BBO Discussion Forums: Parity Cue Bids - BBO Discussion Forums

Jump to content

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

Parity Cue Bids

#1 User is offline   Cascade 

  • PipPipPipPipPipPipPipPip
  • Group: Yellows
  • Posts: 6,765
  • Joined: 2003-July-22
  • Gender:Male
  • Location:New Zealand
  • Interests:Juggling, Unicycling

Posted 2019-October-08, 15:06

I have seen lots of mention of awm's description of parity cue bids but try as I might I cannot find his original posts on the topic by searching. Does anyone with better search skills or more knowledge of this than I have links to his orginal post(s).

Thanks
Wayne Burrows

I believe that the USA currently hold only the World Championship For People Who Still Bid Like Your Auntie Gladys - dburn
dunno how to play 4 card majors - JLOGIC
True but I know Standard American and what better reason could I have for playing Precision? - Hideous Hog
Bidding is an estimation of probabilities SJ Simon

#2 User is offline   rbforster 

  • PipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 1,611
  • Joined: 2006-March-18

Posted 2019-October-08, 20:37

Here’s the basics. In a relay sequence, after asking for AKQ=321 points,

https://bridgewithda...IMprecision.pdf

Quote

Once RPs are shown we show parity of top three honors in each suit by halting if even parity and skipping onwards if odd parity. All suits will be scanned (excluding voids but including singletons) in order of length in describer’s hand with higher suits first in case of ties. In the case where RPs are shown with the first step (which has two possibilties), we stop with the minimum, and zoom into answering parity.
After showing all parities, we next shows the number of RP in the first suit with only one honor, if such suit exist, and stop if odd (A or Q) or skip a step if even (K). We may zoom into answering this, but we never zoom beyond this ask.
After answering the parity question, we next ask for jacks.


Also try searching just this forum (Non-Natural) for the word “parity”. You should see a collection of posts PCB## from 2017 where different hands are bid, as well as another set on Auction Termination from 2013 with examples. There are only 2-3 pages of threads in total.
1

#3 User is offline   Cascade 

  • PipPipPipPipPipPipPipPip
  • Group: Yellows
  • Posts: 6,765
  • Joined: 2003-July-22
  • Gender:Male
  • Location:New Zealand
  • Interests:Juggling, Unicycling

Posted 2019-October-09, 03:55

View Postrbforster, on 2019-October-08, 20:37, said:

After showing all parities, we next shows the number of RP in the first suit with only one honor, if such suit exist, and stop if odd (A or Q) or skip a step if even (K).


Thanks.

What happens if there is no such suit?

So I finish the parity showing responses at say 4H and partner relays with 4S, now 4NT says my first suit with one honour has A or Q and 5C says it is even. Do I have to bid 5D to say there is no such suit?
Wayne Burrows

I believe that the USA currently hold only the World Championship For People Who Still Bid Like Your Auntie Gladys - dburn
dunno how to play 4 card majors - JLOGIC
True but I know Standard American and what better reason could I have for playing Precision? - Hideous Hog
Bidding is an estimation of probabilities SJ Simon

#4 User is offline   foobar 

  • PipPipPipPipPip
  • Group: Full Members
  • Posts: 511
  • Joined: 2003-June-20
  • Gender:Male

Posted 2019-October-09, 10:48

View PostCascade, on 2019-October-09, 03:55, said:

Thanks.

What happens if there is no such suit?

So I finish the parity showing responses at say 4H and partner relays with 4S, now 4NT says my first suit with one honour has A or Q and 5C says it is even. Do I have to bid 5D to say there is no such suit?

The response showing the K-parity is tacked on to the last step in the parity scan. Say that we scanned a 5=4=2=2 shape and parity happened to even-even-even-even, with 4H as the terminal step. Since all suits have even parity, there's no presumptive candidate for the first suit with odd parity, so we stop at 4H with AQ in spades (or nothing) or bid 4S with the AK / KQ (longest+highest suit in the scan order). Also, note that when scanning singletons, we stop with a stiff AK(Q) and skip without (the latter is more likely and the decision on whether to skip the stiff Q depends on your QP evaluation methods).

BTW, awm had a recent minor tweak to address the problem of identifying KQ combinations in two suits with even parity, but am not completely sure of the exact mechanism.
0

#5 User is offline   straube 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,082
  • Joined: 2009-January-18
  • Gender:Male
  • Location:Vancouver WA USA

Posted 2019-October-09, 20:49

I think awm doesn’t show king parity with all even (Atul and I do) and I think he treats a singleton honor as odd (we treat as even for the reason Atul said).

