Join Bridge Winners
All comments by Nicolas Hammond
You are ignoring the author of this comment. Click to temporarily show the comment.
Bob Hamman asked me to play with him at an NABC. He had left the main event early and was playing in the regional KOs. We were at breakfast and the event started at 1pm. I spent the rest of the morning studying the various system notes I had created over the years.

We met one hour before game time to create a card. I play with a lot of last minute pick up partners at our Sectionals/Regionals and I usually play their card and was expected Bob to say something similar, “show me your card and we will play that”.

But, no.

Bob start to describe what we were going to play, no real discussion allowed and none ventured. Who was I to disagree on the best treatment for a bid? I had spent some of my time in the morning re-studying 4 card majors, but we were going to be playing 5 card. After a few minutes, we had most of the right side of the card filled out and then we got to the 2D section.

“Do you play Flannery?”, he asked.

“Yes”. I had spent some considerable time that morning going over all my Flannery notes. I knew when in the fourth round of bidding if 4H is to play or 4H is a transfer to 4S, so yes, I knew Flannery. I tried to suppress a smile when I said it.

“Good. Here's how we are going to play it.” My smile went away.

Bob proceeded to give a detailed explanation of all the responses he wanted starting with 2NT and higher and all the follow-up bids. I had written Flannery in the small line in the 2D section in the ACBL convention card and started to write all of Bob's responses for 2NT and higher in the “Responses/Rebid” section. Very quickly I was out of room and starting to write smaller and smaller. I started writing into the 2H and 2S Response/Rebid section hoping that when we got to those bids that the responses and follow-up bids would not be as complicated. I quickly filled those up and then started writing in the 2C section. When that filled up, I started writing around the sides of the convention card until I had gone 3/4 of the way round in the tiniest font possible and then back again. Bob covered every different possible bid and rebid and scenario. I had used every possible abbreviation and shorthand I knew, including some I made up as he was giving the Flannery that we were going to play. My convention card was covered in tiny hieroglyphics and looked like the Rosetta Stone. Archaeologists years from now would not be able to understand my notes, and I suspect I may not either.

After I had written down all the 2D bids, we barely had any time left for the left side of the card before the event started.

And no, Flannery did not come up… but if it did, I was prepared!
Feb. 10, 2015
Nicolas Hammond edited this comment Feb. 10, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Let me give an example.

Say Jeff Meckstroth is on a KO team, let's say he has 70,000.0 MPs, he plays with 3 other players, they each have 10,000.0 MPs. So the total is 100,000 MPs.

What is the amount of MPs used for this team for calculating the KO MPs?

Answer: 45,000

This is documented in the MPBook (page 5, max is 45,000).

Now he plays with the same team in a compact KO.

What is the amount of MPs used?

The answer is not 45,000. I have not seen the answer documented anywhere. This is the type of information that is lacking from the MPBook. Given the MPBook, anyone should be able to re-create the MPs for an event. Currently this is not possible for all events.
Feb. 7, 2015
Nicolas Hammond edited this comment Feb. 10, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Peg: Not sure why you did not get your results, this would be an ACBL Live issue. You would need to talk to ACBL support.

I introduced Fast Results at the 2012 Memphis Platinum Pairs. I started work on it around November 2011, and first tested it at a Sectional in Atlanta in February 2012. Code took about 2-3 months to develop/test/release. ACBL has had the code since April 2012. It was part of the ACBLscore+ contract; I put it in Phase 1 of the contract so that there would be an easy quick win so ACBL could start to show benefits from the ACBLscore+ work.

Jay Whipple introduced his own version about a year later - known as “Fast Results”. Very popular with Regionals and Sectionals. I think he got frustrated with ACBL not releasing anything.

ACBL introduced their own version in Providence. Took them 2 1/2 years. I think they used their own version. They did not want to use the software previously created :-)

I'll see if I can dig out a backup so that you can compare the results presentation from what is now 3 years ago to what ACBL are offering.
Feb. 7, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.

It is fairly simple. Similar to what is done now, but automated.

