[hrothgar]

benlessard, on 2013-July-24, 17:59, said:

I think relaying a 10-14 pts opening hand is the most pertinent for everybody. What will work for 10-14 is likely to work for any limited range.

Ive started to scout for all unbal 10-15 on a slam or near slam hands on the WBF web page of the last tournament.

I disagree

If you are going to make an objective comparison of auction termination mechanisms, you need to control external sources of variance.
In this case, eliminate shape resolution and range information from the picture.

By focusing on very limited hands you are eliminating one of the major (relative) flaws of your methods (the fact that slam points provide additional information about the overall strength). Moreover, the fact that different methods will resolve shape at different levels means that we aren't just contrasting DCB and RKCB

The appropriate way to proceed is to

1. Define a hand for the strong club opener
2. Define a known shape for relay responder
3. Define the level at which that shape is resolved
4. Define a minimum strength for the relay responder such that investigating slam is reasonable

Deal 50 or so hands consistent with these conditions

Analyze the results and then start varying the starting conditions

If you want to be a bit more flexible from the get go, vary the the strength of the strong club opening as well while keeping the same shape.

[benlessard]

♠ AQJxx
♥ A
♦ xxx
♣ Axxx

♠ Kxxxx
♥ Qxxxx
♦ AKQ
♣ void

This one is cute for us too.

1C--1H (15+,S W or GF)
1S--1Nt (15-20, S+C GF)
2D--2H (C void at least 4s trumps (2C is normally forced the only reason to bypass is void), Reverser S=>C)
2S--3D (ask, 5134)
3H--4D (S rkc, 2+Q no KD)
4NT--5K (JofS ? yes but no J of D)
5H--5NT (extras ? --yes but only 1, (extras here are A of C or stiff K) IRL I dont think ive ever got the chance to ask for a K in a void.
7S

There is no need to ask for A of C because 7S is still good.

[benlessard]

I do agree that we all start at the same place (obvious), however I disagree about the non-limited part because most of the time the pts are limited, if you play a strong club 1D/1H/1S are going to be limited if responder is a passed hand hes limited. If you open a strong club & responder make a semi-pos bid and you managed to relay hes limited (not sure Moscito can do this however). In my system its the opener that is most of the time limited (15-20).

For me 19 times out of 20 relaying auction one side of the partnership is limited. Ex 1C-1D... 1H/1S/2C arent forcing because they are 15-20. So with a marginal GF responder can slow down easily.

I cannot speak for other system and I dont really care anyway since I need to do my minors where opener range is 15+ and responder is an unpassed hand.

(13)45 is 3C for me and a really annoying start for me would you take this as a good starting point ?

[Free]

nvm

[hrothgar]

benlessard, on 2013-July-24, 20:26, said:

(13)45 is 3C for me and a really annoying start for me would you take this as a good starting point ?

5-4-3-1 hands are relatively common and typically show rather than ask.
Seems like as good a shape for RR as any.

It seems reasonable to use a 3D bid for shape resolution, given that is the level at which symmetric relay resolves 5-4-3-1 patterns
Indeed, your last example showed showed a 5=1=3=4 resolving at 3♦.

I don't agree that we should limit RR's strength.
Once again, we want to purely compare RKCB with Denial Cue Bidding across a full set of hands rather than a constrained subset.

As a next step, we should probably

1. Fix the strength and shape of the relay asker
2. Decide whether we want to inspect ANY 5431 pattern or (for example) 3=5=1=4 patterns
3. Agree on an algorithm to fix the minimum strength for Relay Responder

Once we do so, I'll post a dealer script along with 50 or so hands that folks can bid them using their preferred auction termination mechanisms..

[hrothgar]

FWIW, I think that the most interesting results would be generated if

1. Opener holds a 3=5=(3-2) hand
2. Responder can hold any 5-4-3-1 pattern
3. The minimum combined strength is sufficient that the relay asker wants to actively explore slam rather than letting RR relay break with a max. (I suggest that relay asker has 12 slam points and relay responder has 6+)

This way, the best strain could include any of the 4 suits as well as NT

[hrothgar]

Also, I recommend setting up a separate thread for each of the hands that get bid.
It means a lot of threads to keep track of, but ultimately, it should be easier if multiple hands generate much discussion.