They’ve actually stopped using the king parity question altogether. Instead, they ask to clarify which longest/ranking suits have the KQ in one and zero in the other whenever that possibility exists. Failing that, they ask a different question when K by itself is in one suit and the Q by itself is in another (again focusing on the longer/ranking suits). If neither question applies, the question doesn’t exist.

I think they’re trying to solve the “marriage problem” But Atul and I found using that addendum very difficult. We switched back to king parity.
0

#6 User is offline   Cascade 

  • PipPipPipPipPipPipPipPip
  • Group: Yellows
  • Posts: 6,765
  • Joined: 2003-July-22
  • Gender:Male
  • Location:New Zealand
  • Interests:Juggling, Unicycling

Posted 2019-October-11, 22:14

View Poststraube, on 2019-October-09, 20:49, said:

I think awm doesn’t show king parity with all even (Atul and I do) and I think he treats a singleton honor as odd (we treat as even for the reason Atul said).

They’ve actually stopped using the king parity question altogether. Instead, they ask to clarify which longest/ranking suits have the KQ in one and zero in the other whenever that possibility exists. Failing that, they ask a different question when K by itself is in one suit and the Q by itself is in another (again focusing on the longer/ranking suits). If neither question applies, the question doesn’t exist.

I think they’re trying to solve the “marriage problem” But Atul and I found using that addendum very difficult. We switched back to king parity.


After a strong club, what is your base QP for responder? (And what is your minimum strength for a positive response and for opener's 1?
Wayne Burrows

I believe that the USA currently hold only the World Championship For People Who Still Bid Like Your Auntie Gladys - dburn
dunno how to play 4 card majors - JLOGIC
True but I know Standard American and what better reason could I have for playing Precision? - Hideous Hog
Bidding is an estimation of probabilities SJ Simon

#7 User is offline   straube 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,082
  • Joined: 2009-January-18
  • Gender:Male
  • Location:Vancouver WA USA

Posted 2019-October-12, 07:48

My understanding of IMprecision is that its 1C base is 9 (and they combine 9 and 10 with the first step and separate with the next bid) but they use semipositives that range from 2-6 QPs (2-3 are combined similarly as the first step). They don’t have positives, so opener’s hand is relayed or they use their Opening NT system if opener rebids 1N.

It’s a very strong system. In a word, Adam and Sieong figured out an extremely useful reason to break relay...which is to suggest part scores and investigate/invite game. Another way of putting that is that if you play a relay system such as standard symmetric and find yourself rebidding 1S after a 1H response 80+ % of the time, you know you’re losing sequences.

Adam might explain the rationale differently, emphasizing how the club responses get distributional information quickly for the appropriately 5-11 hcp range of responding hands.
0

#8 User is offline   awm 

  • PipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 8,371
  • Joined: 2005-February-09
  • Gender:Male
  • Location:Zurich, Switzerland

Posted 2019-October-12, 08:15

A game forcing positive for a 16+ club would start at 4RP. We generally compute this by taking HCP for a balanced hand, subtracting two (jacks and or shape upgrades) and working out the lowest RP possibility. This gives:

5+ HCP (semi positive) for 2+ RP
9+ HCP (GF positive) for 4+ RP
11+ HCP (opening) for 5+ RP
17+ HCP (strong club) for 9+ RP (have to start using Kings here because there are only four queens; lowest RP way to get to 15 is 3K+3Q)

It’s possible to come up with hands satisfying the HCP without the RP of course, but they tend to be very “quacky” hands that may be worth downgrading. There are also hands that get upgraded (we open lots of unbal tens and some nines) but usually it’s bad to upgrade hands without controls so these upgrades tend to be higher in the RP front anyway.

All this said, Straube is correct that we do not generally play GF positive responses to 1C in the current incarnation of our system.
Adam W. Meyerson
a.k.a. Appeal Without Merit
0

#9 User is offline   straube 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,082
  • Joined: 2009-January-18
  • Gender:Male
  • Location:Vancouver WA USA

Posted 2019-October-12, 11:45

I forgot you have GF single-suited club responses.
0

#10 User is offline   DinDIP 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 117
  • Joined: 2008-December-13
  • Gender:Male
  • Location:Melbourne (the one in Australia not Florida)

Posted 2019-November-19, 22:05

I’ve been testing PCB and keep on running into marriage problems: how to distinguish between suits with two honours and those with zero.

Responder to an IMPrecision-style 1 held
AJxx A AJxxxxx x
Teller showed 5-2-2-4 with 11RP, then stopped four times to show an even number of honours in all four suits. Teller clearly has to have the CAK and KQ + KQ.

How can asker find out which? Even if using the refinements played by Sam and Adam -- If there is a pair of suits where one has KQ and the other has none, the first such pairing (length order, followed by high to low) is picked; stop if first suit has KQ, skip if second suit has KQ -- all asker finds out is that teller has the SKQ. How does asker find out which of the red suits is xx and which KQ?
What do experienced PCB users do?

