Join Bridge Winners
All comments by Richard Willey
You are ignoring the author of this comment. Click to temporarily show the comment.
Like it or not, going forward circumstantial evidence is going to need to be sufficient for establishing the burden of proof.

Assume for the moment that some pair is cheating, however, rather than using a primitive side channel like board position they chose to use some kind of electronic signaling device.

Unless you are lucky enough to catch the players in possession of the device via bug sweeping or some such, you are NEVER going to be able to have direct evidence. You will never be able to observe the signals nor establish a link between the a precise signal and a card being played.

The only thing that you will have available to you is a set of pairs who seems to be abnormally lucky over long periods of time.

You might not like this. (I certainly don't). However, the the bridge community is unwilling to start making some changes, we're going to see a repeat of the last 15 or so years, with all of the major events irrevocably tainted…
March 18, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I find myself more convinced by Greg Herman's dissent
March 18, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> Oh I didn't know they existed or were available, hopefully also sunday :)

Given how little information you have about any of this and the paucity of the effort that you devoted to basic research, whatever made you think that you were qualified to provide expert testimony for this investigation.

There are plenty of people on this list who have advanced degrees in statistics, machine learning, and the like. Yet somehow, we didn't see the need to insert our own opinions into the proceedings.

It takes real chutzpah to provide expert testimony after such a half assed effort…
March 17, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
As I understand matters, the ACBL has a separate corpus of video recordings that have not been made available to anyone during the data analysis.

If some of these videos include hands that F+S played, they would make an excellent test set.

It would be useful to if the ACBL has videos of F+S and, if so, whether these boards are consistent with the hypothesis.
March 17, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I agree that the world is awash with shades of grey, however, when you're talking about deliberate signaling that simply isn't true.

Either the pair is cheating or they are not. I don't see any way to to split the difference here…

If we're looking at things from a statistical perspective, we probably will never be completely certain of what's going on.
March 17, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Strongly agree with Kieren. It would have been better to focus the discussion on the simplest part of the case rather than leaving room for doubt by dragging in extraneous information.
March 17, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
>There is nothing “courageous” about writing anonymous letters!

I don't care whether the individuals posting this are courageous.
I care whether they are providing useful information.

It is for this reason that most large companies provide formal mechanisms for providing anonymous input to the senior management.
March 17, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> Anything else hasn't worked so far, and must be relegated
> to the category of either wishful thinking or sci-fi.

“There Are None So Blind As Those Who Will Not See”

Two quick observations about the current cheating scandal

1. There is a strong incentive to cheat in top level bridge play
2. The advent of video recording means that “naive” cheating methods can be detected

One possibility is that players will stop cheating.

The other is that they will make more extensive use of electronics. (FWIW, I firmly believe that they only reason that people haven't been found to be using electronic transmitters and receivers is that we don't look for them).

All your arguments about observation go completely out the window if the side channel can't be detected via video or audio recording.

March 17, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I'd carefully ask E/W if they want to require that N/S be in game.
If they insist, I'd adjust the score to 3N == and ask them if they need anything else…
March 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> Amir, look at the video sequence with Rezo's “deposition”.
> I understand that he wanted to bring forward boards from
> other competitions, but SEC refused to go into this.

One possible interpretations is that the SEC is worried about jurisdiction. Another is that they don't want to waste time with Rezo and are seizing on a pretense.
March 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> those who insist that the matter has already been proved should
> remember that the judgement of kit, ish and the others came in a
> non-adversarial situation where no one from the side of the accused
> challenged the arguments and the evidence. even those who find
> the evidence absolutely compelling should realize that its validity
> wasn't tested back there by a competent and organized defense.

Any of the accused had ample opportunity to engage opportunity to present evidence in their defense, to dispute the claims being made against them, or make a public statement. They chose not to (probably under the advice of their lawyers or, alternatively, in the belief that the opinions of the individuals on public forum like this one are largely irrelevant)

That's neither here nor there. However, it is ridiculous to discount these finds because it was non-adversarial in nature.

March 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Thanks for the translations
March 13, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> All I get is certain people being offencive to me and I fail to understand why.

Here's one possible explanation:

1. Prior to March 8th of this year, you had never made a posting on Bridge Winners. Your account says that you joined back in 2013, however, for the vast majority of this time you've been lurking. No one knows who you are or has anything to base an opinion on.