Have all seeds go to the seeding table. Seeds entered into computer. These seeds are manually seeded, possibly with some computer help, by the seeding committee. Very similar to the current process. Seeds would get a monotonically increasing number, starting at say 501.

All others pairs get a monotonically increasing number, similar to what is done for KOs. Players would enter their MP total, just like as in a KO. Yup, going to use MP for the moment as the simplest seeding method for the anti-seeds.

At game time, TD decides the number of sections and number of pairs in each section.

Computer assigns seeded players across the sections. Computer assigns non-seeded players to fill in remaining seeds.

Just like a KO is started at ACBL, would display your pair number and table assignment.

Only I'd use a projector to display it all. Possibly two projectors - one for seeds, one for non-seeds.

Probably about a week of work to do seeds, a week of work to do the non-seeds. I"ve already got the code for projecting data. Probably about 2 weeks worth of work (this would include testing etc.)

If I thought that ACBL would realistically implement a solution, I would implement it. (They would not own the copyright :-)

If I have time, I'll see if I can get something done for New Orleans.
Feb. 5, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Kevin: You would have to look at each event and decide if they used 3 or 5 as the first seed. Take a look who was sitting in seat 5 for some sections and you can probably determine fairly quickly if this was the 1 seed or the 4 seed.

As best I recall the decision to use 3 or 5 was the DIC. May be different for each pairs event within the same NABC.
Feb. 3, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
The Vanderbilt/Spingold have defined seeding procedures.

Each time seeding is done for the pairs events it is the subjective decision of the people on the seeding committee that day. I'm not sure that any records of kept of the seeding information from the National Pairs events.
Feb. 3, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
The ACBL use different rules for NABC+ KO events, these are seeded across the entire field suing a published seeding method.

The ACBL NABC+ pairs events are seeded, but the seeding is manual. Only the teams that have requested a seed are seeded.
Feb. 3, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.

The reason for scoring within a group of sections is historical. Only so much data input per terminal per person. With the introduction of Electronic Scoring Devices (Bridgemates etc.), there is a typical limit of 6 sections per Bridgemate.

ACBLscore is non-networked and single-threaded. Each scoring machine can only handle one ESD server. So each scoring machine scores up to 6 sections. Typically a top of 77= ((6*13) - 1).

ACBL could score across all sections. With the current ACBLscore technology, players would have to wait until all scores were in from all sections. Then a TD would sneaker-net the game files from all scoring machines to a single scoring machine, import the data, merge it, score across all sections. Then sneaker-net the data back to the original scoring devices because there is only one printer hooked up per scoring device. The printers take 22-23 seconds to print each page. Each section typically requires 4 pages (2 for NS, 2 for EW). Typically about 90 seconds per section. Then they are manually hung up.

TDs want to find scoring corrections as quickly as possible (you won't believe how many there are).

Using the current technology to score across all sections would add a significant lag time to getting results out to players.
Feb. 3, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Get them to provide their ACBL number(s) as part of payment/registration process, e.g. the same information that is currently collected in round 1. Use caddies/TDs/volunteers/players to enter the player number(s) to a computer system. For all non-seeded players, rank via masterpoints (or whatever other system you want), then you assign all these non-seeded players to the rest of the field, e.g. top pair to table A2 (or whatever the next ‘seed’ table is), next pair to B2 and so on.

Means a change to the current process of signing up for NABC+ events, but it is a fairly simple system to implement. I currently do this for KO events for Bridgescore+. System works very well. At times I have had everyone assigned to a table within 10 seconds after game time. Not much extra work to convert to a pairs event. Can easily handle large numbers of players - it was used to start Gatlinburg KOs last year, more players in the first round of a Gatlinburg KO event than there are in an NABC+ pairs event.

Fairly simple stuff.

If I thought ACBL would use it, I'd probably write the code. Probably about a week of work for a prototype, then need some testing and real-world use. Probably about the same amount of work to get the seeded players done. Got lots of similar code already written, including all the projector/TV display output so is pretty easy to do.
Feb. 2, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
For ACBLscore+, I spent a lot of time at NABC+ pair events looking at the seeding and also the seeding/assignments for day 2. I probably know as much about the process as any player (not a TD).

