Join Bridge Winners
All comments by Richard Willey
You are ignoring the author of this comment. Click to temporarily show the comment.
You've just hit on a real sore spot with me..

Transfer Walsh responses to 1C are so simply that they don't require any kind of suggested defense. They're also almost identical to MOSCITO style opening bids which are so insidious that no suitable defense can be devised.

Personally, I think that the difference in treatment has an awful lot to do with the fact that, on the one hand, Chip Martel likes to play transfer Walsh and, on the other, members of the committee have stated that they won't allow MOSCITO to be played in the US.
April 18, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
I have a real issue with creating different sets of rules for different classes of bridge players.

As a practical example, I vehemently dislike creating a special set of rules to punish players who forget a convention, but apply a completely different standard for players who forget a natural bidding sequence.

As a followup, I don't think that you could every impose this sort of rule for new players or for the duffers, so I don't think you should roll this out on a wide spread basis.

FWIW, I don't think its unreasonable to apply a different set of standards for certain types of events. For example, if you're playing in the Spingold or trying to qualify for the Bermuda Bowl or some such, I don't think its unreasonable to expect that you and your partner should have agreements for common circumstances.

Moreover, if you and your partner are suddenly forced to deal with some random convention, I don't think its unreasonable that there is a less of an expectation that your partnership have as comprehensive a set of agreements…

April 18, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
> My final thought is let's have a look at how casinos manage
> cheating risks, which obviously is at the core of their
> business model: cameras everywhere.

Casino's also invest heavily in image recognition software. (In a previous life I worked for MathWorks. It's interesting how some folks are using Image Processing Toolbox + Statistics Toolbox)

The reason that I bring this up is that camera's aren't enough. You also need a mechanism to move all the information into a database that you can then mine. This is quite complicated stuff. There's a lot of separate pieces of information to capture and you don't have good training sets.

I don't believe that the WBF is technically qualified to do this sort of work. Nor do I believe that they have the resources necessary to pay for such a project.

I could be be wrong. Maybe one of these days we'll see mention of a WBF project intended to address this issue. Until then, I'm going to promote the cheap, effective, working alternative that could be put into place next year without much effort or expense.





April 17, 2014
Richard Willey edited this comment April 17, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Screens might have once served a purpose. In this day and age, I fear that they are near useless against any determined cheats. It is trivial to build a low power transmitter/receiver into a item of clothing (I'd use shoes myself).

For the life of me, I can't understand why people are so willing to believe that the game is full of cheats while simultaneously acting as if they're all idiots. I said it before, I'll say it again.

The systems that the WBF have in place to prevent cheating are woefully inadequate. The only reason you caught W+E because they were stupid.

If you actually want to catch cheats, you're going to need perfect record keeping over large numbers of boards so you can detect pairs that are consistently “lucky”. Coupled with this, you're going to want to physically separate members of the same partnership to increase the power requirements for such a transmitter.

I would strongly recommend that folks familiarize themselves with various cheating scandals that have taken place in chess involving illicit electronic communications.



April 16, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Coming to this thread rather late…

In an ideal world, it might have been better to conduct an A/B type test.

1. Create two separate version of each hand.

Version 1 contains information from the auction.
Version 2 contains an additional information (does partner have shortage or is he balanced) / (does partner know that I have shortage or that I am balanced)

2. See whether there is a statistically significant difference in the lead by hand and see what lead was chosen at the table.



April 16, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
> The players deserve a “best case scenario” winner, and that
> would be the finalists. To deny them a title, rewards the
> offending team at the expense of everyone.

You might as well be handing out trophies at random.

I'm sympathetic to what you want to do, however, I think that its ridiculous to pretend that these sorts of proposals have merit, nor that any trophy awarded through such a system has real value.

FWIW, I spent some time this afternoon looking at cheating occurences in other mind games. There have been a significant number of chess cheating scandals that have involved electronic transmitters.

If people are really concerned with these sorts of happenings, its probably better to spend your time thinking about methods to make this sort of cheating more difficult rather than focusing about who should have gotten what meaningless trophy.

April 13, 2014
Richard Willey edited this comment April 13, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
I agree with the desire to have a clearly articulated policy in place for cases where a tournament has been “fouled” do to a cheating conviction. I also understand the appeal of simply adjusting all of the results up by one or some such. However, from my perspective, it is completely inappropriate to do so.

Bridge tournaments are highly random affairs. I agree that on average the most skillful teams will win, but there’s a lot of luck involved as well. Minor perturbations in the pairings and the schedule can ripple through the system.

I don’t think that you can play hypothetical “what if” games and credibly predict what the results would have been absent cheating. I thing that there are much more powerful arguments in favor of voiding the entire event rather than elevating other players.

I agree that this all sucks. I wish that it weren’t so. But having pairs cheat really spoils the entire proceedings.

This is part of the reason that I am so disappointed that players value pushing pasteboard rather than putting real security measures into place.
April 13, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
FWIW, I think that the WBF handled this in precisely the right way.