[hrothgar]

I generated a quick dealer script for the purpose of illustration (you should be able to run this script in BBO's hand dealer program)

predeal    north SKQ2, HAQT94, D82, CAKJ

Responder = (shape(south, any 5431) and spades(south) == 1 
            and 
            (3 * (hascard(south, AH) + hascard(south, AD) + hascard(south, AC))) +
            (2 * (hascard(south, KH) + hascard(south, KD) + hascard(south, KC))) +
            (1 * (hascard(south, QH) + hascard(south, QD) + hascard(south, QC))) >= 6)

            or

            (shape(south, any 5431) and hearts(south) == 1 
            and 
            (3 * (hascard(south, AS) + hascard(south, AD) + hascard(south, AC))) +
            (2 * (hascard(south, KS) + hascard(south, KD) + hascard(south, KC))) +
            (1 * (hascard(south, QS) + hascard(south, QD) + hascard(south, QC))) >= 6)
            
            or

            (shape(south, any 5431) and diamonds(south) == 1 
            and 
            (3 * (hascard(south, AH) + hascard(south, AS) + hascard(south, AC))) +
            (2 * (hascard(south, KH) + hascard(south, KS) + hascard(south, KC))) +
            (1 * (hascard(south, QH) + hascard(south, QS) + hascard(south, QC))) >= 6)

            or

            (shape(south, any 5431) and clubs(south) == 1 
            and 
            (3 * (hascard(south, AH) + hascard(south, AD) + hascard(south, AS))) +
            (2 * (hascard(south, KH) + hascard(south, KD) + hascard(south, KS))) +
            (1 * (hascard(south, QH) + hascard(south, QD) + hascard(south, QS))) >= 6)


condition Responder

action
print(south)

Here are a couple representative hands.

In each case, the relay asker will hold

♠KQ2
♥AQT94
♦ 82
♣ AKJ

Hand 1, relay responder holds

♠AJ85
♥K62
♦KT643
♣3

Hand 2, relay responder holds

♠ AJ985
♥ K
♦ A95
♣ T832

[benlessard]

One small caveat. Its 8 times more likely to have 15 or 16 hcp than 19 for a balanced pattern and 15-16 is 4 times more likely than 19-21 (if you play a weak NT) so I not sure about 19hcp represent a "base" starting hand.

But its ok for me and Im trusting you to do the shuffling if you want.

Normally 19 pts is where I start thinking about making an extra move (IMO its better to make an extra move with 19 to cover 13 pts by reponder than the other way around for frequency reasons.

GTG now ill clarify scanning my method later

[hrothgar]

benlessard, on 2013-July-25, 10:27, said:

One small caveat. Its 8 times more likely to have 15 or 16 hcp than 19 for a balanced pattern and 15-16 is 4 times more likely than 19-21 (if you play a weak NT) so I not sure about 19hcp represent a "base" starting hand.

But its ok for me and Im trusting you to do the shuffling if you want.

Normally 19 pts is where I start thinking about making an extra move (IMO its better to make an extra move with 19 to cover 13 pts by reponder than the other way around for frequency reasons.

GTG now ill clarify scanning my method later

With a minimum hand, the relay asker is going to use an auction termination mechanism rather than investigating slam.
Relay responder will relay break any time they hold 10+ slam points.
I thought that the 19 HCP hand was sufficiently strong that the relay asker would chose to investigate slam opposite many of the 5431 patterns.

I'm not particularly weed to this hand, this shape, or this strength if someone else would like to suggest an alternative.

[straube]

I hope you will test Parity Cue Bidding along with DCB and spiral scans. I'm actually not very sure how DCB and spiral scans are different, but I've the impression that with spiral scans the captain hand skips certain steps so as not to ask for cards he holds.

[akhare]

benlessard, on 2013-July-24, 19:42, said:

1C--1H (15+,S W or GF)
1S--1Nt (15-20, S+C GF)
2D--2H (C void at least 4s trumps (2C is normally forced the only reason to bypass is void), Reverser S=>C)
2S--3D (ask, 5134)
3H--4D (S rkc, 2+Q no KD)

What is the order of suits for the RKC ask? Is it by the relative length of suits with some tie breaker for equal length suits?

