Elinescu-Wladow were stupid. You haven't found the smart cheats...

[PhilKing]

I was not going to post this but in view of some of the previous posts have decided to do so. I am not a mathematician and only have rudimentary undergraduate grasp of statistics. I asked a friend, a retired professor of stats who has many publications to explain the references and posts regarding Bayesian analysis. His comment was that anyone who attempts to apply this to real life situations does not have the slightest idea what he is talking about and is using pseudo mathematical knowledge to attempt to obfuscate his audience. (His words not mine). This is a bit sad really.

 

I thought that was what Bayes' was for (applying it to real life - not the obfuscation bit, although that is clearly more common, and is in itself probably a good example of the Bayes' in action).

[Zelandakh]

Hog, in bridge terms, Bayes Theorem (BT) ~= restricted choice. There are many cases where it is practical, even essential, to apply BT to the real world. That said, I would not try to use it in the way used here because the 1% value is so arbitrary as to be useless and potentially misleading. Some of the maths here is also perhaps somewhat questionable due to the initial signal being assumed to count whereas this is perhaps somewhat arbitrary, akin to the difference between rolling 6 6s in a row or any number 6 times in a row; the point being that we tend to think of the former when it happens when we would in fact have found any number noteworthy. It does not change the fact that the incidence of non-coughing to no shortage is remarkable, even before we start delving into the matches for the code itself.

 

As an aside, it should be noted that E-W played a complex method that involved using difference systems in different circumstances. If I had been coaching them to cheat I would certainly have used this to produce different schemes to match the system in use. For example using coughing for one system and tray signals for the other or simply different coughing codes (reversing ♠/♣ and ♥/♦ for example). In practise I suspect this would be almost impossible to break and subsequently prove in the relatively short time frame of a single match.

 

Perhaps I am wrong but I am confident that I could get away with some forms of cheating practically indefinitely if I wanted to. And I would be more than surprised if that was not the case for a significant number of expert pairs, or at least more than anyone would ever suggest. It does not take much to obtain an edge. It is a bit like watching the current Tour de France knowing that certain techniques that we know about - mini transfusions, etc - are effectively undetectable in addition to whatever chemicals have been developed to replace GPO that we do not know about yet. The whole thing leaves me feeling sadly jaded about bridge - there just has to be a better way out there to make things cleaner, even if it means offending the purists who want to keep the personal touch and look their opponent directly in the eye...

[Trinidad]

I was not going to post this but in view of some of the previous posts have decided to do so. I am not a mathematician and only have rudimentary undergraduate grasp of statistics. I asked a friend, a retired professor of stats who has many publications to explain the references and posts regarding Bayesian analysis. His comment was that anyone who attempts to apply this to real life situations does not have the slightest idea what he is talking about and is using pseudo mathematical knowledge to attempt to obfuscate his audience. (His words not mine). This is a bit sad really.

If I would have to feed every mathematician who told me that his particular corner of math cannot be applied to real life, I would be bankrupt very soon.

 

That is why we have other scientists (physicists, chemists, engineers, ...). They can do what mathematicans can't. And, so far, it seems to work quite well.

 

Rik

[helene_t]

That said, I would not try to use it in the way used here because the 1% value is so arvitrary as to be useless and potentially misleading.

But the 1% wasn't Arend's point. The Bayes factor of 10000 was the point. If you prefer you could use the example that if the prior odds were one to a million in favour of non-cheating, you would still have posterior odds of 1 to 100 in favour of non-cheating. Or whatever.

[olegru]

...

Now, however, the translation of what Elinescu wrote about this incident:

....

 

"After North artificial 2NT opening East thought for at least 1 minute before passing.

It was obvious that he was rankled to get into the bidding ...

 

South, who could not have escaped noticing the long huddle, opened a trap by a deceptive 3♣ bid.

 

It could be true or it could be not true.

Point is: Every single board could be explained by "luck", "table presents" and so on.

Stats cannot be.

[cherdano]

So in the same thread one poster implied I don't know German, another implied I have only pseudo-mathematical knowledge. I have been accused of many things, but these are certainly a first for me.

I really don't get it - why post in a forum if after thousands of posts, you don't know the most facts about the identities of your fellow posters?

