> <tr> <th class='clubs'>♣</th> <td> Jx </td> </tr> </table> </th> </tr> </table> </td> <td>  </td> </tr> </table><!-- EASTWEST end -->

 

Here no slam has real play, though 5♠ should almost always make. Note that this hand will make all the same bids as the hand above it until you are past 5♠!

 

-------

 

Admittedly these are made-up examples. Next time I practice relaying with someone I'll try to grab some more "random" hands. However, my feeling is that this issue arises fairly often. The main problem is when one of:

 

(1) You're not sure of the proper trump suit, because a 5-3 or 5-4 fit is likely to be "safest" but your only possible slam is in a 4-4 fit in another suit. If you relay for controls it is hard to find out about the important queens, but if you bid keycard you must commit to a trump suit too early.

 

(2) The key card for a making slam is a queen outside the trump suit, which covers a critical loser in declarer's hand.

 

(3) You really want a "quantitative" invite to slam in a suit, but the "max" partner needs to accept is not so strong that partner would pull a signoff attempt.

 

Another nice thing about AKQ points is that it reduces the frequency of game forcing "negative" responses. Assuming you require 2 controls for a positive, there are hands like K+Q+Q+Q that are quite good but have to start with a negative. You can still construct such hands if you require 4+ AKQ points for a positive, but there are fewer of them (i.e. K+Q+J+J+J+J or Q+Q+Q+J+J+J).

Playing Moscito with AKQ-ask, none of the example hands pose a problem. But it is not the AKQ-ask itself that makes the difference. It is because the AKQ point count is used to determine the difference between a positive or semi-positive response. In the first and last example responder will show a semi-positive which keeps you from getting too high. In the second and third example responder will show a positive in Moscito.

From these examples one might conclude that Moscito is better than Symmetric Relay in bidding slams! And I thought the real benefit of Moscito was when you don't open 1♣...

 

OTOH if responder always has to clarify his range, that is a flaw in your system. It is better to use the step as control ask (responder should always show his controls even with minimal strength) and the step+1 as a weak relay (responder shows a mininum or runs on with a maximum).

 

This will solve your problems in the last two examples, provided you apply the optimalization suggested by Echognome in this thread (start showing queens from the nth suit onwards where n is the number of control CARDS you have).

Responder will show 3 controls with 3NT. Now 4♣ starts the denial cuebidding.

Since responder has 2 control cards (K of spades and A of hearts) he starts showing queens from his second suit onwards and bids 4NT in the third example(showing K of spades, ace of hearts and queen of diamonds) and 4♠ in the forth example (not Q diamond).

 

Unfortunately this will not solve your problems in the first two examples, unless you have an agreement to show a maximum with 3 controls and more than 6 AKQ points (AK and 2Qs or KKK and 1Q at least). Then you can show a non-minimal hand on the second example and get to 6C.

 

Steven

Well first of all moscito uses symmetric relay.

 

The main features of moscito:

Light Forcing club opening, with symmetric relay continuations (Residues are shown in reverse lexagraphic order which is backwords from normal, but it doesn't really matter)

Immediate Semi-Positives and double Neg (This is new in moscito)

x-fer openings, with flexible symmetric relay continuations (majors are shown in natural order, as is 4M-6m )

Agreesive 4 card majors (can be balanced) with agressive 3 card raises (Paul likes mps more than I do).

 

 

In any case, you are misssing the point: If you require only 15 for the strong club and more like 10 for a positive response (so that a min positive has 5 SP instead of 4) then merely move a Q from openers hand to responders hand in my example, and you have the same problem.

 

Every system will have a an awkward range to deal with, so its a question of minimizing the awkwardness.

 

Josh

 

p.s. There are many interesting paradoxes in relay methods. I used to play that there was no final relay (ok maybe 7H). E.G. If partner bid 7C, 7D was still a relay. We soon discovered that at about the 5H/5S level we often couldn't afford to make one more ask, since a negative response might prevent us from signing off in 6m. So ironically, we found that having no relays above 5N actually helped us find grands on average, since it let us relay one more time safely. (Of course when our suit is spades, we are better off using 6 level relays, but there is no way of setting trumps in DCB methods...) But this observation was merely "on average". Just the other day I was playing the old method (you can keep relaying at the 6 level) and that enabled partner to find a key J and bid a grand. (We had a spade fit). Some methodfs are better for some hands and other are better for others...

