Join Bridge Winners
All comments by Richard Willey
You are ignoring the author of this comment. Click to temporarily show the comment.
> My post was intended for this particular tournament, which
> (according to multiple posters above), had a regulation
> barring openings “a king below average”. That's 7 HCP
> or less.

> I do not KNOW the exact purpose of this regulation, but I am
> guessing it was to prevent very light third seat NV openers,
> (which to be playable by the light openers, require methods
> discussed, undiscussed, perhaps developed through partnership
> experience) which cannot be fully and fairly disclosed in the
> short time frame of qualifying and even knockout rounds.

I'd bet dollars to donuts that he intention was to ban strong pass systems and that the impact on 3rd seat openings is collateral damage
Sept. 15, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> While some of you are diddling to fix this, may I suggest
> a fix to prevent the computer hands from “randomly”
> running a full session with cards running entirely N-S
> or E-W, as even if that is 100% random it ruins the game
> for half the players and this is an apparent regular happening
> at all levels of ACBL tourneys.

If cards are dealt randomly, there are going to be “runs”.
This is generally considered a good thing.

Consider the following: Suppose that one were to adjust a dealing program such that the number of HCPs were balanced out between N/S and E/W across the duration of a round. This would enable players to start counting cards. (I'll keep a running total of the number of HCPs that N/S held across the first N boards of the event and then use this to predict how strong their hands are on the remainder.)

Life gets order's of magnitude worse when you start consider that you normally generate 36 or so boards, however, you need to make sure that any set of 27 consecutive boards also need to be balanced. pretty soon, your trapped in a world where E/W always get 20 HCPs, as do N/S.

And, of course, its much more difficult to write a program that will do this in a manner than is fair and secure.

Sorry, but the very suggestion strikes me as badly misguided…

(I'll note in passing that anyone who has access to a computer as you apparently do, has access to a random deal generator)
Sept. 15, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
There might be value in having some kind of system in which the seed for a tournament was developed in a manner that was public, collaborative, and verifiable.

I had originally been speculating about this type of scheme for that point in time in which major tournaments transition from using pasteboard to a fully electronic playing environment. Who knows, there might be value to introducing this earlier.

1. At the start of the tournament select four teams at random

2. Each of these teams (plus the chief tournament director) deals out a bridge hand and creates a private copy that they will keep secret.

3. The chief enters his bridge hand into the dealing program, then each of the four teams do the same. The dealing program hashes each of the five hands together and uses the output to seed the RNG for the event.

4. At the close of the round, all five hands are publicly posted. Anyone who wants can verify that the boards played match the hand that were input.

I'm sure that this system can be improved upon, however, here's what I am going for

1. Recovering the seed requires either that the computer running the event is compromised or that four different teams as well as the chief TD are in collaboration with one another.

2. The process can be validated at the end of the match.

From my perspective, the best way to crack this type of system would be a social engineering attack. As such, I'm not sure that implementing this would be worth the bother until there is a computerized system in place. But your milage may vary…
Sept. 15, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> And, I was publicly impressed with the (hypothetical) software that
> Nic was involved with, so we'll always have that :)

I think that you are significantly over estimating the difficulty of cracking this program.

Most of the crypto types that I discussed this with considered this a pretty simple system to break. In particular, the fact that the program was written quite some time ago and was almost certainly designed for efficiency rather obfuscation probably helped…
Sept. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> The ACBL has really been a leader in tracking down and prosecuting cheaters.

Interesting theory…

Let's look at the record that Fantoni and Nunes racked up over the last decade

First place finishes
2003 Jacoby Open Swiss Teams
2004 Wernher Open Pairs
2004 Mitchell Open BAM Teams
2005 NABC Open Swiss Teams (Roth)
2006 Kaplan Blue Ribbon Pairs
2007 Reisinger BAM Teams
2011 Spingold KO Teams
2012 Spingold KO Teams
2012 Reisinger BAM Teams
2013 Reisinger BAM Teams

Second place finishes
2004 Jacoby Open Swiss Teams
2006 Spingold KO Teams
2010 Spingold KO Teams
2011 Reisinger BAM Teams
2014 Spingold KO Teams
2015 Spingold KO Teams

Add in Balicki-Zmudzinski and Fisher and Schwartz and pretty much every major event that the ACBL ran for decades has a known cheater or placing.

Next, please recall… We've only caught the stupid cheaters. The smart ones are still out there.

I've been on the record saying that folks are being cheated blind for well over a decade. Sadly, the ACBL preferred to stick its head in the sand rather than taking reasonable proactive measures.

The ACBL reaction when I informed them that their hand generator was insecure is a classic example of the organization's failure mode.
Sept. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I am all in favor of such a system.

I'd just like to note that designing the database on the backend and making sure that it has good search capabilities is likely to be every bit as important as data capture. (I don't like the thought of having to teach SQL and grep to WBF officials)

You probably also want to make sure that this system is adopted during team trials and the like.
Sept. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I am not aware of any evidence that anyone has used such a program. (If you wanted to cheat with such a program, its really not the sort of thing that one would want to publicize)

With this said and done, I would not be at all surprised to find out that someone else has developed this same program and made use of it.