The current seeding process is totally manual, not automated. This is prescribed by ACBL BOD. Some of it could be automated, and automated fairly easily while still having humans make the final decisions on seeding. I had a plan/design for it, but it wasn't part of ACBLscore+ Version 1.0 so not implemented.

Having to wait 40+ minutes for the boards to come out at a NABC+ Pairs Event for the first round is a little embarrassing. Usually this isn't the seeding committee's fault, though some assume it is.

The current manual process is actually quite efficient, given it is a paper system.

Obviously a system with pre-registration required for seeded pairs 15 minutes prior to game time would be more efficient. Got to allow for human override of computer assigned seeding (e.g. foreign players).

I have seen yelling matches between players and members of the seeding committee (who typically are playing in the event as well). This is not productive or healthy for either the player or the member of the seeding committee or other players in the event. “How dare you not make me a #1 seed!!”. Just a comment for some of the readers of this site. Instead of yelling and causing a scene, show them up and win the event.

The current ACBL process for determining qualifiers and getting ready for Day 2 screams for automation. Improving the process was not strictly part of the ACBLscore+ project, but the code was designed to significantly make the TDs job much easier. It can typically take 5-6 directors about 1-2 hours after the evening session to get ready for the next day. Sometimes even much longer. I think the longest I saw was 4 hours. Players never see this. Thank the TDs the next time you see them, few have any idea the process they are forced to go through.

Kevin: Typically only the odd numbers are seeded. The seeding committee will get, say 150 seeded entries. The DIC will tell them there are 14 sections. So there are 28 seed 1 (either sitting at table 3 or, more recently, 5), 28 seed 2, 28 seed 3, 28 seed 4, 28 seed 5, and 10 put into seed 6. All other pairs will be randomly assigned so anything below a 6 seed isn't really seeded. There is no way of telling the number of seeds by looking at the event; the information is not recorded anywhere so you cannot determine after the event. If the order is 3, 9, 13, 5, 1, 11 then any even numbered table is not a seed, same with table 7 and some of table 11.

For a long time, ACBL ran a web movement for the last section to be seated, so all the others were 13 table sections. Not sure if that is still the case. The web movement is typically not fully seeded. Small point, probably won't affect the stats too much.

In at least one of the recent day 1, there was a problem with seeding, and most sections did not have an E/W Table 1, causing lots of changes to the movement in rounds 2 and 3. One event would not mess up your stats.

If even you or BridgeWinners get some software to automatically seed with manual override, I've got tools to automatically create the ACBLscore game file so we can get the game started much quicker. ACBL has the same tools as well. Just a question of time/priority. The NABC+ events are the ACBL's showcase to foreign players, but make up a small fraction of overall pair events played in ACBLland.
Feb. 2, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
At some of the recent NABC+ events, the top seed was at table 5, not table 3. With corresponding changes to the other seeds.

You would probably need to look at the event, look at some of the players at tables 3 and 5 and can probably easily determine what the seeding policy was for that event.
Feb. 2, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Damn those Superbowl ads.

I just poured coke into my laptop.

Nothing changed.
Feb. 1, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
You are ignoring the author of this comment. Click to temporarily show the comment.

ACBLscore+ has built-in support for web services.

Take the ACBLscore+ that ACBL has, put it on-line (cost was $12/month), TDs/Clubs can upload game files (been supported for 18+ months), then access the game with web services. You can do everything that you described above. I used JSON for web service output, in some places it is XML, but normal web service stuff.

For example, see ./app/views/events/index.json.rabl for an example of code to create a JSON response. Code is trivial. Try http://(url)/clubs/1/events.json?start=1022766800&end=2526392000 to see the output. ACBL has the code, so should be trivial for them to put this up. This will return all the events associated with a club from a start and end time. It is used for the calendar feature for clubs. See last few seconds of for the club calendar feature. Data is fed for each month using the JSON described. I guess I need to post another Youtube video showing how fast it is. The 1 in the URL is the club_id, change for whatever club_id you have defined.This will get you a formatted JSON reply. It is trivial to add more web services. Whenever you need a Web Service, just add

def index
respond_to do |format|

to the controller, cut/paste the aforementioned index.json.rabl to the model. Very easy.