Are cases like this exceptional it seems they arise only when asker has three suits with 0-1 honour? Are they just the equivalent of problem hands for other methods of denial cue bidding?
0

#11 User is offline   straube 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,082
  • Joined: 2009-January-18
  • Gender:Male
  • Location:Vancouver WA USA

Posted 2019-November-20, 00:12

It's been a long time since I used the alternative of DCB. Isn't the issue with DCB that it takes more steps for completion? DCB gives more concrete information, but it's more often information that opener may already know by looking at his hand. If I were a kibitzer or an opponent, I'd appreciate DCB being used more than PCB. So PCB is more ambiguous but more efficient when knowing many of the high cards that the slave hand doesn't have.

We run into marriage problems, too, but not often enough for me to stop thinking PCB is a great tool. Virtually all of our relay auctions proceed through it. That's partly, however, because I wanted us to master it before considering adjuncts to PCB. Also, AQ pairs are also (rarely) a problem.

PCB is primarily 5 questions, odd/even for each suit and an additional clue to disambiguate. We could add another question to help further disambiguate but at the cost of getting too high and delaying a jack ask. Not sure if the marriage problem is frequent enough to merit adding a 6th question. And would two final clues would we choose?

Wondering, however, if asker could break relay to ask clarification. Say teller bids 4S to show even spades. Asker foresees a marriage problem and bids 5C to ask if this even is 0 or if it is 2. Of course you'd have to have preset rules about when asker could skip suits to ask vs skip suits to suggest a contract. Conveniently, teller stops on even which is the point where asker could consider a relay break.
0

#12 User is offline   foobar 

  • PipPipPipPipPip
  • Group: Full Members
  • Posts: 511
  • Joined: 2003-June-20
  • Gender:Male

Posted 2019-November-21, 17:42

View PostDinDIP, on 2019-November-19, 22:05, said:

I’ve been testing PCB and keep on running into marriage problems: how to distinguish between suits with two honours and those with zero.

Responder to an IMPrecision-style 1 held
AJxx A AJxxxxx x
Teller showed 5-2-2-4 with 11RP, then stopped four times to show an even number of honours in all four suits. Teller clearly has to have the CAK and KQ + KQ.

How can asker find out which? Even if using the refinements played by Sam and Adam -- If there is a pair of suits where one has KQ and the other has none, the first such pairing (length order, followed by high to low) is picked; stop if first suit has KQ, skip if second suit has KQ -- all asker finds out is that teller has the SKQ. How does asker find out which of the red suits is xx and which KQ?
What do experienced PCB users do?

Are cases like this exceptional it seems they arise only when asker has three suits with 0-1 honour? Are they just the equivalent of problem hands for other methods of denial cue bidding?


Can you please post some hands that were especially problematic? This should be general interest to this thread and we might get clarity on the refinements if Adam or Sam chime in. FWIW, having switched to PCB from DCB, it's definitely seems superior AFAICT, but it'll be good to have an objective discussion.
0

#13 User is offline   straube 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,082
  • Joined: 2009-January-18
  • Gender:Male
  • Location:Vancouver WA USA

Posted 2019-November-22, 00:44

Let's say 4H asks PCB and 4S shows even. Can't think of a situation where asker now wants to sign off in 5C. Isn't it a free question?

If we define 5C as an ask, what should it ask? Like any of the following?

1) stop with 0, continue PCB with 2
2) stop with 2, continue PCB with 0
3) stop with 0, skip with 2, asker will then p/c
4) stop with 2, skip with 0, asker will then p/c
5) skip your next suit because asker has the AKQ
6) give K parity
0

#14 User is offline   foobar 

  • PipPipPipPipPip
  • Group: Full Members
  • Posts: 511
  • Joined: 2003-June-20
  • Gender:Male

Posted 2019-November-22, 12:17

View Poststraube, on 2019-November-22, 00:44, said:

Let's say 4H asks PCB and 4S shows even. Can't think of a situation where asker now wants to sign off in 5C. Isn't it a free question?


Is this a truly generalized situation? For example, what if the PCB for the longest suit (s), and relay captain was looking for AKQ to determine whether to bid 6?
0

#15 User is offline   straube 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,082
  • Joined: 2009-January-18
  • Gender:Male
  • Location:Vancouver WA USA

Posted 2019-November-22, 13:14

Well, the relay break (ask) could only apply after an even...so it wouldn't pertain to AKQ.

OTOH, you could certainly make the relay break ask anything. Could ask about the longest suit (for example) even if it's odd.

