Join Bridge Winners
A New Structure
(Page of 15)

Lessons to Date

Depending on the final verdict (and someone will have an opinion to the contrary, regardless) what will have occurred is either:

(a) A victory for justice with citizens taking to the streets with pitchforks and torches.

(b) An online lynching.

Now, to give all those lawyers, who were horrified at how this started, their due (sigh. . .), it is quite possible that a system that allows for (a) can easily become (b). And sometimes, even a victory for (a) can degenerate to (b) and a slew of innocents executed. Just think French & Russian revolutions.

So yes, we do need proper structures, procedures and (sigh. . .) lawyers. But we need the lawyers in the context of a system that actually preserves people’s rights without it becoming a slave to process. Nor can we let process create a vehicle that can be hijacked for the purpose of burying or obfuscating the truth.

There is a lot to be said for keeping cheating, ethics and recording issues under wraps. The rest of us can all be ignorant of any accusations or details and all the proceedings closed. Then, the powers that be can just publish a guilty blurb in say, the last bulletin of the NABC without any supporting information.

You may want to ask Mike Passell how well that sort of system can work.

"There is only one good, knowledge, and one evil, ignorance." - Socrates

We must consider that if a pair is guilty of cheating, that it can take an extraordinary effort by many, many people to prove that point. Lets assume innocence for now for the current pair under discussion as there have not been any completed official proceedings by authorities. If they are innocent, despite all the on-line evidence to date, there has to be a mountain of contrary evidence yet to come out. Either way, it should be obvious that no small, behind closed door committee is ever going to have the investigatory powers nor the resources to make certain cases.

We need a workable system to protect the integrity of the game. I think that the overall crime rate is very low. So does Jeff Meckstroth. If you Google “safest country in the world”, you will find it is Iceland. (Congratulations; apparently volcanoes are not part of the metric.) Iceland still has a police force and court system.  I agree completely with Tarek Sadek that prevention is the most important issue.  Smokey the Bear and preventing fires is the first line of defense.  But we still need firefighters.  And a low crime rate is not a reason for being lax.  

The absolute requirement for justice is knowledge. Knowledge is not possible without data, facts and analysis. And not just some data or some facts. When one is trying to make a determination, the smaller the sample size the greater the chance for error. That means a system that might not convict the guilty or possibly damn the innocent.

The following is a sketch of a workable system that interfaces regulating authorities (R.A.), members, directors, expert committee members, all the world’s top experts, and,

of course. . .

bridgewinners/RA_support (Ta da!)

After all, how did we survive before this site was created?

In a nutshell, here is what I would want from a system if I were a regulating authority (RA)

  • Convenience
  • Privacy & Security, yet. . .
  • Communications and access by relevant personnel
  • A complete record of all proceedings, analysis and data

First, let me make clear that everything is really under the auspices and control of the appropriate regulating authority. Bridgewinners simply acts as the IT support and user interface. All the data is property of the specific RA handling the complaints and filings. I believe that this data can also be copyrighted. That would means that no one who had access could disseminate the specific data without automatically being in some legal trouble. (If there are any attorneys out there who I haven’t completely offended yet, perhaps you could confirm this point. Obviously a wiki-leak anonymous leak is a bit harder to address.)

The advantage of Bridgewinners is they have a lot of the system and supporting software actually developed, meaning we could save. . . (let me calculate this). . . probably at least a million and half dollars in software development. It would be much cheaper for the R.A.s to employ BW to provide the interface and ongoing tech support. Plus, almost everyone already knows how to use and post to BW. You can IM or email reply via the system, etc. Drop in hands and links. A “recorder form” would essentially look like an article. You can run like 20,916 pages without crashing the server. I know because some of my posts are that long.

This is the 21st century. The idea of paper forms, telephone calls and meetings during the NABCs is, well lets just say that stuff it is like Culbertson’s Approach Forcing System. Once upon a time, bear skins and stone axes was all we had.

Time to move forward

Going Through The Process

A player, lets call him BB, thinks a pair - hereinafter: Slee & Zee - might be cheating. This is the most expansive scenario. The ACBL (or other RA) could use the same format for ZT or those all those evil psych recordings. Obviously all the files could be organized by class (suspect, conduct, psych (yeah, Ozzie is rolling his eyes someplace)) and by type (open, pending, active, resolved, archieve).

In the case of a recorder form, one contacts the tournament TD. He opens a case file with everyone’s player number, date & tourney info. The player can access the form within, say 48 hours to complete it. It is in the system ready to be on file or acted upon. The TD can add any facts he knows to the filing. The proper ACBL committee gets an e-mail alert and can look to see if it is a problem that needs to be addressed, whether any similar issues arose with this player or pair, etc. The TD and everyone on the committee has immediate access to the data, analysis and witness testimony and commentary.