Glad that ACBL are going to move to a model where the data is more accessible.

Previously they have said the opposite.

There are lots of web tools associated with ACBLscore+/Bridgescore+. At some point, I'll put them online so that others can use them. Glad you have ACBL moved in the right direction.

BTW, the first implementation of “fast results” was part of ACBLscore+. Showcased at the Memphis NABC in March 2012. It used a similar approach. Has taken ACBL 2+ years to get the code in place to display results, still not as good as was displayed 2+ years ago but a step in the right direction. I'll see if I can dig out those results so that everyone can see what was showcased so we can compare with what we have now. March 2012 was one month before the ACBLscore+ project started… ACBL have had the code since April 2012.

Please make sure that your use of tense is correct as it is starting to sound like you have become an ACBL management mouthpiece. “similar to what Jay Whipple did” should be “similar to what Jay Whipple is doing”. It sounds like you have killed his project, which I know that ACBL is trying to do, but is not the case. Just stay away from their kool-aid.

ACBL have had this (web services technology) available as part of ACBLscore+ for almost 3 years. Take a look at the ACBLscore+ project Wiki design documentation and GitHub design documents/notes for more details. Much cheaper to pay $12/month, put this stuff on line, let everyone access the data through the web services.
Jan. 29, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Even though this is an updated MPBOOK, it is still not what ACBLscore implements. TDs base decisions on events to maximize MPs for players, but the MPBOOK is still wrong.

We really need ACBL to produce an MPBOOK that both describes the actual MP implementation, but also includes all of the MP eligibility and assignment rules. And we need full explanation of how the MPs are really calculated.

We also need a spreadsheet/web site where anyone can double check the MP rules.

The MP rules are v. complicated (I probably no more than anyone else), but I think we have a right to know what they are and how they are implemented.

The MPs was one of the biggest holdups for ACBLscore+. Even at the end of the contract, ACBL still never delivered the specs for masterpoints.
Jan. 29, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Thank you everyone for your help.

The last few got ‘approved’ by Wiki editors today.

Here is the list of 171 players that were added (and approved) over the last 3-4 months:

