[billw55]

So if one suspects confirmation bias, investigates, and finds confirmation bias, then ...

[mikeh]

phil_20686, on 2015-September-09, 04:09, said:

Comparing it to Ferguson is hardly an apples to oranges comparison. One is all about the context of a specific incident, the other has been repeated in reasonably controlled conditions hundreds if not thousands of times.

Code breaking, as with most learning techniques, comes in two stages. In the first you literally do attempt to analyse every movement to get the best explanation of the results possible. In the second, you take your hypothesis about the code to a second, unrelated set of boards, and see which parts of your hypothesis stand up to this independent test. If, broadly, your hypothesis is predictive, then you can be reasonably sure that you have discovered a signal. This is called "training your model" and "testing your model", and there is a usually a third part, validating. If we put woosley and ish in the "training your model", then approaching it with the belief that they are cheating is the right thing to do, the next step is to take the code to some unrelated set of boards and see if its predictive on boards that the code breakers have never seen.

I agree with you about how it ought to be done. But that isn't how either did it. When they looked at the 'new' set if boards, when the predictions were not borne out, they rationalized the results into being confirmatory anyway. They even purport to explain what F and S were thinking when they gave no signal or ignored a perceived signal. The sad thing is that there seems to me to have been no reason, other than bias, to have done so. Simply report the null results along with the actual positives, and there would still have been compelling reason to conclude that cheating was going on. By taking null results and twisting them to fit the theory, they risk having their credibility and reliability rejected should they be invited to testify, and that applies to anyone who uncritically endorses their approach. I hope that whoever presents the case against FS before any tribunal will do the job independently of the flawed approach we see on BW. I see that the IBF is subject to the Israeli judicial system. I deal with 'experts' as witnesses all the time. I would have a field day with Ish and Kit if I got to cross examine them, and I suspect that FS have counsel who are at least as competent as I am in that area.

[phil_20686]

mikeh, on 2015-September-09, 07:18, said:

I agree with you about how it ought to be done. But that isn't how either did it. When they looked at the 'new' set if boards, when the predictions were not borne out, they rationalized the results into being confirmatory anyway. They even purport to explain what F and S were thinking when they gave no signal or ignored a perceived signal. The sad thing is that there seems to me to have been no reason, other than bias, to have done so. Simply report the null results along with the actual positives, and there would still have been compelling reason to conclude that cheating was going on. By taking null results and twisting them to fit the theory, they risk having their credibility and reliability rejected should they be invited to testify, and that applies to anyone who uncritically endorses their approach. I hope that whoever presents the case against FS before any tribunal will do the job independently of the flawed approach we see on BW. I see that the IBF is subject to the Israeli judicial system. I deal with 'experts' as witnesses all the time. I would have a field day with Ish and Kit if I got to cross examine them, and I suspect that FS have counsel who are at least as competent as I am in that area.

Sure, but we have roughly an infinity more hands to test on......

[PhilKing]

phil_20686, on 2015-September-09, 07:24, said:

Sure, but we have roughly an infinity more hands to test on......

I don't think there are that many sets on video (unless we include unreleased footage).

[hrothgar]

mikeh, on 2015-September-09, 07:18, said:

I agree with you about how it ought to be done. But that isn't how either did it. When they looked at the 'new' set if boards, when the predictions were not borne out, they rationalized the results into being confirmatory anyway. They even purport to explain what F and S were thinking when they gave no signal or ignored a perceived signal. The sad thing is that there seems to me to have been no reason, other than bias, to have done so. Simply report the null results along with the actual positives, and there would still have been compelling reason to conclude that cheating was going on. By taking null results and twisting them to fit the theory, they risk having their credibility and reliability rejected should they be invited to testify, and that applies to anyone who uncritically endorses their approach. I hope that whoever presents the case against FS before any tribunal will do the job independently of the flawed approach we see on BW. I see that the IBF is subject to the Israeli judicial system. I deal with 'experts' as witnesses all the time. I would have a field day with Ish and Kit if I got to cross examine them, and I suspect that FS have counsel who are at least as competent as I am in that area.

Hi Mike

FWIW, I agree with everything that you have written above. I've been trying to caution Kit about his approach with limited success. I am very concerned regarding what a talented defense attorney might be able to do in a court of law.

With this said and done, I'm not sure whether this should all come down to a question of "law" and "Beyond a reasonable doubt". It occurred to me that casinos have enormous latitude to bar players from playing their games. I wonder whether there might be an option to change the institutional structure of organized bridge to achieve the same end.

For better or worse, I don't need to be 100% certain that L-S were cheating to be willing to bar them from participating.

[MrAce]

hrothgar, on 2015-September-09, 07:53, said:

Hi Mike

FWIW, I agree with everything that you have written above. I've been trying to caution Kit about his approach with limited success. I am very concerned regarding what a talented defense attorney might be able to do in a court of law.