I can only image how unpleasant it must have been for the American team to sit down and play a match knowing precisely how you were being cheated on board after board. (I’m not sure if I would have had the patience for this). With this said and done, what makes this cheating prosecution so impressive is the level of rigor that was applied.

The WBF has a specific, testable hypothesis in place BEFORE any data was captured. From a stats perspective, this is much more compelling than “after-the-fact” data mining.
April 7, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
As I've said before, my primary concern isn't the use of cards per see, but rather close physical proximity between members of the same team.

Based on discussions on this forum, people have been suspicious of Elinescu and Wladow for years. However, they weren't convicted of cheating until

1. People kept track of the coughing
2. The WBF was able to collect confirmatory video

Lets assume that W+E were a little smarter and used a low power electronic transmitter in their shoes. Same code, but they push down on their right heel to send a bit of information. Their left shoe delivers a slight shock to transmit the signal.

Had they been using such a system, I'm guessing that no one would have been able to bring charges.
April 4, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
The $64,000 question involved deliberate collusion between the people running the show and the contestant. This is very different from trying to prevent a pair of contestants from cheating. The fact that you can't specify an adequate control for the first case doesn't mean that you can't improve security in the second.

FWIW, I currently work for the Information Security group at Akamai. I spend my days dealing with these types of issues. The following blog post provides a decent overview of the methodology that our group employs when approaching these sorts of problems.

https://blogs.akamai.com/2013/11/so-you-want-to-secure-something.html
April 4, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
There's a couple points that I think you are missing

1. Part of the reason to move to an electronic playing format is to improve record keeping. The best way to detect if something untoward is going on is via statistical inference, however, this works best if you have a lot of data to work with.

2. The reason that I advocate separating players on the same team by a significant amount of space is to increase the power requirements to send and receive data and make this easier to detect


April 4, 2014
Richard Willey edited this comment April 4, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
I can trivially build a pair of low power electronic transmitters/receivers that will fit into my shoes.

Protecting against such a system requires physical separation. You need to increase the distance between players significantly enough that you can listen for chatter.

Even with this, you're going to want a complete record of the bidding and card play so you can perform statistical analysis of the hands and look for folks who seem a bit too lucky.

April 3, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Couple additional points worth noting:

1. I am making recommendations for the top level of the game. WBF championships. The Vanderbilt. The Cavendish. Events where prestige and money create a strong incentive to cheat. No oneis suggesting that your local club or home game should be required to use computers.

2. A few years back, I compared the number of tables in play on an average day on BBO with ACBL Nationals. BBO had significantly more, and its only grown since then while the Nationals are shrinking. Folks are welcome to make claims about the “death of the game”. However, the area that is dying is the traditional F2F game.



April 3, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Elinescu-Wladow were stupid. You haven't found the smart cheats…

Please consider the fllowing:

http://bridgewinners.com/article/view/elinescu-wladow-were-stupid-you-havent-found-the-smart-cheats/
April 3, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
You are ignoring the author of this comment. Click to temporarily show the comment.
The most useful information is board results in comparable events. Lots and lots of board results.
April 3, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Question for the peanut gallery: Just how much interest is there in solving this problem?

I have access to some very sophisticated machine learning software at my current place of work. (http://www.redlambda.com/products/metagrid-platform/neural-foam/) better yet, I know their CTO. He's a fantasy football buff (He has the number one rated fantasy football team in the world according to CBSSports.com)

I'm quite sure that between us we can come up with a very accurate set of seeding estimates. Here's the rub:

Lets assume that I am able to come up with a seeding model, along with some kind of objective accuracy model. Is their any chance that this would be of practical use? If this would actually be used, I'd be happy to invest some effort… If its going to be ignored because Horn Lake likes the status quo, I'll pass…


April 2, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
I am often quite critical of the way that cheating charges have been levied and cheating investigations conducted. Its nice to see a case where the side making the claim seems to have done things correctly.

From my perspective, what I found most impressive is the fact that the USA team was able to present a testable hypothesis and that the WBF was then able to validate this during a future session.
April 2, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Interesting reading to say the least.

Professionally speaking, I like the way in which the WBF built their case.

They started by specifying a testable hypothesis: W+E are cheating using the following code.

They followed this by collecting data to verify this specific hypothesis.

The combination of specifying the code in advance and then collecting data consistent with this is very powerful.

In an ideal world, there are a couple additional points that I would have liked to have seen addressed:

• Were there any occurrences during which W+E were monitored that were not consistent with the “code”. I’m most certainly not alleging that there was selective presentation of data. However, addressing this in advance is good form
• A summary of the hands in question

o W+E were monitored over “foo” hands
o X hands were consistent with the code
o Foo – X hands were inconsistent with the code





March 28, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
If misinformation has been provided, the offending side is obligated to point this out at the first opportunity.

In this case, LHO should have noted that the alert was mistaken after the bidding had finished and before the opening lead was made.

It’s worth mentioning that the precise same procedure should apply regardless of whether the misinformation involved a conventional sequence or a natural sequence.
March 26, 2014
.

Bottom Home Top