Consider the following

1. The hand generation program in question has been in use for a significant length of time

2. The ACBL published a description of the random number generator being used in the ACBL Bulletin. This information was more than sufficient to point to serious flaws in the implementation

3. It took an extremely short time for folks to develop the crack

4. Top level tournament bridge is sufficiently lucrative that there is clear motive

5. We know for a fact that similar types of attacks have been used against online poker services

6. We know that top level bridge players cheat

From my perspective, this type of attack only started to become really practical once we had cheap parallel computing as a service Say 2007. Someone might have been able to get a beowulf cluster before then or some such. Online Vugraph probably made this easier.

So yeah. I'd guess that some one or a few someones have been cheating for a while…
Sept. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I suspect that it would increase the search cost
Sept. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Isn't there some kind of verbiage in the Laws about “external aides to computation” or some such?
Sept. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I first noticed this on the crypto mailing list.

I held off from publishing this on BW, however,

1. Nic posted a link in a previous thread
http://bridgewinners.com/article/view/acbl-and-usbf-hand-records-are-plausibly-insecure/

2. Someone else suggested that this should egt a thread of its own
Sept. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I think that Godel might differ with your faith in mathematics…
Sept. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
To me, the most critical issue in this whole discussion is the frequency with which B+L have the opportunity to choose to open holding 0-7 HCPs in 3rd seat. B+L play an extremely aggressive opening style in first seat. Presumably, many of their opponents are behaving much-the-same in second seat. The net result of this is that the conditional probability that B+L are dealt 0-7 HCP following two passes is extremely small. As I understand matters, we have a corpus of roughly 1500 hands stretching back four years or so. Of these, a grand total of five hold (0-7 HCPs) following the auction Pass – (Pass).

Why is this important? Lall states that he and Bathurst don’t have an explicit agreement to open with extremely weak hands. If we are willing to accept this claim at face value then our concern is whether or not these five hands are sufficient to establish some kind of implicit agreement. Even if L+B chose to open all five of these hands (and they did not), I would hesitate to claim that this is sufficient to establish an implicit agreement. Five hands spaced out over four years just doesn’t feel like enough. Maybe if we had access to all of the various practice boards that B+L played I’d reach a different conclusion, however, for now I just don’t think that we have enough data.

This takes me to my second point: Before doing anything else, I’d really want to understand to what extent B+L’s behavior differs from that of other World Class pairs playing strong club. Personally, I’d have a real issue if the WBF were to prosecute one specific pair for a bidding style that is in wide spread use in top levels of the game. Personally, I think that it would be worthwhile to conduct such a broader analysis, not as part of some witch hunt, but rather to see whether expert practice is consistent with the rules as they are currently written (If a conflict is found then I think that the rules should probably be changed)

Let’s turn now to the Spanish team. Speaking as an American, I have absolutely no problem with the Spanish team’s decision to publically introduce these charges during a session, nor do I consider it problematic that the Spanish team wanted B+L benched. As I have mentioned in the past, I think that the American teams engage in a lot of gamesmanship during these events and I absolutely believe that turnabout is fair play. With this said and done, I think that the way in which the Spanish team brought these charges forward was inept at best. There clearly does not seem to be enough data to draw any kind of a firm decision. If you are going to publically pull this kind of stunt then you damn well better have your shit together. The Spanish team clearly did not.
Sept. 13, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> “A King or more below average strength” is different for
> third seat than for other positions.

Defining average strength based on conditional probabilities seems unworkable.

More significant, the same regulation that defines a HUM defines an “Average Hand” as one “containing 10 high card points (Milton Work) with no distributional values”
Sept. 12, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Must be even more sad to see your country lose to that team over and over and over again…
Sept. 12, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I doubt that the Spanish team had any issue coping with B+L's system. I think that they were engaged in gamesmanship.
Sept. 12, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I strongly suspect that they mean 500 hands in total.
Sept. 12, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
As I recall, they played Strong Club when the opponent's were vulnerable
Sept. 12, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> So which is the explanation…..incompetence or corruption?
> I can't see a third possibility. But I'm willing to be convinced.

A system designed by Kaplan to be run by Kaplan doesn't work very well without Kaplan.

As I understand matters Kaplan did an awful lot to make sure that he has a much flexibility as possible in interpreting the Laws. This is all fine and dandy if you have a singular talent like Kaplan at the helm. I don't think that this system works nearly as well in the hands of Endicott or Wignall…
Sept. 12, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> I agree with your list of points and I am sure Ignacio will delighted
> with the checklist you have given him, but I can't find anything in
> the Conditions of Contest that says all complaints should be done
> with public statements, nor that BridgeWinners is preferred
> vehicle for publishing such statements.

It's under Section 8, subsection 4A of the Conditions of Contest which reads:

“If a team ‘screws the pooch’ by engaging in a poorly researched theatrical display on international vugraph they really might want to try to get ahead of the scandal before they are tried in absentia in the court of public opinion.”
Sept. 12, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
The America teams has engaged in all sorts of gamesmanship when it comes to international competition and there are definite examples when the US team has “worked the ref” during discussions surrounding system regulations.

Form my perspective, the worst such example was an incident where Brink-Drijver from the Dutch Team were playing an overcall structure called “Holo Bolo” over convention club openings. One of the American pairs was playing a short club that showed 2+ clubs. Not only did they managed to get their club opening redefined as “natural”, they also tried to nail the Dutch pair for playing an excessive number of Brown Sticker Conventions and force them to play standard.

Simply put, if your NBO is going to play these sorts of games then you lose your right to complain when other folks pull the same sort of shit…
Sept. 12, 2016
.

Bottom Home Top