With this said and done, I'm not sure whether this should all come down to a question of "law" and "Beyond a reasonable doubt". It occurred to me that casinos have enormous latitude to bar players from playing their games. I wonder whether there might be an option to change the institutional structure of organized bridge to achieve the same end.

For better or worse, I don't need to be 100% certain that L-S were cheating to be willing to bar them from participating.

Yes, I read what you wrote Richard. I also tried to warn Kit despite my lack of knowledge about statistics. So I showed you and Mr.Lawrence as reference and Mike as an expert player, without giving his name. I also want, just like you and Mike that if they are going to go fully statistical, they better make a case that is statistically very strong to be beaten by a good lawyer and an expert on statistics.

[hrothgar]

MrAce, on 2015-September-09, 08:33, said:

Yes, I read what you wrote Richard. I also tried to warn Kit despite my lack of knowledge about statistics. So I showed you and Mr.Lawrence as reference and Mike as an expert player, without giving his name. I also want, just like you and Mike that if they are going to go fully statistical, they better make a case that is statistically very strong to be beaten by a good lawyer and an expert on statistics.

I am not a lawyer (though I am taking classes at Harvard Law and might even eventually get a JD)

Were this to go to trial, my advice would be simplify, simplify, simplify.

There are way to many conjectures and claims floating around out there.
You don't need to convict L-F on everything. One hand was sufficient for the racecars.

There's no point in introducing claims about coughing and clubs if it won't help your case.
Identify your most compelling claim and focus on this.

FWIW, I am very glad that Greg Lawler is participating in the thread, since his qualifications dwarf mine.
(I am even more glad that he hasn't mocked any of my claims)

[mikeh]

billw55, on 2015-September-09, 06:32, said:

So if one suspects confirmation bias, investigates, and finds confirmation bias, then ...

so true. However, I had been encouraged by Kit's early postings and expected that he was going to do an objective job. I went to the videos with a sense of expectation that the code had been cracked and now we'd see the most important part: confirmation. What I saw made me exclaim out loud...the taking of the non-spade signal and asserting that the failure to meet expectations was confirmation. My heart sank as I continued, and I was incredibly disappointed to see Kit adopt Ish's approach of literally making up internal thinking processes by FS to explain away why a predicted signal wasn't made or wasn't followed when it seemed to be made. Trust me, I was hoping to see confirmation of cheating, not confirmation of bias. I think I saw some of the former and I know I saw lots of the latter. I have long respected Woolsey, so it was disappointing to see such incompetence.

[billw55]

I haven't read Kit's material, so forgive an obvious question. Was he checking for and reporting matches of the hypothesized signals to (1) the actual lead, or (2) the nonleader's hand? This seems like a big difference.

[cherdano]

Mike, I agree that Woolsey could have done a better job. (Especially given that he sometimes seems to project himself as the ultimate arbiter of such cheating accusations.) But it's also not nearly as bad as you make it out to be! For example, the hypothesis he was testing was clearly whether 3rd hand was making a signal. So any comments about the opening leader were just illustrative side remarks.

Also, the hypothesis he was testing was: "If given the chance by opening leader, 3rd hand will make a signal. His options are 1. Leave tray on the table to indicate no preference. 2. Take tray and keep board on his side to indicate club preference. ..." Any board where 3rd hand has no preference and does leave the tray on the table is confirmation of that hypothesis! Let me rephrase: if 3rd hand leaves the tray on the table on precisely the boards where an expert would not want to make a signal, then that is a coincidence that is much more likely to happen if they cheat, then if they not cheat. Thus, it is (probabilistic) evidence of cheating. You instead seem to want to ignore it!

[cherdano]

I also don't understand what all this talk about a trial is about. Nobody plans to take F-S to a criminal court! If they get banned by, say, the EBL, they may sue. But I doubt the court would rule on whether the EBL had proof beyond reasonable doubt - it would just rule on whether the EBL followed proper procedure.

I actually think the case against F-S is more solid than the case against the doctors (just because board placements are more objective to verify on a video than who coughed how many times).

[cherdano]

phil_20686, on 2015-September-09, 04:09, said:

Code breaking, as with most learning techniques, comes in two stages. In the first you literally do attempt to analyse every movement to get the best explanation of the results possible. In the second, you take your hypothesis about the code to a second, unrelated set of boards, and see which parts of your hypothesis stand up to this independent test. If, broadly, your hypothesis is predictive, then you can be reasonably sure that you have discovered a signal. This is called "training your model" and "testing your model", and there is a usually a third part, validating. If we put woosley and ish in the "training your model", then approaching it with the belief that they are cheating is the right thing to do, the next step is to take the code to some unrelated set of boards and see if its predictive on boards that the code breakers have never seen.