[Free]

hrothgar, on 2013-July-24, 17:38, said:

♠ AQJxx
♥ A
♦ xxx
♣ Axxx

♠ Kxxxx
♥ Qxxxx
♦ AKQ
♣ void

FWIW, I am attaching the MOSCITO auction because its a pretty amusing one
(Lots of different options in bidding the hand, but all roads lead to 7S)

1♣ - 1♦ (With a 5-5 shape and 9 slam points South's hand is a bit strong for a limited opening)
(1♦ = art GF, denies certain hand types)
1♥ - 2♣ (1♥ = Relay: Even with a minimum, opener prefers not to RR with a 5530 shape)
(2♣ = SS with Clubs or 2 suited with Blacks)
2♥ - 2♠ (2♥ = Reverse Relay, 0-1 Clubs and 4+ Spades)
(2♠ = Relay)
2N - 3♣ (2N = Spades and Hearts, 3♣ = Relay)
3♦ - 3H (3♦ = 5/5 in the majors, 3♥ = relay
3N - 4♣ (3N = 5=5=3=0, 4♣ = QP ask)
4♦ - 4♠ (4♥ = 9 slam points, 4♠ = DCB)
5♦ - 7♠ (5♦ = 1-2 Controls in ♠ + ♥, 0/3 Controls in Diamonds

Alternatively, I could use RKCB in Spades, in which case I'd bid 4S rather than 4C

Here, the auction would continue

5D - 5H (5D = 2 Keycards no queen, 5H = CAB in Hearts)
5N - 6C (5N = AK or Q in Hearts, 6C = CAB in Diamonds)
6N - 7S (6N = AKQ in Diamonds)

I really don't like this auction for shape resolution. I'd just show my shape with opener's hand because:
- the chance is real that partner shows a balanced hand after which you can't reverse relays anymore.
- you have ♦AKQ which will require only 1 scan to eliminate this suit completely. After that you only have 2 suits left to scan.
- you have a void, which makes denial cuebidding very poor (can't calculate the QP's in ♣, which might make other suits dubious), and afaik you don't have exclusion blackwood available. On this hand it doesn't matter because you have all keycards, but after 1♣-1♦-1NT you're screwed when you'd appear to miss 1 or 2 keycards.

This is the first time I see reversed relays in action after 1♣-1♦-1♥-something (not 1NT). I've seen them in the old scheme, used them a couple of times, but it doesn't make much sense to do it here with a minimum hand. Usually opener without extras starts showing shape immediately, and this hand is no exception imo. So reversed relays should show 12+ QP's in this case.

[hrothgar]

straube, on 2013-July-25, 11:47, said:

I hope you will test Parity Cue Bidding along with DCB and spiral scans. I'm actually not very sure how DCB and spiral scans are different, but I've the impression that with spiral scans the captain hand skips certain steps so as not to ask for cards he holds.

I'll be posting the hands. If you're interesting in parity cue bids, please feel free to participate.
(Personally, I don't know the method so I can't really bid them using PCBs)

[straube]

hrothgar, on 2013-July-25, 13:07, said:

I'll be posting the hands. If you're interesting in parity cue bids, please feel free to participate.
(Personally, I don't know the method so I can't really bid them using PCBs)

ok, I'm in for PCBs unless awm would like to do it. We learned it from him.

[blackshoe]

straube, on 2013-July-25, 11:47, said:

I hope you will test Parity Cue Bidding along with DCB and spiral scans. I'm actually not very sure how DCB and spiral scans are different, but I've the impression that with spiral scans the captain hand skips certain steps so as not to ask for cards he holds.

He can do that, yes. More to the point, if he does do that, he has affirmed that either he holds those cards or does not care about them.

[benlessard]

Quote

What is the order of suits for the RKC ask? Is it by the relative length of suits with some tie breaker for equal length suits?

Yes when its equal its HSCD. 1444 = the order is H,C,D. There is some case where opener has showned or denied 3 card support in the first suit. So if a 5521 & i know partner got 3S = S before H if he denied 3S its H before S. Sometimes we get a preference, ex show S+H and get a H preference, if shapes is 54,55,65, its going to be H before spades. 64 and 74 will be S before H. (but to check DCB vs scan its irrelevant... ill just mentionned it when ive lost a step to gloat on how good my general system is )

There is an improvement on spiral scan I want to try... S is trump

we are in 4C

4D= 1st ask
4H= 2nd ask
4S= to play
4Nt =4th ask !! (not the 3rd)

if you want to do the 3rd ask you bid 4H..even if you dont care about the 2nd card.
partner will...

return to 4S without 2nd over wich 4Nt is asking for 3rd.
bid 4nt with 2nd but not 3rd card
bid 5C with 2nd and 3rd but not 4th etc. So you lose no space whatsoever. The only drawback is really small.. sometimes partner bypassed the K of H and you have it, therefore you know hes got a stiff H and later when hes asking for extras you know that the Q elsewhere is better than the Q of H.