[Cyberyeti]

I do have what is in essence a statistics degree, but have a suspicion of Bayesian methods. While you can't make them prove anything you want, with (in)appropriate choice of priors you can prod them in a particular direction. This is a general point, I haven't read the detailed statistical stuff on this case.

[rhm]

Depends how much time is taken for explanations, no? Given you are a pair known for poor disclosure and that therefore the opps may ask for extra clarification on every detail it does not seem unlikely that this non-thinking period might last a while.

I am sorry but this is getting either a bit naive or a bit ridiculous.

Do I have to tell you that if you play long team matches in a world championship convention cards are mandatory and published well in advance.

Now a convention card can not cover all agreements a sophisticated partnership might have, so later bids may require explanations.

But conventional opening bid agreements have to be described exhaustively.

I am sure anyone, who plays a long team match against a pair in the world championship finals will prepare himself and inspect convention cards of his opponents well in advance.

 

Do not tell me that Eddie Wold did not know what the 2NT bid meant.

It is obvious that it is the preempt in conjunction with the colors and his hand, which created a serious problem for him and therefore the huddle.

That's why people devise such weapons.

 

Rainer Herrmann

[hrothgar]

I do have what is in essence a statistics degree, but have a suspicion of Bayesian methods. While you can't make them prove anything you want, with (in)appropriate choice of priors you can prod them in a particular direction. This is a general point, I haven't read the detailed statistical stuff on this case.

 

I agree that you can twist a Bayesian model to do most anything that you want through the choice of a prior.

However, you can do the same with frequentist methods and its a lot easier to hide what you're doing.

 

What you view as a weakness, I view as a strength.

If you're using Bayesian methods, justification for your choice of prior needs to be central to any discussion.

 

The Bayesian techniques do a much better job framing the subsequent conversation.

[helene_t]

I [...] have a suspicion of Bayesian methods. While you can't make them prove anything you want, with (in)appropriate choice of priors you can prod them in a particular direction. This is a general point, I haven't read the detailed statistical stuff on this case.

Oh well you can always call it a likelihood ratio if you are concerned that the audience doesn't like Bayes.

 

Arend apparently found that Woolsey's data are 10000 times more likely under a hypothesis of cheating than under a hypothesis of non-cheating. This is what is (potentially) interesting.

[Fluffy]

So in the same thread one poster implied I don't know German, another implied I have only pseudo-mathematical knowledge. I have been accused of many things, but these are certainly a first for me.

I really don't get it - why post in a forum if after thousands of posts, you don't know the most facts about the identities of your fellow posters?

 

Why is a beginner trying to explain this expert issue anyway?, if you want to practice your german find a school or something. Or better go watch baseball on the tv.

 

(note to that poster who impled the things, this is intended as a joke)

[Trinidad]

So in the same thread one poster implied I don't know German, another implied I have only pseudo-mathematical knowledge. I have been accused of many things, but these are certainly a first for me.

I really don't get it - why post in a forum if after thousands of posts, you don't know the most facts about the identities of your fellow posters?

Wer sind Sie? Und wie viel?

 

;)

 

Rik

[nige1]

[sNIP] Now, however, the translation of what Elinescu wrote about this incident: [sNIP] During the play East realized the trap he had fallen into and called the director and complained in a rude manner "South is strong" and North has given the verbal explanation "weak". North contradicted that he never said weak but "pass or correct", which was proven. The director let the score stand, but East raged on, that North lied and this led to a loud exchange between North and East, which continued after the end of the segment [sNIP]

IMO Elinescu must be mistaken that East continued to accuse North of lying, loudly, after the director ruled. Otherwise...

• If East is an experienced player, who knows the rules and deliberately flouted them, then the director would impose a DP.
• Anyway, official witnesses would have presented this video-evidence to afford the tribunal a balanced view.

[rhm]

IMO Elinescu must be mistaken that East continued to accuse North of lying, loudly, after the director ruled. Otherwise...

• If East is an experienced player, who knows the rules and deliberately flouted them, then the director would impose a DP.
• Anyway, official witnesses would have presented this video-evidence to afford the tribunal a balanced view.

As I understand this happened in the first segment, where to my knowledge there was no video taped yet. It was the last board of the set.

I can neither confirm nor disprove the account given by Elinescu.