Actually I am with Mike here - Woolsey was claiming to do the second part.

[Flem72]

hrothgar, on 2015-September-09, 08:41, said:

I am not a lawyer (though I am taking classes at Harvard Law and might even eventually get a JD)

Were this to go to trial, my advice would be simplify, simplify, simplify.

There are way to many conjectures and claims floating around out there.
You don't need to convict L-F on everything. One hand was sufficient for the racecars.

There's no point in introducing claims about coughing and clubs if it won't help your case.
Identify your most compelling claim and focus on this.

FWIW, I am very glad that Greg Lawler is participating in the thread, since his qualifications dwarf mine.
(I am even more glad that he hasn't mocked any of my claims)

Were this to go to some kind of formal hearing, I'm certain there would be two lines of evidence besides the testimony of the participants: statistical analysis and expert opinion. At first I was bothered that so many top-level players were willing to say 'so clear they are cheating, we don't need a how' (not least b/c they appear to have no defense, on this strategy, against a slander per se action), but, on reflection, who gonna be the experts for FS?

[cherdano]

Sorry one more post: the argument "this expert testimony would never hold up in cross-examination" has exactly one place. Namely when you are considering whether to bring this expert testimony to a court where the expert would be cross-examined.
There is all sort of bullshit (including a lot that used to be called "forensic science") that holds up well in cross because it's internally consistent and because lawyers are unfamiliar with its basic problems. And there is all sort of testimony that is very much on target, but would fall apart in cross-examination because it includes irrelevant details, some of which may be not correct.

If you just care about finding out the truth, this question is a complete waste of time. (And if it's your main frame of reference for judging what is true - well then I can't help you.) It's about as relevant as "would make for a good soundbite" is for understanding what makes good policy.

[Flem72]

cherdano, on 2015-September-09, 09:52, said:

Sorry one more post: the argument "this expert testimony would never hold up in cross-examination" has exactly one place. Namely when you are considering whether to bring this expert testimony to a court where the expert would be cross-examined.
There is all sort of bullshit (including a lot that used to be called "forensic science") that holds up well in cross because it's internally consistent and because lawyers are unfamiliar with its basic problems. And there is all sort of testimony that is very much on target, but would fall apart in cross-examination because it includes irrelevant details, some of which may be not correct.

If you just care about finding out the truth, this question is a complete waste of time. (And if it's your main frame of reference for judging what is true - well then I can't help you.) It's about as relevant as "would make for a good soundbite" is for understanding what makes good policy.

In a technical case, we trial lawyers are generally only as good as the bullshitters on our side. Litigators don't know much about anything, they just learn it fast...

[zasanya]

All of us who are participating in Transnationals at Chennai have to sign a commitment form. It has the following 2 clauses among other things
"2. I am aware that the Championship is governed by the Laws of Bridge, the Conditions of Contest, the WBF Disciplinary Code and all other Rules & Regulations of the World Bridge Federation (WBF) and I undertake to comply with such rules, all of which are fully known to me.

3. I therefore agree to be submitted to such rules, regulations, and procedures to the jurisdiction of the bodies and persons that are in charge of applying them. After the exhaustion of internal remedies within the WBF system, I acknowledge that the Court of Arbitration for Sport (Lausanne, Switzerland) has ultimate jurisdiction for all disputes in connection with the Championship and agree that recourse to the ordinary courts of law is prohibited."
Do the players submit any such forms in other major bridge events? If yes then recourse to ordinary courts of law is not possible?

[mikeh]

cherdano, on 2015-September-09, 09:52, said:

Sorry one more post: the argument "this expert testimony would never hold up in cross-examination" has exactly one place. Namely when you are considering whether to bring this expert testimony to a court where the expert would be cross-examined.
There is all sort of bullshit (including a lot that used to be called "forensic science") that holds up well in cross because it's internally consistent and because lawyers are unfamiliar with its basic problems. And there is all sort of testimony that is very much on target, but would fall apart in cross-examination because it includes irrelevant details, some of which may be not correct.

If you just care about finding out the truth, this question is a complete waste of time. (And if it's your main frame of reference for judging what is true - well then I can't help you.) It's about as relevant as "would make for a good soundbite" is for understanding what makes good policy.

I think we agree on more than you may think, but we disagree about how it is that the bridge authorities will have to deal with FS.

I readily admit that I have no expertise in how the WBF or the IBF, to name two likely authorities, have to act. I do see, from BW, that the IBF processes are subordinate in some fashion to the Israeli court system, but I don't know anything much about the governing principles of that court system. I do assume that the system operates on more or less the English/US presumptions and basic evidentiary rules. That may or may not be valid but in the absence of reason to reject it, it is all I have to go on. I assume that the WBF is not subject to any particular national jurisdiction, other than perhaps that of the country in which it has its headquarters.