[lowerline]

From these examples one might conclude that Moscito is better than Symmetric Relay in bidding slams! And I thought the real benefit of Moscito was when you don't open 1♣...

As alway, I'd like to make an argument in favor of more precision:

 

MOSCITO is a bidding system. MOSCITO can be contrasted with Precision, Viking Club, Acol, yada, yada, yada..

 

Symmetric Relay is relay structure. Symmetric Relay can be contrasted with nummeric, Ice relay, Albarran, ....

 

Denial cue bidding is an auction termination mechanism. Denial Cue Bidding can be contrasted with RKCB, CABS, etc...

 

To some extent, these structures can be mixed and matched. In theory, there's nothing to prevent me from playing MOSCITO using Ice Relay. (It would confuse people, but it wouldn't impact the core of the system).

 

For what its worth, I don't think that these hands demonstract that MOSCITO is "better" than a Symmetric Relay based precision style. Any bididng system has its cracks... Hands that fall into them won't necessarily be handled well. I don't find it surprising that hands which are intended to demonstrate problems for one scheme are handled easily by a different bidding systems which uses different demaractions between positives/semi-positives/neagatives.

I disagree. Symmetric Relay is a strong club bidding system first published in 1980 by Walter Jones and Roy Kerr. It was the relay structure that people remembered, but it was presented as a complete system.

 

The opening structure of "Symmetric" is something like:

1♣ = 16+ any shape

1♦ = 11-15 2suited (1♥ or 1NT as relay)

1♥/♠ = 11-15 5+ (1NT as relay)

1NT = 12-15 balanced (2♣ as relay)

2♣/♦ = 11-15 6+

 

When I refer to Moscito I mean the latest Paul Marston variant:

1♣ = 15+ any shape

1♦ = 9-14 4+♥

1♥ = 9-14 4+♠

1♠ = 9-14 4+♦

1NT = 12-14 balanced

2♣ = 9-14 6+♣

 

There is a big difference when you don't open 1♣. There is also an important difference in the responses to 1♣. In Symmetric you respond 1♦ with 0 or 1 control. All the rest is positive with 2+ controls. In Moscito you respond 1♠ with 0-2 QPs, 1♦ with 6+ QPs and most of the other responses are semi-positives with 3-5 QPs. The Moscito approach is superior (IMHO) because of the use of queen points to differentiate the 3 ranges after the big club opening. In the given examples it is exactly the ability to differentiate a positive from a semi-positive that allows you to get to the right contract, not the relay structure and not the denial cuebidding.

 

Ofcourse you can use this in your symmetric system as well, but this has a bigger impact on your system as you might think. First you must sacrifice some symmetry, second you must decide whether you stick with the AK-ask or switch to the AKQ-ask and change your denial cuebidding accordingly, third you must ask yourself if it still makes sense to ask for min/max when using AKQ.

 

And there is also the possibility to have it all and relayer decides which to use...

 

Steven

[Free]

That relaystructure you mention where 1♦ is GF and 1♠ is double negative has nothing to do with moscito. It just happens moscito uses this, but before it used symmetric relays. Me and my f2f partner still use symmetric relays. MOSCITO is a term used to describe the system style, not the responses. 'Major Oriented Strong Club In Transfer Opening' which basicly means a form of MAFIA style transfer openings combined with a strong ♣. 'Moscito relays' however are another kind of relays, which aren't symmetric, but those have nothing to do with moscito itself, it can be adopted over any type of openings. These are not as easy as symmetric imo, and also not as efficient I think.

 

Still, after shape is known, you can again choose what slammethods you use. It's completely independent of your relay system, except end signals perhaps. When you have relays going up till 4♥, you need other auction termination methods than if all your auctions end under 3NT. Responding positive or semipositive isn't such a big difference after all, but when you waste an entire level to know if responder is minimum (which is most frequent) and still don't know how many controls he has, you're clearly looking at an inferior method. I've been playing AKQ ask and denial cues for a few years now, and I haven't had much problems in determining slam or not, since I don't need to ask if partner is min or max before I know his 'useful strength'.