2. Our introduction to you and your content was poorly drafted “Statistical Opinion” that was made available for download on the Neapolitan Club web site. http://neapolitanclub.altervista.org/eng/fisher-schwartz-files-statistical-opinion.html

The main thing that the opinion demonstrates is that you did not bother to familiarize yourself with the discussion threads surrounding the Fisher - Schwartz case. Most of the topics that you bring up have already been discussed and dealt with in far more detail.

3. In a similar vein, the videos that you posted above don't show what you claim that they show.

You seem to have some belief that folks on this site have some obligation to invest a lot of time and effort correcting your claims. Simply put, you've done nothing to capture people's interest or convince them that it worth spending time engaging with you.

If / when you are able to make an interesting or even provocative contribution, you might find that people will engage with you more seriously.
March 13, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
My big issue with Liran's document is that it represents the opinion of an isolated individual.

I don't completely agree with all of Kit's methods, however, he got an awful lot of stuff right. One of the most important was using multiple independent individuals for both measurement and evaluation, the result of which is that we can have much more confidence in the reliability of the results.

F+S have produced one individual who happens to disagree.

If you search long enough, you're going to find fair coins that appear to be loaded. In a similar vein, if you shop around for experts, you can find one that will support most anything.
March 13, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Sorry, I should have mentioned that this was MP
March 13, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Sorry, it doesn't work that way. Its late, so rather than trying to spell out the difference I'm going to point you at https://www.stat.berkeley.edu/~aldous/157/Papers/shmueli.pdf

The key quote is the following:

In explanatory modeling the focus is on minimizing bias to obtain the most accurate representation of the underlying theory. In contrast, predictive modeling seeks to minimize the combination of bias and estimation variance, occasionally sacrificing theoretical accuracy for improved empirical precision.
March 10, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
FWIW, I think that people are being incorrect when they are complaining about overfitting.

The expression “overfitting” is typically used in one of two cases:

1. People are including too many explanatory variables in a parametric regression.
2. People are choosing an overly sensitive spanning or smoothing parameter in a non parametric regression.

In either cases, the result is a model that will fit both the signal in the data set as well as the noise. Training sets and validations do get used to control for over fitting. You develop a model using the training set. You test the accuracy of the model using the validation set.

In this example, we're dealing with a question of hypothesis testing rather than developing a predictive model. Here once again, you want to use a different data set to develop your model than the one that you use to validate the model. However, people don't talk about “overfitting”. Rather, the issue is with the accuracy of traditional statistical tests.

FWIW, the classic quote in the area is the following (Fermi attributed it to von Neuman)

“With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.”



March 10, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
FWIW, I've been thinking about some of the statistical analysis that has been going on. I think that it might be useful to simplify / change the base hypothesis that is being considered.

What if we changed the null hypothesis to be the following

Players who do not use board position as an out of band signal either

1. Place boards deterministically (always in precisely the same position) OR
2. Place boards at random based on some continuous probability density function

In contrast, players who are using board position as an out of band signal will place the board precisely at one of a small number of discrete locations.

Note that the is no claim here about what precise piece of information is being signalled. We don't need to related board location to hand type. Rather, we are claiming that the behavior with respect to board placement is sufficiently unusual that this, in and of itself, is sufficient to establish that a pair is signalling.

Personally, I believe that this will make the stats and the analysis a lot more simple, with the obvious caveat that some people will find the claim that varying board position is equivalent to cheating too hard a row to hoe…






March 10, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> Still calling BS on the (wannabe) lawyers claiming doom and desctruction
> from lawsuits filed by (suspended) cheaters.
> Please provide relevant jurisprudence for your fantasy stories if you can?


Blubaugh v. Am. Contract Bridge League, No. IP 01-358-C H/G, 2001 WL 699656 (S.D. Ind. June 20, 2001) is the most obvious reference.

I believe that the ACBL folded under the threat of a lawsuit from Steve Sion. Other folks on this list are better positioned to comment on this then I.

It will be interesting to see how your brain is able to deal with facts that conflict with your preferred world view.
March 9, 2016
Richard Willey edited this comment March 9, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> So ask yourself this question: to whom did the cheaters cause foreseeable
> economic loss as a consequence of their wrongful behaviour?

In many cases, the cheaters belonged to a team with a sponsor.
The sponsor paid significant amounts of $$$ to win a title.
Said sponsors are being stripped of their titles.

Two of the pairs that are being prosecuted for cheating were rated as some of the strongest pairs on the planet. Presumably, the decision to cheat strengthen their perceived value and made it more difficult for pairs who did not cheat to gain contracts.

March 9, 2016
.

Bottom Home Top