I am somewhat familiar with the ACBL system, at least at the District level, having been counsel on one hearing and having been part of the list of possible jurists for the District for many years, but never actually having to sit on a hearing.

I am generally familiar with the basic concepts of 'natural justice' as in play in many parts of the Commonwealth and am fairly sure that these rules are similar to some of the due process rules in the US.

All of this is by way of saying that I strongly suspect that if FS insist upon formal process by any bridge authority with the power to limit their ability to play professionally, there will be a hearing run according to procedural and evidentiary rules that will parallel and may well be essentially identical to the rules that would be in effect in a court hearing...in a trial.

Now, the standard of proof is an interesting but very different issue. Procedural and evidentiary rules govern what evidence is admitted and how evidence is tested. Balance of proof goes to how satisfied the tribunal has to feel about whether the accusations have been made out. In the jurisdictions with which I am familiar, the accuser, in a civil matter, only has to show that it is 'more likely than not' that the case is made out...the balance of probabilities approach. In criminal matters, the standard is proof beyond a reasonable doubt....which, by the way, is an extremely difficult standard to explain in simple terms and is often grossly misunderstood by reporters covering criminal trials.

My main concern is that regardless of the standard of proof, there will eventually be a hearing constrained by rules that to the layperson may seem almost designed to limit the evidence that is to be considered....and in a very real way that impression is correct. The rules exist to ensure, as best as rules created and enforced by humans can do, that the issues are decided on 'real' evidence, and not rumour, gossip, and so on.

So, if our concern is to rid high level bridge of cheaters, we can do this in one of two ways.

We can lay charges and have due process (a phrase I don't like and never use in my practice, but we have a lot of Americans involved and they love that concept) and hope for the just outcome or we can engage in open or hidden character assassination, with no real opportunity for the accused to get a fair opportunity to refute the charges, and/or hope that all the sponsors in the world will decide never to hire them again. Good luck with that, if the experience of Katz, Cohen (not the famous and highly ethical Larry we all respect!) and Cokin are any basis for comparison. They all resumed professional bridge careers after getting back in the ACBL...Katx and Cohen did so after suing the ACBL!

Personally, I want to see a hearing, with all of the safeguards and protections that would accompany such a hearing, and then, if the evidence warrants, a conviction and a banishment for all time from the game. What I don't want to see is a hearing in which the accused are convicted on the basis that 'everyone knows' that they are cheats.

Since my take is that any hearing will, at least in form, be the 'due process' type of hearing, I do think that my concerns about process to date have a real basis. That isn't to say that the prosecutors won't do it right. But while the hearing may be constrained by proper rules, and may be subject to judicial review if done improperly, there is a real risk that the 'prosecutors' will be amateurs. They either won't know how to properly put together a case, because they lack the legal training, or they won't understand the bridge issues in a way that allows them to articulate the reasoning that ought to flow from the evidence (or to cross examine any evidence from the FS side, including FS themselves).

The fact that so far the experts who have, commendably, spent a huge amount of time on this matter, are screwing up in profound ways, in the context of admissible evidence, doesn't create a lot of confidence that a case will be made, as opposed to there being a case that could be made....I am pretty sure of the latter, worried about the former.

I hope this explains to Arend and to others of like mind why it is important to worry about how the proof is being put together. Or do we want to live in a world in which we punish people without a hearing and on obviously biased 'evidence'? I don't. Do you?

[captyogi]

As regards Coughing ( not necessarily in this case, otherwise also ),
What we can watch is if the players coughing during play, were coughing in between rounds also or when they were resting pair.

If otherwise normally they were not coughing that day, we can safely say, why you get cough only during Round ?

Same thing goes for other acts, like scratching etc.

[captyogi]

As regards Code Breaking Method, What Do We Expect should be the Success Rate ? 75 % ? 80 % 85 % 90 ?

[captyogi]

As regards Lawyer taking Cross of an Expert Witness in Court, I can assure you, Lawyers can tear apart the Expert Witness, they go to any level, they build up Cross slowly slowly, Experts Witness ( not only in Bridge, each and every field ) are thorough in their field and knowledge, but facing Lawyer standing in witness box is totally different ball game altogether, Expert Witness might not be Psychologically strong enough to receive missiles from unexpected corners, and once he has been demoralised, every likelyhood he will faulter in answering questions in his own field, Defence Lawyer's mission accomplished.

So anybody planning to go as an Expert Witness newly, i.e. you are not a regular professional Expert Witness, better Do lot of homework with your own Lawyers. And get trained what questions Cross Lawyer is allowed to ask, which he might ask, only to provoke, instigate or destabilize you.