Bridge Bidding

[ADavidBailey]

Thorvald and I were chatting about his interest in deals for game-level preemptive overcalls. He wanted to compare and contrast how various systems, conventions, and agreements worked with these deals. Here's my response:

I want to do exactly this for lots of scenarios.

• Jacoby_Super-Accept
• Jacoby_2N w/void vs good suit
• Mixed Raises in Competion vs ...
• Soloway Jump Shifts vs. Weak Jump shifts...
• Different styles of Weak 2 bids
  ... many, many more

How well are 500 deals of a single scenario bid with these different systems, conventions, and agreements? Which produces the better results?

I've written code to produce a summary of the 'notes' in BBA's archived pbn file and 'DD Bidding' from DDS's saved csv file. It's really crude and difficult to compare. It's made more difficult because every file does not have all possible 'note' or 'DD Bidding' value. Thus, the rows in my summaries don't align -- I have to manually cut/paste to fix it. Here's an example:

Jacoby_Super-Accept GIB61 vs EPBOT

I thought it was particularly interesting that GIB relied much more on Cue-Bidding and EPBot on Key Card.

All of the files in my GitHub BBA folder are efforts to compare/contrast different systems/conventions/agreements. It’s not going very well.

We’re all working on different aspects of the same problem.

• I’m creating deals for various bidding problems
• Edward is bidding the deals with various systems/conventions/agreements
• Bob is doing a through analysis of the results
• Thorvald uses the results to train BEN and discovers issues

Together, we can build a structure for a serious investigation of bidding and results. We can produce quantified metrics of the differences. Are you interested in pursuing this?

Each time I create or make a significant change to a scenario, I have python code to:

• Extract the Dealer Code from the BBOalert wrapper (via Python)
• Run the Dealer Code through BBO’s Dealer to produce a pbn file (via CMD file)
• Comment the Statistics (via Python)
• Rotate the pbn for 4-handed play (via Python)
• Create a rotated lin file for import to BBO (via Python)

For a single scenario, this takes a few minutes.

I’m working on a CMD file to run pbn file(s) through BBA to create the BBA archive. And then hope to run the BBA archive(s) through DDS (is this possible?)

Here are some things that I would like your help with:

• For comparing 500 similar deals, DDS needs to summarize the results at the bottom of the archived csv files. I'm sure it would greatly improve the summary.
• Same for the BBA notes
• BBA needs lots more 'notes' in the archive. The bidding sequence, would be nice, too. How many hands were bid 1N P 3N.. or 1M P 2N P 4?... -- maybe just the top 20 for each input file.

@Rick-Wilson join this discussion

[ADavidBailey]

I talked with Rick late yesterday afternoon. He pointed out that the bidding is in the BBA archive. I'll have to figure out how to use it.

[ADavidBailey]

Rick has been working on a command file to run files in my pbn folder through BBA. I'll see if I can include that in my chain of programs. If so, then each time I update a scenario (PBS folder), I'll create corresponding files in the following folders:

• dlr
• pbn
• pbn rotated
• lin rotated
• bba

I'll put the other stuff that's in the BBA folder somewhere else.

Bob, is it possible to run DDS from the command line with parameters for the input BBA pbn file and the output DDS csv file?

[ThorvaldAagaard]

I think it is fine using this for discussions, but perhaps you should tag persons you want comments from like this
@ADavidBailey

• but I am not sure we will get notified

I am currently more focused on fixing bugs in the robot as the bidding is fine (Except for the issues I have created for Edward to solve)

[ADavidBailey]

Thanks, Thorvald. One deal at a time... :)

Both @Rick-Wilson and I are really interested in looking at lots of deals.

Rick made a suggestion that had not occurred to me. He suggested reading files from my (soon to be) bba folder and filtering those that had the desired bidding sequence and par result. (Rick's interest is in files for teaching)

[ThorvaldAagaard]

I have made several filtering scripts.
Currently I remove some "bad" boards before training BEN, and it has improved the bidding.