Same thing if H is trump and we are in 4C

4D... 1st ask
4H to play
4S = 3rd ask !! (not 2nd ask)

if i want for the 2nd ask I bid 4D even if Ive got no interest in the 1st card. partner will bid 4H and ill bid 4S asking for 2nd card.

-----------------------------

The other thing I want to check is asking for Aces not keycards. After Ace ask ill do KQ of trumps (instead of a Qtrump).. responses will be

0 or 2
1 without the next card
1, +1
1, +2
etc.

I hope that most of the time im going to have one of the big trump wich mean that my ask is for the KorQ missing. I also hope that if ive got none ill be able to figure 0 from 2 most of the times. If we are missing 1 I wont know if its the K or the Q but i hope it wont cost anything.

So ill do both standard RCK with Q trump ask (SSCan 1) and ACES ask with KQ trump parity ask (sscan 2). Ill be curious to check is 14,30,2 is good order when you only ask for Aces.

[hrothgar]

Free, on 2013-July-25, 12:36, said:

I really don't like this auction for shape resolution. I'd just show my shape with opener's hand because:
- the chance is real that partner shows a balanced hand after which you can't reverse relays anymore.
- you have ♦AKQ which will require only 1 scan to eliminate this suit completely. After that you only have 2 suits left to scan.
- you have a void, which makes denial cuebidding very poor (can't calculate the QP's in ♣, which might make other suits dubious), and afaik you don't have exclusion blackwood available. On this hand it doesn't matter because you have all keycards, but after 1♣-1♦-1NT you're screwed when you'd appear to miss 1 or 2 keycards.

This is the first time I see reversed relays in action after 1♣-1♦-1♥-something (not 1NT). I've seen them in the old scheme, used them a couple of times, but it doesn't make much sense to do it here with a minimum hand. Usually opener without extras starts showing shape immediately, and this hand is no exception imo. So reversed relays should show 12+ QP's in this case.

Just to clarify in case it isn't clear

Using the first relay break (in this case, showing shape over 1♦ rather than asking for shape) is limits hand strength. I've never played that failure to use a relay break changes the expectation for the minimum strength of the hand (for example, I'll ak with some quite weak balanced hands). It might be reasonable to play a style in which the failure to relay break resets the QP if you are later showing shape, however, I hadn't explicitly considered this.

The second option to apply a reverse relay occurs after RR has made his first bid showing shape. I use this relay break to show a misfit.

If RR has only promised one suit, the relay break shows a sngleton or void in that suit
If RR has shown a two suited patter the relay break shows the other two suits
If RR has show a balanced hand, relay break shows a balanced pattern, wishing to transition to natural bidding and explore stoppers

The goal of this relay break is to identify a misfit and be able to bail in NT

(I thought that I had posted these schedules before. If not, i can definitely respost them)

On this hand, when I initially looked at the 5530 hand, I said to myself "Well, its definitely a minimum but do I really want to show shape at such a high level"? Why don't I ask instead and see what happens. Then, when RR showed a club suit, my reaction was: This has just gotten a lot worse. Odds are that partner has the single suited in clubs. I'm going to relay break to show my singleton / void and (hopefully) slow him down. If partner has the two suiter with clubs and spades, he'll be well positioned to explore slam.

[hrothgar]

I am going to start posting blocks of hands.

I'm going to start 5 new threads with what I hope are obvious titles using the bidding code I posted earlier.

[straube]

It might be nice if you also started a thread for participants to describe their bidding methodologies including (after pattern has been shown) what S1, S2, S3, 3N, 4D, etc mean.

Something I've wondered is how you would address subjectivity regarding the following....let's say I know partner has 5413 and I can use DCB or RKC. The choice here is subjective and whether I do well depends on which choice I make as well as which RKC bid is assigned for which suit. It would also be tempting to reverse engineer outcomes and say "Well, I'd use RKC here" because it happens to give me the best result.

How do we get rid of this bias?

I think it would be most useful to only compare S1 continuations. I.e. I would only look at QP asking (my S1) and then see what PCB offers. Someone else might do control asking and then DCB. Etc.

Anyway, before you start offering deals, just a suggestion that we know what methodologies we're testing and make sure that a cross-comparison will give us useful results.