However, to me the explanation that

 

1) East huddled noticeably to be detectable by South that East had a serious problem over the preempt

2) knowing Wladow/Elinescu style a little bit it is quite plausible that South decided to set up a trap white against red

3) that East assumed South to be limited by his "pass or correct" action and was not happy thereafter and upset

 

makes a lot of sense and looks to me quite logical and most important fits the Bridge events on this board.

 

And let it be said if this is what happened on this board, Eddie Wolds account at Bridgewinner about this board, which I cited in full and was headed Diagnosis: Foul Play

left something to be desired for to put it mildly.

It gave a lopsided account by leaving out the above points, most important that East (Eddie Wold) took a long time for his initial decision to pass.

If you accuse someone of cheating you should not be "economical" with the facts in your account because these facts do not support your accusation.

According to Eddie Wolds own words "The final hand of the set totally convinced me the doctors were cheating."

This board does not convince me any more.

 

Rainer Herrmann

[Cascade]

IMO Elinescu must be mistaken that East continued to accuse North of lying, loudly, after the director ruled. Otherwise...

• If East is an experienced player, who knows the rules and deliberately flouted them, then the director would impose a DP.
• Anyway, official witnesses would have presented this video-evidence to afford the tribunal a balanced view.

 

 

Nigel you know directors rarely impose disciplinary penalties.

[cherdano]

So first Rainer takes Wold's account of this board at face value, and the board as strong evidence that E-W are cheating, even though it was always plausible that the 3♣ bid was made as a gamble or based on table presence or whatever, or that Wold's account may have been incomplete. Now he does a complete turnaround and is suddenly convinced that Elinescu's account is 100% complete and accurate, and that Wold's account was very lopsided.

 

If you look at the facts with some distance, you should never have treated this board as strong evidence for cheating. It was and is some mild evidence, but it just does not compare to coughing-exactly-when-you-have-shortness-14-boards-in-a-row. Elinescu's account does not cast a doubt on the trial, since (to my knowledge) this board was never cited as compelling evidence in the two rulings. It does cast doubt on the judgement of whoever thought this board was strong evidence (such as rhm), and perhaps a little bit of doubt on the methodology of the German commission (which I never found as convincing as either the evidence used in the trial nor the study by Woolsey).

[rhm]

So first Rainer takes Wold's account of this board at face value, and the board as strong evidence that E-W are cheating, even though it was always plausible that the 3♣ bid was made as a gamble or based on table presence or whatever, or that Wold's account may have been incomplete. Now he does a complete turnaround and is suddenly convinced that Elinescu's account is 100% complete and accurate, and that Wold's account was very lopsided.

 

If you look at the facts with some distance, you should never have treated this board as strong evidence for cheating. It was and is some mild evidence, but it just does not compare to coughing-exactly-when-you-have-shortness-14-boards-in-a-row. Elinescu's account does not cast a doubt on the trial, since (to my knowledge) this board was never cited as compelling evidence in the two rulings. It does cast doubt on the judgement of whoever thought this board was strong evidence (such as rhm), and perhaps a little bit of doubt on the methodology of the German commission (which I never found as convincing as either the evidence used in the trial nor the study by Woolsey).

Please tell me why I should not have taken Eddie Wold's account at face value?

 

Do you want to give support to Elinescu's claim, that this accusation is nothing but a conspiracy by american professionals, who can not bear losing a world championship to some amateurs?

 

I said : ...I remember impressed me as "Bridge evidence" at the time when it was brought forward by Eddie Wold on Bridgewinner

 

However Eddie Wold said "The final hand of the set totally convinced me the doctors were cheating."

I never said it is strong evidence for cheating.

 

Elinescu rebuttal was to the German commission, which cited boards from Bali, the commission found suspicious.

This was one of them.

To my knowledge Wladow / Elinescu have so far refused to publish anything with regard to the WBF trial.

 

The only thing I try is to keep an open mind.

And if someone is accused of cheating I want a fair process.

I am not claiming they are innocent.

 

Rainer Herrmann

[helene_t]

If I would have to feed every mathematician who told me that his particular corner of math cannot be applied to real life, I would be bankrupt very soon.

 

That is why we have other scientists (physicists, chemists, engineers, ...). They can do what mathematicans can't. And, so far, it seems to work quite well.