I am using PBN-files as input, so that could be a good starting point for you.
https://github.com/lorserker/ben/blob/main/src/badcontracts.py
but it use some other libraries from BEN (Including Double Dummy), so perhaps you will need some help

[EdwardPiwowar]

I see this thread on my site and I am waiting for events to develop, but in reality I am working hard on eliminating the regression.

[ADavidBailey]

Thanks, Edward. Hopefully something will come along to interest you.

[ADavidBailey]

@EdwardPiwowar I trying to run BBA from the makeBBA.cmd -- here are the contents:

P:\BBA.exe --ARCHIVE_FILE !P:\pbn\2N_Smolen_Leveled-S.bba! --AUTOBID --HAND !2N_Smolen_Leveled-S.pbn! --CC1 GIB-ADB.bbsa -- CC2 GIB-ADB.bbsa --DD 0 --SD 1
P:\BBA.exe --ARCHIVE_FILE !P:\pbn\Opps_2-Suited_Overcalls-S.bba! --AUTOBID --HAND !Opps_2-Suited_Overcalls-S.pbn! --CC1 GIB-ADB.bbsa -- CC2 GIB-ADB.bbsa --DD 0 --SD 1
P:\BBA.exe --ARCHIVE_FILE !P:\pbn\SCS_1C_any_0-4_Resp-S.bba! --AUTOBID --HAND !SCS_1C_any_0-4_Resp-S.pbn! --CC1 GIB-ADB.bbsa -- CC2 GIB-ADB.bbsa --DD 0 --SD 1
P:\BBA.exe --ARCHIVE_FILE !P:\pbn\Splinters-S.bba! --AUTOBID --HAND !Splinters-S.pbn! --CC1 GIB-ADB.bbsa -- CC2 GIB-ADB.bbsa --DD 0 --SD 1

when I run it, the BBA program opens and it starts bidding the hands. It doesn't stop after 500 (the number of deals in each file). I click start and it stops. I then close the BBA window and it pops again. The same thing happens 4 times.

I notice as it's running each file that the CC's are NOT as specified in makeBBA.cmd

Do you have any suggestions?

[ThorvaldAagaard]

Change
!P:\pbn\Splinters-S.bba! to P:\pbn\Splinters-S.bba
The example I gave you was with variables
Also for input

[ADavidBailey]

@EdwardPiwowar I see that I don't have the right path for the CCs.

[ADavidBailey]

@ThorvaldAagaard it's better, now. It opens the BBA UI runs through 1 file and stops. When I close the BBA UI it runs the next. Lots of other issues; but, I have a question. Should it open the UI? I sure hope NOT!!

I'm taking a break for a while.

[ThorvaldAagaard]

It will open the UI, but you can have multiple running

I created a cmd file, that will start multiple BBA

@echo off
setlocal enabledelayedexpansion

for /L %%i in (1,1,6) do (
    set "input_file=.\input_%%i.bba"
    set "output_file=output_%%i"
    
    set "command=start "BBA" /D . BBA.exe --ARCHIVE_FILE !output_file! --AUTOBID --HAND !input_file! --CC1 GIB-Thorvald.bbsa --CC2 GIB-Thorvald.bbsa --DD 0 --SD 1"
    echo !command!
    
    !command!
    echo BBA Started for input file %%i
    timeout /t 2 /nobreak >nul
)

endlocal

Using start "BBA" /D . will start a new BBA - number of running instance is dependent on our hardware

[ADavidBailey]

I want to run this on over 190 input files! The UI is a big issue.

[ThorvaldAagaard]

I will create a request for BBA, but one idea for you is to combine the 190 input files into one file, and then split it after the bidding.
I think you can use board number to keep track og the file where the deal originates from.

[ThorvaldAagaard]

EdwardPiwowar/BBA#868

[ADavidBailey]

When BBA is run from the Command line, the archive file is not created AND the scores.txt file is not updated.