If the player does not complete the recording, it gets auto-tossed. If a player does not know how to use the internet we can give them a write in form and key it for them. Although that could be evidence that they also don’t understand what they thought they saw at the table.

Back to Cheating Issues

BB can file his collection of evidence and his supporting analysis online. The expert committee - which is watched by RA officials - can decide a course of action. If the evidence is deemed sufficient for action, it could be referred to the appropriate CDR people. That sort of structure could be a la Chris Willenken proposal, for example.

The expert committee might also want to launch and investigation or a general review. In an investigation - say some coughing players need to be watched - that data can be recorded by officials who have been brought into the inquiry by the committee. They also have easy access to information and data collected by other sources.

Keep in mind that a comprehensive system starts with addressing prevention (T and could end in legal proceedings. In-between we need reporting, review, additional data collection, investigation, analysis and inquiry. A report might That investigation might start off covertly

If committee feels the matter could benefit from an expanded review, the issues and data could be accessed for evaluation by either additional experts brought in at the behest of the expert committee or it my be more open an any interested WC expert who has been approved could participate. The player or pair who is the subject of the expanded review would, of course, be notified and able to respond to any inquires as well as submit any counter data or examples they feel relevant.

Since the RA is controlling access will simultaneously enable anyone who can provide data or insight, this will be a very efficient material data collection process. All the commentary and data is available to everyone. We know who said what to whom. There is no behind the back whispering now. Also, if anyone wanted to revise or rescind their comments, they would append subsequent commentary. For example, Rodwell had a post that he properly went back and added an edit that said something to the effect of: “This analysis is no long valid in light of. . .” Which is exactly proper methodology to follow.


This creates a now open source discovery process. Ultimately the seated expert panel is responsible for making a recommendation to their RA. If a conduct case is actually heard, that process would be as it is now; an in-person, formal hearing. The “prosecution” would be presenting what they considered the relevant case information from what has been compiled. Nothing that has not been “discovered” could be presented by the prosecution during the committee conduct hearing.

If the official inquiry resulted in a not guilty verdict, that transcript would be attached to the head of the discover data and anyone who had participated in discovery and analysis would be informed. The file could be accessed later, if needed, by RA & relevant committee members.

If the inquiry ended in some form of disciplinary action, the verdict would either seal or open (at the defendant’s request(* ) the discovery files to any interested party. (* if a pair or more and just one defendant wants it open, it should be opened).

Note that im a case file of something like: “Board Dropped on Floor” is opened to all by the defendant, anyone and everyone can now look at the entire discovery file. No speculation, no potential re-writes of to what was actually attested and claimed. The case and verdict is properly detailed. Because a proper verdict should always sum up the case for the decision. Not some bulletin blurb like: “Doctor Stockmann has been declared an enemy of the people”.

I think this overall method, while obviously the details need to be nuanced, strikes the best balance between protecting people from broadcast allegations while allowing for an open, broad and complete investigation that will be often required to make a determination.

As a concomitant advantage, all the RA s around the world can save a lot of money, now spared from the costs of developing their own network setup. While Bridgewinners can make a few bucks providing the service. (My information - can’t vouch for accuracy - is Steve will work for food and cans of Red Bull, real talent like Greg costs a little more.)

This structure provides for easy and organized access by authorized personnel at every stage. It would contain everything that has been alleged against a player. If that is determined to be baseless, it stays out of view of gossips. And if a defendant feels railroaded, he can demand it be open for examination. Secret tribunals and proceedings not generally being a favorite of the not-guilty.

Ultimately, some cases require an extraordinary amount of work and analysis by many experts to produce the proper outcome. One needs a structure to enable that effort. The bridge world can benefit from a properly structured open source system in these matters.

Related Observations -

Extended committee expert support has been heretofore referred to as “Top 100 Players”. This was coined by LC simply to make a point (but picked up by other posters, such as myself) of the skill level required to make evaluations. I think LC and the rest of us will agree while that is a great designation to make a point, it is a bad title to use in real life.

The way I would handle extended experts is based upon approvals from the original expert committee. If half the original group ok’d a person, that should be enough. Note that doesn’t require a committee meeting. An application can be made on-line. Also, anyone who has entered a recording obviously has access to at least their segment of data for additional psoting or replies.