Adam Wildavsky
Agnes Gordon
Albert Weiss
Alicia Kempner
Allan Siebert
Allan Stauber
Arnie Fisher
Arthur G. Robinson
Arthur S. Goldsmith
B. Jay Becker
Bart Bramley
Ben Fain
Bernie Chazen
Beth Palmer
Betty Ann Kennedy
Billy Seamon
Björn Fallenius
Bobby Nail
Brad Moss
Brian Glubok
Carol Sanders
Charles Coon (bridge)
Charles J. Solomon
Cheri Bjerkan
Chip Martel
Chris Willenken
Christal Henner
Christal Henner-Welland
Chuck Burger
Curtis Cheek
Dan Morse (bridge)
Dan Rotman
Daniela von Arnim
Debbie Rosenberg
Disa Eythorsdottir
Donna Compton
Doris Fuller
Doug Doub
Ed Manfield
Eddie Wold
Edith Freilich
Edward Hymes
Edward O. Taylor
Emma Jean Hawes
Eric Greco
Eric Kokish
Eric R. Murray
Fred Hamilton (bridge)
Fred Stewart (bridge)
Gail Greenberg
Garey Hayden
Gary Cohler
Gavin Wolpert
Gaylor Kasle
Gene Freed
Geoff Hampson
George Jacobs (bridge)
George Rapée
Gerald Caravelli
Grant Baze
Harold Harkavy
Harold Lilie
Harry Fishbein
Helen Utegaard
Hemant Lall
Howard Weinstein
Hugh Ross (bridge)
Ivar Stakgold
Jacqui Mitchell
Jan Martel (bridge)
Janice Seamon-Molson
Jeff Glick
Jenny Wolpert
Jill Levin
Jill Meyers
Jim Krekorian
Jim Mahaffey
Jim Robison
Jo Morse
JoAnna Stansby
John Diamond (bridge)
John Gerber (bridge)
John Mohan
John Schermer
John Sutherlin
John Swanson (bridge)
Josephine Culbertson
Juanita Chambers
Judi Radin
Karen McCallum
Kay Rhodes
Kerri Sanborn
Kitty Cooper
Larry Mori
Larry T. Cohen
Lee Hazen
Lew Mathe
Lisa Berkowitz
Lorenza Lauria
Lou Bluhm
Lynn Baker
Malcolm Brachman
Marc Jacobus
Margaret Wagar
Marilyn Johnson
Marion Michielsen
Mark Lair
Mark Molson
Mary Jane Farell
Matt Granovetter
Meike Wortel
Merwyn Maier
Meyer Schleifer
Michael Seamon
Mike Becker
Mike Kamil
Mike Passell
Mike Smolen
Mildred Breed
Morrie Elis
Nancy Gruver
Nick Nickell
Olive Peterson
Paul Hodge
Paul Swanson
Peggy Solomon
Peggy Sutherlin
Peter Boyd
Peter Leventritt
Peter Pender
Peter Weichsel
Petra Hamman
Phil Feldesman
Pierre Zimmermann (bridge)
Renee Mancuso
Rhoda Walsh
Rich DeMartino
Richard Coren
Richard H. Katz
Richard L. Frey
Rita Shugart
Robert F. Jordan
Robert Lipsitz
Roger Bates
Ron Andersen
Ron Von der Porten
Rose Meltzer
Ross Grabel
Rozanne Pollack
Russ Arnold
Russ Ekeblad
Ruth Sherman
Sabine Auken
Samuel M. Stayman
Seymon Deutsch
Shawn Quinn
Sherman Stearns
Sidney H. Lazard
Sidney Silodor
Sol Mogal
Steve Garner
Steve Landen
Steve Robinson (bridge)
Sue Picus
Sylvie Willard
Theodore Lightner
Thomas K. Sanders
Tobias Stone
Waldemar von Zedtwitz
William Grieve (bridge)
Zeke Jabbour
Jan. 22, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I'm not going to be at the Wilmington Regional so don't know.

If there is someone going, that is interested in running Bridgescore+, has a projector, couple of laptops with Internet connection, printer, let me know.

You'll need permission from the TC and DIC.

KOs and Swiss are the best ones to run.
Jan. 21, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Ask your DD when they were told about outside counsel's review of the contract. All will become clear then.
Jan. 21, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Ed: I spent Sunday at an ABA tournament in Atlanta. I can recommend. They know how to have fun…

I ran Bridgescore+ for them at a two session Swiss event, projectors displaying assignments/results etc. etc. They loved it.
Jan. 19, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Kevin: re: “stories/rumors I heard”.

If I know first hand, I can state it; anything else I hear second or third hand is a story or rumor.

Perhaps poor choice of works, but let me give you an example.

After ACBL dropped ACBLscore+, I heard various ‘stories’ on the reasons why (remember: I know why they dropped it!). One “story” or “rumor” that the software was too slow. Prior to each board meeting for the 2 years of the ACBLscore+, I'd set up a site on the Internet for the board to look at the software. I always set it up in “developer mode” because it's much easier to change anything on the fly. My immediate reaction, was, “duh, they just ran a developer version”. Some time later I looked at the logs and saw that there was nowhere near enough traffic to justify a proper review, which means that ACBL had installed the software themselves somewhere and ran it (I assume). No-one at ACBL, at least AFAIK, knew how to configure a production version, so I still don't know what they looked at. Then I heard that it didn't run on XP. Well… we'd delivered a version on XP. Each time I heard a ‘story’ or ‘rumor’, I'd put a Youtube video up with evidence to the contrary. All the stories and rumors I've heard are just that.
Jan. 19, 2015

Bottom Home Top