These seem to be working: -- CC1 --CC2 -- HAND --DD --SD

Here's a link to my .CMD file: https://github.com/ADavidBailey/Practice-Bidding-Scenarios/blob/main/makeBBA.cmd

[ThorvaldAagaard]

Open BBA and make sure Achive is selected - it used to be saved, but I just tested it and it is removed on close, and that is probably the problem.
@EdwardPiwowar

[ADavidBailey]

I observed the BBA window after starting my makeBBA.cmd. Archive is selected for each runs (I have to close BBA after the first.)
@EdwardPiwowar

[ThorvaldAagaard]

I think extension for outputfile is added by BBA, and are you sure directory XXX exists?

[ADavidBailey]

Yes, XXX exists. I was directing it to the Practice Bidding Scenarios\bba folder. When I discovered it wasn't working, I created a new folder in Windows. If I run BBA manually, everything works.

[ADavidBailey]

I just realized that I was using .pba as and extension for the archive file. I changed it to .pbn and it's the same.

[EdwardPiwowar]

Paste a BBA screenshot shortly after the command is invoked.

[ADavidBailey]

This is hard to do. Must be quick. Command window overlays it. ...

[EdwardPiwowar]

C:\BBA\BBA --CC1 C:\BBA\GIB-ADB.bbsa --CC2 C:\BBA\21GF-ADB.bbsa --DD 0 --AUTOBID --HAND P:\pbn\2N_Smolen_Leveled-S.pbn --CURRENT_ARCHIVE P:\pba\ --ARCHIVE_FILE 2N_Smolen_Leveled-S.pbn
or
C:\BBA\BBA --CC1 C:\BBA\GIB-ADB.bbsa --CC2 C:\BBA\21GF-ADB.bbsa --DD 0 --AUTOBID --HAND P:\pbn\2N_Smolen_Leveled-S.pbn --CURRENT_ARCHIVE P:\pba\ --ARCHIVE_FILE 2N_Smolen_Leveled-S

[ADavidBailey]

I just realized that I was using --ARCHIVE instead of --ARCHIVE_FILE. DUH!!

[EdwardPiwowar]

Yes, --ARCHIVE is only from your dream today...:)

[ADavidBailey]

--ARCHIVE_FILE... is working. The archived files have the correct name.

I've tried both of the following: note that I've changed 'pba' to 'bba'

--CURRENT_ARCHIVE P:\bba
--CURRENT_ARCHIVE P:\bba\

The archived files are in P:, my Practice-Bidding-Scenarios & GitHub root directory, instead of P:\bba

[ADavidBailey]

@ThorvaldAagaard
I added 5431 After NT. I see it's on BBA CC's. Which Systems/Robots use it?

[ThorvaldAagaard]

GIB?

[ADavidBailey]

GIB plays Splinters after NT. It's very much like 5431 after NT; but, different. I created a separate scenario.

→adb42 (Private): --- Splinters After Notrump
→adb42 (Private): South opens 1N and responder has a void/singleton ♠ ♥ or ♦ with at
→adb42 (Private): least 4-cards in each of the other suits and 12+ TP.

from https://sites.google.com/view/bbaenglish/description-of-conventions/5431-after-1nt

→adb42 (Private): --- 5431 After Notrump
→adb42 (Private): South opens 1N and responder bids 3M，showing
→adb42 (Private): . a singleton or void in that major，
→adb42 (Private): . exactly 3-cards in the other major，
→adb42 (Private): . at least 4-4 in minors，
→adb42 (Private): . and 8+ HCP (if 1NT 15-17 HCP).

[ThorvaldAagaard]

Yes, you are right. So no robot play 5431
BTW I play it bidding the 3-card major to avoid a lead directing double with one partner :-)

[ADavidBailey]

So no robot play 5431

That's interesting. So, why does @EdwardPiwowar include it in his BBA documentation and CC? :) That's the only place I recall seeing it.

Years ago, I played fragments. No one I play with, today, would touch that.

[ThorvaldAagaard]

It might be standard in polish club. Remember BBA is not started from robots but from human bidding

[ADavidBailey]

This morning I woke up thinkin' 'bout vulnerability...

When I generate deals at a BBO Practice Table, I have no control over vulnerability. So, none of my dealer code considers vulnerability.