From Magnus Olafsson via Mangus E Magnusson, #9 Delay mechanism for all matches on BBO (and other similar sites) is a very good idea. Though I would hate if the commentators actually have advance knowledge before the transmission delay because “the curse” is part of the fun :)

On the plus side, if the segment started with say a half-hour delay, it could finish with a 10 minute delay and the broadcast move along a little more briskly for the viewers.

#5 - Lead recording is critical. Yes, as David Burns points it will often be the J-. But this is more important for better players and they will get it right. Also, the system can easily tell you what N/S pairs are not properly recording leads. So it can be made right.

Having access to as much digital data as possible is critically important. However. . .

Statistical Analysis Issues

Subjecting data that you understand to statistical analysis can prove a many things.

Subjecting data that you do not understand to brute force analysis will generally only prove one thing - that you do not understand the data.

For example, I just number crunched ALL the weak-two bids from the last 40 years for high-level players. And the most SUSPICIOUS DATA EVER points the fickle finger of fatuous facts at. . .


Larry N. Cohen!!!


My computer modeling found that Mr. Cohen might pass a weak-2 bid with 17 HCP!! Yet on another hand he might bid game with a mere 9-10 HCP and not even with 4-card support. (What happened to THE LAW?) ?


And I have literally hundreds of examples to prove this. All these hands with Marty Bergen and Ron Gerard. Are these partners in on it too??

(Just in case anyone doesn’t know, two Bergen preempts put together is still a pass from Gerard.)

OK, so you say “minor glitch, lets control by partnership”.

Next up with anomalies is Robson-Forrester. Andrew’s preempts may well look like a Bergen preempt. But Bergen was “colors are for children” and Andrew is all about timing, position and situation. Meanwhile, Tony’s preempts are usually middle of the road. Each player preempts according to his style and the other knows it.

So adjusting for the partnership is not enough. Because the Robson-Forrester data has to be adjusted by which player bid what. Also what was the vulnerability and position?

If you think HAL is going to number crunch a pile of data and come to any conclusions, you will likely find something like HAL decides not to open the pod bay doors.

Or, what else might one find?

My Old Bridge Partner Marc. . .

. . . had a college roommate who managed to get a lot of computer time on the University mainframe (PS’s were just coming out). His roomie crunched a ton of random data and discovered a huge problem to which he dedicated his stat thesis.

He discovered a nearly 1.0 correlation between (if memory serves) the change in the rate of change in the suicide rate and which league won the World Series. Evidentially, when the NL won, the change in the rate of change increased positively. And it was always conversely when the AL won.

His proposal to solve this problem was to ban the NL from competing. Obviously the AL fans took the game more seriously and the WS should be limited to just to AL teams to keep distressed and disappointed people from offing themselves. It isn’t that ridiculous a proposal. Heck, right now we don’t even invite anyone from any other country to play. So what is the big deal about not inviting the NL? They only won 7 of the last 29 All-Star “games” anyway.

(My Chicago Cubs, saving lives since 1908.)

Compiling as much bridge data as possible is necessary because that material might need to be reviewed. More evidence is more compelling than limited data. But unless one knows what they are actually analyzing and why, then any statistical analysis useless. If one does not have context, all one actually has is a random number.

Casino Methods

That casino or poker rooms employ some sort of statistical analysis to try to spot cheats is first of all easy. This is because the underlying data is supposed to be random. We therefore already understand the data without needing to know virtually anything about the game.

Bridge is not random.

(Well, maybe at the club. But not at higher levels)

Also, the casino decision process is based on completely different metric: MONEY. I know of examples where a casino was so incensed to f-i-n-a-l-ly identify team blackjack play, that frustrated they were not smart enough to actually identify the spotter, the casino barred and threw everyone at the table out.

Why? First, because they could. They have the legal authority to do so in Nevada. No reason ever needs to be given. But the other reason is it makes FINANCIAL sense to screw innocent people. Not everyone would be a spotter; that would incredibly dumb. The casino knew that but their priority was to rid themselves of the guilty party. The people whose vacation the casino just ruined were never going to lose anywhere near the money the card counters were going to take from them.

And card counting is not even cheating. Casinos just react to it like using your brain should be illegal. The good news for the casinos is that based on actual data, there are lots of law-biding citizens. But be aware that the casino is simply making a financial decision. It has nothing to do with justice.

Understanding the data is crucial. The casino personnel watched while a team took $100+ grand before they figured out what was happening, I know of no competent blackjack player who cannot spot a card counter nearly instantly.