 

To give you an example of improving the method: show your controls first, and if you have the space you can still invite with a bid below game and which isn't a relay. Example with AKQ points: shape shown with 3♦, 6SP shown with 3♠, now relayer can bid 4♣ as start denial cues, 4M or higher to signoff, but 4♦ is still open, to use for slaminvite. When partner is minimum he just bids step 1, otherwise he starts denial cuebids with step 2 and higher. You only waste 2 steps, and you know if partner is min/max. Same thing can easily be used for AK controls, since the same principles apply: 4♦ is a useless bid.

[joshs]

From these examples one might conclude that Moscito is better than Symmetric Relay in bidding slams! And I thought the real benefit of Moscito was when you don't open 1♣...

As alway, I'd like to make an argument in favor of more precision:

 

MOSCITO is a bidding system. MOSCITO can be contrasted with Precision, Viking Club, Acol, yada, yada, yada..

 

Symmetric Relay is relay structure. Symmetric Relay can be contrasted with nummeric, Ice relay, Albarran, ....

 

Denial cue bidding is an auction termination mechanism. Denial Cue Bidding can be contrasted with RKCB, CABS, etc...

 

To some extent, these structures can be mixed and matched. In theory, there's nothing to prevent me from playing MOSCITO using Ice Relay. (It would confuse people, but it wouldn't impact the core of the system).

 

For what its worth, I don't think that these hands demonstract that MOSCITO is "better" than a Symmetric Relay based precision style. Any bididng system has its cracks... Hands that fall into them won't necessarily be handled well. I don't find it surprising that hands which are intended to demonstrate problems for one scheme are handled easily by a different bidding systems which uses different demaractions between positives/semi-positives/neagatives.

I disagree. Symmetric Relay is a strong club bidding system first published in 1980 by Walter Jones and Roy Kerr. It was the relay structure that people remembered, but it was presented as a complete system.

 

The opening structure of "Symmetric" is something like:

1♣ = 16+ any shape

1♦ = 11-15 2suited (1♥ or 1NT as relay)

1♥/♠ = 11-15 5+ (1NT as relay)

1NT = 12-15 balanced (2♣ as relay)

2♣/♦ = 11-15 6+

 

When I refer to Moscito I mean the latest Paul Marston variant:

1♣ = 15+ any shape

1♦ = 9-14 4+♥

1♥ = 9-14 4+♠

1♠ = 9-14 4+♦

1NT = 12-14 balanced

2♣ = 9-14 6+♣

 

There is a big difference when you don't open 1♣. There is also an important difference in the responses to 1♣. In Symmetric you respond 1♦ with 0 or 1 control. All the rest is positive with 2+ controls. In Moscito you respond 1♠ with 0-2 QPs, 1♦ with 6+ QPs and most of the other responses are semi-positives with 3-5 QPs. The Moscito approach is superior (IMHO) because of the use of queen points to differentiate the 3 ranges after the big club opening. In the given examples it is exactly the ability to differentiate a positive from a semi-positive that allows you to get to the right contract, not the relay structure and not the denial cuebidding.

 

Ofcourse you can use this in your symmetric system as well, but this has a bigger impact on your system as you might think. First you must sacrifice some symmetry, second you must decide whether you stick with the AK-ask or switch to the AKQ-ask and change your denial cuebidding accordingly, third you must ask yourself if it still makes sense to ask for min/max when using AKQ.

 

And there is also the possibility to have it all and relayer decides which to use...

 

Steven

Facenating. Now Moscito has been around for 15 years, but Paul Marsden just switched to the 1S double negative and the 1D full positive about 2 years ago (someone correct me if its older than that). So what system was he playing for the first 13 years??

 

I think there is some confusion about the "overall structure of a bidding system" vs "the implimentation of certain components of the system".

 

There are also, various inversions that are possible in symmetric relay that do not effect the symmetric nature of the system:

a. re-ordering of the suits

b. switching HML shortage to LMH

c. Switching the meaning of the 2H reverser to be "higher ranking suit is longer"

d. Making direct responses of 2S and higher 1 suited, and bid 2C then 2H or higher with both minors and a direct 2H as 3 suited

etc.