But, I do have control when I map the deals into duplicate boards. Boards 1, 14, 11, and 8 all have the same vulnerability with respect to the dealer. So, when I'm creating pbn files , rotated-pbn files, and mixed-lin files I could map the deals to boards in a way that they maintain their dealer/vulnerability relationship.

@ThorvaldAagaard this is all your fault!! When discussing Game_Overcalls, you mentioned 'losers would be better' and the Rule of 1-2-3-4-5 (is that what you called it?) which is based on vulnerability.

It would be an interesting challenge to generate deals based on vulnerability. Is it worth it?

I discovered that I'm not the first to be frightened awake by this nightmare!
https://tedmuller.us/Bridge/Esoterica/BoardVulnerability.htm (this one is really interesting)
https://www.bridgehands.com/V/Vulnerable.htm

[ThorvaldAagaard]

Yes, you should always consider vulnerability :-)

With one of my partners I play weak NT, except 3rd hand if alone vulnerable, then it is 15-17 - some practice deals for that will show you the requirements.

[ADavidBailey]

Yes, you should always consider vulnerability :-)

I agree. One should always consider vulnerability when bidding.

Let me rephrase my question. Should I attempt to create hands that a bidder (considering vulnerability) would always preempt according to the 'rule'?

And, another. You mention using losers. Do you think I should add that?

I played Precision for years. Added Canopy style and played that for a long time. We played 11-14 NT. We opened all 11 HCP hands. Once people figured out how to play against it (years) it became less fun. I've forgotten most of it.

[ThorvaldAagaard]

I think losers are important when preempting,

[ADavidBailey]

I think losers are important when preempting,

Is this correct?
winnersSouth = 13 - losers(south)

[ThorvaldAagaard]

Depend on what you want to use winners for :-)

I have a small tool, that calculates some metrics for a hand
https://handeval.snippen.dk/#/home

[Rick-Wilson]

You can specify the vulnerability for any given hand in the PBN file. You may choose to use the default vulnerabilities, but you don't have to. Some software will honor the vul as specified, and other software will just use the default value anyhow.

The same goes for the dealer. When we rotate the hands, the dealer is no longer the standard, default value. For face-to-face, using the dealing machine and duplicate boards, I will tell the class hand-by-hand who is dealer. I could do the same for the vulnerability, but for these beginning and intermediate classes I don't need to.

[ADavidBailey]

Did you look at how I do the rotation for the mixed.lin?

[ADavidBailey]

I've been testing. The Dealer at a Bidding or Practice table ignores 'dealer' and 'vulnerable'; so, I cannot control or determine the dealer or vulnerability. Thus, I'm giving up on adding this logic to my scenarios. Maybe someone can show me that I'm wrong.

If I were just creating pbn and lin files, I could do it.

[Rick-Wilson]

I wrote a script bba_one.cmd with these contents - it runs BBA for the passed-in file, waits 5 seconds, then kills BBA. This could be put in a master script to run all 190 of your files. This could be a placeholder until the AUTOEXIT is implemented in BBA. This assumes the BBA folder is in your PATH environment variable. If not, specify the full path to BBA.exe.

powershell -executionPolicy bypass -command "& {$process = start-process $args[0] -passthru -argumentlist $args[1..($args.length-1)]; exit $process.id}" bba --HAND %1.pbn --ARCHIVE_FILE %1_Archive --AUTOBID

set pid=%errorlevel%

timeout /t 5 /nobreak > NUL

taskkill /pid %pid% /T /F

[ADavidBailey]

BBA has NEVER seen an 8-card suit it didn't like!! Interestingly, playing on BBO, I bid 2H and West bid 2S all pass. With manual bidding in BBA, after 2H West bids 3H and they get to 4S. BBO's double dummy says EW make 5S and NS make 2H.

N/S is vul and South has 7-losers. Do you count that as 7 losers?

0653ECCA276D2667E27E2E6ABFB1

[EdwardPiwowar]

This is not true:
0653FCCA377D2667E26E2E6AAFA1

BBA sees 7 losers but it's hard to imagine a hand and an auction where 4H will result in a loss.