Basically, I think we have a free opportunity here and am looking for the best question to ask.
0

#16 User is offline   Cascade 

  • PipPipPipPipPipPipPipPip
  • Group: Yellows
  • Posts: 6,765
  • Joined: 2003-July-22
  • Gender:Male
  • Location:New Zealand
  • Interests:Juggling, Unicycling

Posted 2019-November-23, 04:09

View Poststraube, on 2019-November-22, 13:14, said:

Well, the relay break (ask) could only apply after an even...so it wouldn't pertain to AKQ.

OTOH, you could certainly make the relay break ask anything. Could ask about the longest suit (for example) even if it's odd.

Basically, I think we have a free opportunity here and am looking for the best question to ask.


1. If relayer was looking for cAKQ and you showed even then relayer knows you don't have AKQ so might now want to play in 5C.

2. More generally we want to know what the best question to ask is at each stage of the relay. I think there is some room for some work here to determine what are on average the best discriminators for slam and possibly reframe our questions.

3. It might be possible to always have two questions to ask, or conditionally to have two questions to ask based on what information you have received. This would come at the cost of a sign off bid.

Kerr and Jones had this conditional idea in some sequences. They had a rule that if partner responds at the four level showing five or more controls then as well as the normal one step relay, the relayer could bid 4NT asking for queens. The idea is that partner knows all of the controls often after such a response and instead of having to go through suits twice to find the queens can ask immediately for queens - whether none, one, or two controls are missing.

... 4 five controls
4 - keep relaying.
4NT - show me queens, their responses were 5 none, otherwise cue the queen with 5NT showing the club queen and 6 two queens.

They used the same principle with a 5NT bid (non-jump and maybe a jump over 4 - I don't know).

This idea could be extended to other situations, where you have two asks and whatever your style of cue bids - denial or parity or whatever.

I have had situations where the slave hand has gone so many steps that they must have what I want. But what say in the same situation I still need a key jack or something. If the bidding is low enough then it is all but impossible that I want to sign off below slam. Therefore we are free to have multiple asking bids up to at least 5NT under those conditions.
Wayne Burrows

I believe that the USA currently hold only the World Championship For People Who Still Bid Like Your Auntie Gladys - dburn
dunno how to play 4 card majors - JLOGIC
True but I know Standard American and what better reason could I have for playing Precision? - Hideous Hog
Bidding is an estimation of probabilities SJ Simon

#17 User is offline   straube 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 4,082
  • Joined: 2009-January-18
  • Gender:Male
  • Location:Vancouver WA USA

Posted 2019-November-23, 06:15

My hunch is that the opportunity for a relay break is small, that it should only be possible after even is shown for the first suit and that a skip here should ask for 0 or 2 after which asker passes or corrects. I can seldom imagine wanting to stop the scan at this point but can often imagine wanting very much to know whether partner has 0 or 2 in that suit. Using it after an odd/even response could work but I know more and we're higher; I think I would want that relay break as a sign off.

Any thoughts here, Adam? What would you test if anything?
0

#18 User is offline   DinDIP 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 117
  • Joined: 2008-December-13
  • Gender:Male
  • Location:Melbourne (the one in Australia not Florida)

Posted 2019-December-14, 18:23

A marriage query: awm's and sieong's rule states "If there is a pair of suits where one has KQ and the other has none, the first such pairing (length order, followed by high to low) is picked; stop if first suit has KQ, skip if second suit has KQ". Why does teller stop with what seems the more favourable answer? I presume it's a matter of frequency (teller is more likely to have honours in his long suit) so you want to preserve space in such situations. However, I ran into two hands yesterday where this order forced the bidding higher with the less helpful answer (less helpful because the partnership is more likely to play in teller's long suit -- this is especially true in auctions that start 1-1 where 1 shows a double negative or GF BAL or 7+RP UNBAL):

1. Teller held AJ9xxx AKQx K Kx and showed 12RP (we count kingletons as 1RP and stiff Qs as 0) and an odd parity in each suit. Teller's last bid was 5. Asker had xx xxxx Axxx AQx and now knew slam was worth bidding in the major in which teller had AKQ. However, the 5N marriage ask elicited a 6 response, and now we could no longer play in 6.

2. Teller held Axx AKQJxx Q Axx and showed 12RP and an odd parity in H, S and C, and no honour in D (we stop to show singleton honours). Teller then zoomed to answer the marriage question, but this was 5. Not a problem on the actual hand but asker, who again had no honours in either major, would have wanted to stop in 5 if teller had AKQ Axxxxx Q Axx.

Were these just random luck? I note that: they only occurred because asker had no honour in two suits; and if the first auction had been one step higher then stopping with the marriage in the longer/higher-ranking suit would work better. Or is redgrover right and we should switch the responses?
0

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users