 

Essentially these just are a different lexagraphic order to the hand patterns. If you already play symmetric relay, you would not have to learn a new system to do these inversions. I could sit down and play with 2 minutes explanation as to the inversions....

 

Josh

[akhare]

In Symmetric Relay (the system) one asks for the number of controls after the shape is known. In Moscito one asks for the number of AKQ-points. This has consequences on what you deny during denial cuebidding.

 

Anyone tried both approaches? Which do you prefer?

Here's an unedited post by John Sheehan (of Prism Signals fame) about his GOFISH alterative to DCB when using RPs (aka AKQ points). It looked pretty playable the last time I looked:

 

Hi all,

 

Dlr: W

Vul: EW

Auction: Moscito Variation

 

West East

A863 K5

A652 K8

Q A4

KQJ5 A876432

 

1C 2D

2H 2S

2N 3C

3D 3H 2=2=2=7.

3S(a) 4S 10 relay points.

4N(:( 5S No. Go Fish!

5N© 6C Same rank.

7C

 

(a) 'Number of Relay points?' (A=3 K=2 Q=1)

(B) 'Do you have any queens?'

© 'Tell me about your two kings.'

 

Note(B): The first step after the relay points response

asks,'Do you have any queens?'

 

In response to the queen ask:

Step 1 = 1 Q.

Step 2 = 2 Q's.

Step 3 = 3 Q's.

Step 4 = 0 Q's. Go Fish! (or 4 Q's)

 

If The number of relay points is known; and,

The number of queens is known; then,

The exact number of aces and kings is known.

 

 

Note©: 'Tell me about your kings.'

 

The first step after the number of queens response

asks about location of kings.

 

(If you are more interested in queens

than kings, skip to the second step to

ask about the location of queens.)

 

In response to, 'Tell me about your kings.'

 

To show one or three kings, stop at the bid that

corresponds in the scanning order

to the one suit with (or without) the king.

 

 

If two kings are held; then

Step 1 = 2 K's of the same Rank

Step 2 = 2 K's of the same Color

Step 3 = 2 K's of the same Shape (C/H D/S)

 

 

 

 

Dlr: W

Vul: EW

Auction: Moscito Variation

 

West East

A7 KQ85

AQ3 K852

AQJ874 K6

J4 A96

 

1C 2D

2H 2S

2N 3C

3D 3H 4=4=3=2

3S(a) 4S 10 relay points.

4N(B) 5C 1 Queen.

5D© 5N No club king.

6C(d) 6D Spade queen.

7N

 

(a) 'Number of Relay points?' (A=3 K=2 Q=1)

(B) 'Do you have any queens?'

© 'Tell me about your three kings.'

(d) 'What about that queen?'

 

Is Go Fish! for relay points a practical approach to solving the problem

of discoverying and pinpointing honors in slam auctions?

 

Johnny Sheehan

www.prismsignals.com

[david_c]

A question for those of you that play AKQ-points: how do you adjust for honours being in short suits? (For example I've been playing that a singleton K or Q counts zero, but a singleton A is still 3.) I'm particularly interested in the value of a doubleton queen. I'm beginning to think that a doubleton queen should not count if it's in a long two-suited hand (ie. 5521 or 6520). For more balanced types I'm not so sure. Any thoughts?

[Free]

We play that singleton K or Q doesn't count, but Aces do. However, in some situations we can still show 'a' singleton honour (K or Q, not the Ace). For example in 7411's, you can first show high shortage to show K or Q in the lowest singleton. After semipositives you can still show more than 5 SlamPoints to show singleton K or Q. If opps double a bid in our short suit, we can RDbl to show a tophonour. But in most situations we just can't...

 

When you really need to know about the singleton holding, you have to use another slam method (RKC + SAB's)

 

Doubletons count like normal:

KQ = 3

Kx = 2

Qx = 1

Doubletons with a honour can be interesting (especially Kx when partner has Axxxxx and can ruff the suit free), so we don't adjust. Qx can be enough to make partner's suit high (AKJxxx).

 

It's not perfect in all situations, but it's ok in most cases :P