[ThorvaldAagaard]

4HX seems to be at least 800, so I agree, this is not a 4H :-) But it is always hard to make penalty double for the opponents

[ADavidBailey]

@EdwardPiwowar I should have included a :) on my 'BBA has never seen' statement. DDS shows

Earlier, @ThorvaldAagaard suggested preempting based on losers and vulnerability. For 7-losers vul that's the 2-level.

8 & 9+ card suits are interesting. It's hard to double. And, if your partner doubles, it's hard to leave it.

Do you suppose that at a table squirms and hesitations improves the opponents outcomes.

[ADavidBailey]

@EdwardPiwowar this discussion was started on @ThorvaldAagaard's GitHub. The BBA bidding is getting excellent results vs. par. The following is talking about my Game_Overcalls scenario...

I changed the hcp range to 5-11 (it was 9-12). This looks like a significant improvement.

Out of 499 deals, 471 deals start 1x-4x, 23 dealse start 1x-5x, and 5 deals start 1x-3x

IMPs to Par after is 4.33 before is 3.80

P.S.
I asked Bob to include the Total IMPs to Par and Average to the CSV file. He liked the idea - next release.

P.P.S
The file has only 499 deals. My end of file logic is not yet right.

[ADavidBailey]

Interesting observations from my 1N_GIB scenario. https://github.com/ADavidBailey/Practice-Bidding-Scenarios/blob/main/bba/1N_GIB.pbn.txt

18 deals did not open 1N
    1  1C-1S-X-4S
    1  1C-2D-2S-=1=-4D-X  | 1x-(y)-2z forcing
    1  1C-2H-P-3H
    1  1C-P-1N-2S
    1  1C-P-1S-P-1N-P-2C-=1=-P-2D-P-2S-P-3H-P-3N  | Two way new minor forcing
    1  1D
    1  1D-1N-P-P-2H-P-3D
    1  1D-P-1H-P-2H-P-4H
    1  1D-P-1H-P-3H-P-4H
    2  1D-P-1S-P-1N-P-2D-=1=-P-3H-=2=-P-3N  | Two way new minor forcing | Artificial
    1  1D-P-2D-=1=-P-2H-P-5D  | Inverted minors
    1  1D-P-P-1S-2H-P-3D

Only 4 of the 15 that started 1N-P-2N continued   
11  1N-P-2N
    1  1N-P-2N-4S-P-P-X
    3  1N-P-2N-P-3N 

Only 9 splintered 1N-P-3H/S
    1  1N-P-3H-=1=-P-3N  | Splinter
    2  1N-P-3H-=1=-P-4S  | Splinter
    3  1N-P-3S-=1=-P-3N  | Splinter
    2  1N-P-3S-=1=-P-4H  | Splinter
    1  1N-P-3S-=1=-X-3N  | Splinter

Only 2 bid Gerber
    1  1N-P-4C-=1=-P-4H-P-6N  | Gerber, for n
    1  1N-P-4C-=1=-P-4S-P-6N  | Gerber, for n

Only 3 bid slam after Texas
    1  1N-P-4D-=1=-4H-=2=-X-4N-=3=-X-5C-X-P-5H  | Texas | Michaels cuebid | To the partner's longer
    1  1N-P-4D-=1=-4S-5H-P-P-X  | Texas
   10  1N-P-4D-=1=-P-4H  | Texas
    1  1N-P-4D-=1=-P-4H-P-4N-=2=-P-5H  | Texas | Blackwood 0314, for !h
    1  1N-P-4D-=1=-P-4H-P-4N-=2=-P-5S-P-6H  | Texas | Blackwood 0314, for !h
    1  1N-P-4D-=1=-P-4H-P-5D-=2=-X-5H  | Texas | Exclusion, for !h
    1  1N-P-4D-=1=-X-4H  | Texas
   12  1N-P-4H-=1=-P-4S  | Texas
    1  1N-P-4H-=1=-P-4S-P-4N-=2=-P-5H-P-5S  | Texas | Blackwood 0314, for !s
    1  1N-P-4H-=1=-P-4S-P-4N-=2=-P-5S-P-6S  | Texas | Blackwood 0314, for !s
    1  1N-P-4H-=1=-P-4S-P-5C-=2=-P-5H-X-5S  | Texas | Exclusion, for !s
    1  1N-P-4H-=1=-X-P-=2=-P-4N-=3=-P-5C-P-5S-P-6S  | Texas | Denies a !h stopper | Blackwood 0314, for !s
    1  1N-P-4H-=1=-X-P-=2=-P-4S  | Texas | Denies a !h stopper