You probably won't have to feed many statisticians who claim that their work can't be applied to real life. OTOH you might know some Bayesians who would say that frequentist statistics cannot be applied to real life, and vice versa :)

[hrothgar]

Please tell me why I should not have taken Eddie Wold's account at face value?

 

Do you want to give support to Elinescu's claim, that this accusation is nothing but a conspiracy by american professionals, who can not bear losing a world championship to some amateurs?

 

Both Wold and Elinescu are principals in the dispute. As such, folks should be cogniscent that anything these folks is going to be filtered and interpreted through their own sets of biases. Hence, Cherdano's recommendation to focus attention on:

 

1. The original source material (the videos and WBF records)

2. Analyses like Woolsey's that were done in an open and collaborative manner

[nige1]

Nigel you know directors rarely impose disciplinary penalties.

This thread is about cheating. If you know the law and break it, deliberately, to gain advantage, that is cheating.

 

Allegedly, a top-level player continues to accuse his opponent of lying, in a loud voice, after the director rules. If the director judges that the player knows that such behaviour is against the rules and that it could gain advantage by disconcerting his opponent and spoiling his concentration, then the director should take drastic action.

 

I suspect that Elinescu's claim is mistaken. If not, you might expect it to form part of the E-W defence.

 

BTW, I took part in Kit Woolsey's meticulous investigation. Notwithstanding this and other strong evidence, I still feel that E-W deserve a fair hearing.

[hrothgar]

Notwithstanding this and other strong evidence, I still feel that E-W deserve a fair hearing.

 

I doubt that anyone would disagree with this statement.

 

With this said and done, there seems to be significant disagreement over what constitutes a fair hearing.

(Many folks, myself included, feel that E-W received a fair hearing)

[cherdano]

Please tell me why I should not have taken Eddie Wold's account at face value?

 

Do you want to give support to Elinescu's claim, that this accusation is nothing but a conspiracy by american professionals, who can not bear losing a world championship to some amateurs?

For any simple director ruling it is quite common that the story told by one pair is not 100% complete, or that they remember little things differently than the other pair. Why should it be different in this situation? Wold had to play two sets against a team that included a pair he was convinced is cheating. I don't know Wold at all, but I know that I myself might get quite emotional in that situation, and I wouldn't recommend blindly trusting my eyewitness account.

[the hog]

"The final hand of the set totally convinced me that they were cheating." So before this board he was clearly not totally convinced. The last board proves nothing and in fact Wold is made to look foolish by players he possibly regarded as amateurs. So is it surprising he gets angry and plays the cheating card?

[gwnn]

"The final hand of the set totally convinced me that they were cheating." So before this board he was clearly not totally convinced. The last board proves nothing and in fact Wold is made to look foolish by players he possibly regarded as amateurs. So is it surprising he gets angry and plays the cheating card?

Wold did not know that every board where they had shortness they coughed in the auction period and every board where they had no shortness they did not cough. You need careful video analysis for that and not his personal impression in the heat of the moment.

[jdeegan]

 

"After North artificial 2NT opening East thought for at least 1 minute before passing.

It was obvious that he was rankled to get into the bidding, which he finally - after long deliberation - refrained from doing vulnerable in the sandwich position.

South, who could not have escaped noticing the long huddle, opened a trap by a deceptive 3♣ bid.

North alerted and gave the written explanation : Pass or correct

 

Now East intervened with 3♠,which was doubled and went for 1100.

During the play East realized the trap he had fallen into and called the director and complained in a rude manner "South is strong" and North has given the verbal explanation "weak".

North contradicted that he never said weak but "pass or correct". which was proven.

The director let the score stand, but East raged on, that North lied and this let to a loud exchange between North and East, which continued after the end of the segment.

South trap was an in depth but riskless maneuver. To insinuate illegal information is malicious."

 

 

This one hand proves nothing in and of itself. One doctor evidently laid a clever trap for Eddie whose game is more analytical and less psychological than most. He says he was reading Eddie, not his partner. This is legal. Even if his partner did write "weak" at some point, so what if that was their understanding? What should matter, imo., is the weight of evidence from many hands and, possibly, some of these from other matches. Who knows? Maybe the doctors were behind and needed a board?

 

Tit for tat redress for any one match is unlikely. Getting rid of the real cheats over the long run is what matters.