I remember playing a double-deck game and the Chinese man at the other table must have had a huge count. (I wasn’t doing the actual math because “I never count cards at blackjack” - is that a self-serving statement?) Anyway, right before the cut card he took every single chip and flooded the betting ring with a huge pile. I nearly fell out of my chair. He hit.

So I watched the next shuffle and before the end, I knew what the casino had figured out over time with a pile of statistics gathered over many sessions. He was a crazy gambler and was eventually going to lose. And hour later he was broke. My sample size was probably too technically small - a dozen hands, or so - but I don’t need no stinkin’ computer.

Poker Room “Surveillance” Might Be Even Lamer

To begin with, the casino’s interest is still financial. All they care about is that people show up, gamble and the casino can rake the pots. It really doesn’t matter to them from a financial perspective whether or not other players are cheating. The casino will make the same rake regardless of who wins the pot. Cheaters simply cheat other players in poker, not the casino.

But when other players think someone is cheating, it now does matter because these legit players will stop showing up to play. In the end, it only matters to the casino that it appears to players that the casino is policing the game.

Now maybe the casinos are. Poker is still somewhat random. Your cards are supposed to be random. You need to “shift gears” so players cannot be sure what you do. That isn’t random, but which players at the table are involved in pots over time should be predictable over time. When player x is too often betting against y with anything when z actually holds a hand, we have collusion.

So yes, it is a bit easier to police a poker game with statistical methods. And maybe they do. But in the end, it does not matter to the casino whether a player they bar was actually cheating. Only that it makes other players happy that “something is being done”.

Do we want that? In reality, for all you may know, the floor boss was in cahoots with the real cheaters and he trumped up charges against a couple of guys who were suspicious of the real cheaters to get the complainants out after they reported suspicions to their pit boss friend. “Yep, we barred those two. And you know what? They had the gall to claim OTHERS were cheating. Imagine that!”

And you cannot prove that was not what happened because you do not have access to the data, do you? You are just reading the blurb in the bulletin and taking it on faith. Because hey, what are the odds that people running a poker site would actually cheat customers?

Valid Casino Comparison

One very useful activity that the casinos perform in their effort to identify and catch cheaters is they share information with a company called Griffin Investigations. This is a way of pooling all the data and information from all the various casinos into one bucket.

That arrangement is not without some controversy. There is now a private investigation company compiling dossiers on citizens. (OK, not maybe so controversial given Google, Facebook etc.).

But the casinos recognize how critical it is to compile data and share information. We should too. Notice that in this proposed arrangement from bridge, the data and access privileges is under the control of a regulating authority. Though the RA cannot keep that data secret if the target of an administrative proceeding wants it opened. So we have privacy when we need it and an open process when the accused requests it.

Where’s the Beef?

Everyone wants the smoking gun. People feel more comfortable knowing about the how. But, to repeat what Justin Lall posted: “Having to prove how is a totally broken system”. And it is perfectly possible to prove a crime was committed without needing to demonstrate how.

Thomas Capano was a former deputy AG of Delware. He was convicted of murder for a 1996 murder and given (ending with after appeal) a sentence of life without parole in a case where a body has still never been found. I am not forensic expert, but I suspect that absent a body, how is a little difficult to prove.

Just Google “murder convictions without a body”. It has been done in France, England, Australia, New Zealand in addition to the United States.

A jury simply takes the possibility that the alterative explanation for each circumstantial piece of evidence might be true and multiplies them together. When it multiples out to “impossible”. Obviously, we need to know what data is relevant in bridge. That requires as much data as we can collect and often an expert understanding of whether it is relevant.

Force Multiplier

The ability to link and make available hands, data and information from players around the world is going to be critical for inquiries. Our NABCs are open. So are the European championships. High-level players play and reside all over the world. Little pieces of paper and committee meetings in person are passe. And certainly counterproductive from a discovery standpoint.

It seems ridiculous that players could cough for so long and get away with it. But the reason was the old system stinks. Lots of people suspected, but since no one could prove anything, it went on and on.

Then Eddie Wold breaks the code while simultaneous playing bridge. Ask yourself how much data did Eddie actually have before he figured it out? I know the answer is a &^#-load less than what was actually available. Had this information been collected and available for wider analysis, someone and they do not need to be as good a player as Eddie - could have possibly figured it out earlier. This is because they would have a fair bit of data to examine.

A couple of guys coughing in a match is not unusual. The more data one has of coughing, the less likely they were sick in each of the last three tourneys. Last five. . .? The human brain works on pattern recognition. The more data, the easier it is to see a pattern.

Even if a cheater is varying the code with a cipher shift, there still needs to be a method of communication. So there has to be a pattern to tray movements, coughing, finger or p