[ThorvaldAagaard]

When I looked at https://github.com/ADavidBailey/Practice-Bidding-Scenarios/blob/main/bba/1N_GIB.pbn.txt I can see you have extracted the bidding for all hands

Here it could be interesting to add the actual hands in the short PBN formoat for a deal, so it is possible to read the decument as a standalone document

Like this

    2     W       4S       50        N:2.Q732.KJ92.7653 KJT53.A96.653.82 Q9.JT85.Q8.AKQJ9 A8764.K4.AT74.T4    1C-1S-X-4S 

BTW: The link in the document
/Users/adavidbailey/Practice-Bidding-Scenarios/bba/1N_GIB.pbn
is on your local disk, and should probably just be 1N_GIB.pbn

[ADavidBailey]

Can I calculate the board number so it can be pasted into BBA? Would that be better than the short PBN format?

[ThorvaldAagaard]

Yes, you can calculate the BBA-number, and no BBA is not better, PBN is better as it is readable, so you can just look at a line with strange bidding without need for starting another application.

If I use BBA I normally start by loading the full PBN in BBA and then just enter the board-number.

You can copy a deal in PBN-format into BBA

I have created a script that translate PBN to BBA: https://github.com/lorserker/ben/blob/main/src/pbn2bba.py

[ADavidBailey]

OK. Try it now.

I tried pasting the 'deal' and it does not get the Dealer correct.

[ThorvaldAagaard]

Correct, you will have to change that, and also vulnerability
You could use the BEN format, that include dealer and vulnerability - readable and pasteable. I think you can paste just 3 lines from the PBN but then you need it with the [ ]

[ADavidBailey]

How to I get BEN format?

…

On Sat, Sep 28, 2024 at 6:59 PM Thorvald Aagaard ***@***.***> wrote: Correct, you will have to change that, and also vulnerability You could use the BEN format, that include dealer and vulnerability - readable and pasteable. I think you can paste just 3 lines from the PBN but then you need it with the [ ] — Reply to this email directly, view it on GitHub <#39 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BC4TXI7LHXDJPAXZPJAYJJDZY47GNAVCNFSM6AAAAABOOLDKX2VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTANZYGYZTOMY> . You are receiving this because you were mentioned.Message ID: <ADavidBailey/Practice-Bidding-Scenarios/repo-discussions/39/comments/10786373 @github.com>

-- David Bailey 918-520-9700 mobile

[ThorvaldAagaard]

So, for my scenarios where Dealer is always South, I could rotate the vulnerability. The hands are always N E S W. Is that correct?

Yes

[ThorvaldAagaard]

Changing the subject: https://www.google.com/search?sca_esv=a8d031e7882a108f&sca_upv=1&rlz=1C5CHFA_enUS876US877&cs=0&sxsrf=ADLYWIJk4m2Z1KWW_6fwdN-yM8er6l1FDA:1727570926382&q=ukulele+orchestra+of+great+britain+fly+me+off+the+handel&stick=H4sIAAAAAAAAAONgFuLSz9U3MCwqMyjPUuLRT9c3LDTNSso1zcnSEvQtLc5MdiwqySwuCckPzs9LX8RqUZpdmpOak6qQX5SckVpcUpSokJ-mkF6UmliikFSUWZKYmaeQllOpkAtUkZamUJKRqpCRmJeSmgMAcwsSYWoAAAA&sa=X&ved=2ahUKEwjLkY-b9-aIAxVMEUQIHWAhEfwQri56BAgREBg&biw=1134&bih=635&dpr=3#fpstate=ive&vld=cid:5dddab52,vid:0HuSx3z0E6g,st:0&wptab=si:ACC90ny5P9tRq53jTjXAQzu67N-ILh8ve9V7xX-LZ_pXF6KBQYSO6boqGuqhH2owUsj3X97QI0RD3Z8XwEoSCH0rASy_LEnGCu_523h6d8YtJqRFNl5Hs75A-DVXfRSEMr1S969naRJi1R7Gcsss1W0hvF9dFGSDThbiLmwtkTYrEGQS0qNE6MZuzqLfEwyAVFKCsm12d6O9

Link should be https://www.google.com/search?sca_esv=a8d031e7882a108f&sca_upv=1&rlz=1C5CHFA_enUS876US877&cs=0&sxsrf=ADLYWIJk4m2Z1KWW_6fwdN-yM8er6l1FDA:1727570926382&q=ukulele+orchestra+of+great+britain+fly+me+off+the+handel
the above link is not working

[ADavidBailey]

I cleaned up the format and heading (still PBN deal format; but, without the N:). It's easier to read with a fixed-pitch font. Look at the file. https://github.com/ADavidBailey/Practice-Bidding-Scenarios/blob/main/bba-summary/1N_GIB.txt

-- Sorted Board Number --
board declarer contract score par north east south west | auction... | notes
1 S 4S 420 KT963.T65.52.AQJ 874.Q2.7643.9843 AQ5.K984.AQT8.T6 J2.AJ73.KJ9.K752 | 1N-P-2H-P-2S-P-3N-P-4S
2 W 4S 50 2.Q732.KJ92.7653 KJT53.A96.653.82 Q9.JT85.Q8.AKQJ9 A8764.K4.AT74.T4 | 1C-1S-X-4S
3 E 4S -680 8.7643.J9.876432 AQJT5432..A743.Q K7.AT98.T62.AKJT 96.KQJ52.KQ85.95 | 1N-2H-=1=-P-2S-P-P-3C-3D-4C-P-P-4S | Cappelletti, 5m-4m

[ThorvaldAagaard]

Nice

[ADavidBailey]

I'm still interested in the BEN format. If the dealer is N, then hands are NESW. If the dealer is E, the hands are ESWN... Is that correct?

Do you have a BEN/PBN preference? If so, why?

Does the 'Sorted Bidding Sequences' summary interest you?

I'm thinking of simplifying from this: 476 S 1N -100 to this: 476 1NS -100 What do you think? (single spaces which don't display correctly with this font)

[ThorvaldAagaard]

In BEN format cards are always NESW
BEN is more compact and readable, and was used for training BEN, but I now support PBN also for training, so I more often use PBN

The Sorted Bidding Sequences are interesting, when you check, that a convention has all different variations included.

I just had an example where
1N-2C
2H-2S
Was missing in stayman, so BEN did not know what it did show.

I have also decide that for training I discard a sequence if it is seen more than 50 times (using a program), and the list give an idea about how many different sequences there are

[ADavidBailey]

I cleaned it up some and ran it with a file that has Double Dummy. Check it out:

https://github.com/ADavidBailey/Practice-Bidding-Scenarios/blob/main/bba-summary/1N_GIB.txt

@ThorvaldAagaard you mentioned that because this link is on my 'disk' I probably should just use...
1N_GIB.txt but, it doesn't work. What am I doing wrong?

[ThorvaldAagaard]

I am just referring to the first line in the document
/Users/adavidbailey/Practice-Bidding-Scenarios/bba/1N_GIB.pbn

[ADavidBailey]

All of my file folders (PBS, dlr, pbn, pbn-rotated-for-4-players, lin-rotated-for-4-players, bba, bba-summary) have like-named files. And, I have a folder waiting for @Rick-Wilson's bba-filtered files.

Because of the number of files, removing the dealer-suffix was non-trivial. So, let me know if you see any issues.