Join Bridge Winners
All comments by Richard Willey
You are ignoring the author of this comment. Click to temporarily show the comment.
1. For the overwhelming majority of players, the only exposure that they have to “top level events” like the Bermuda Bowl is via Vugraph or event write ups in the Bridge World of The Bulletin. if anything, migrating to an electronic playing environment will dramatically improve these aspects of the game because you get comprehensive Vugraph for free and a complete set of records for after the fact analysis.

2. From what I can tell, the main thing that the top level players want is an end to cheating in top level events. I am sorry to say that I don't see any good way to square the circle. if you want a nice friendly social game, you're going to continue to get taken advantage of.

3. As I have mentioned several times before, I'm not proposing that any of this gets rolled out either in clubs or low level events.






Sept. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Hi Gene,

There very fact that you are asking for a foolproof / perfect solution means that you aren't qualified to engage in a discussion about security. Nothing in security is “foolproof”. Any system can be broken. The only thing that you can do is make things more difficult for the adversary. (I recommend going and looking at the way in which safe's get rated. The ratings describe how long a given safe is able to resist a given set of attacks).

With respect to your second set of demands:

I envision the following type of system for top level bridge events (the Bermuda Bowl, the Cavendish, the three major NABC titles, the European championships, the Transnational teams).

1. The entire event, from start to finish, is run over BBO (or some equivalent). The event is run off a local server with the option to fall over to a networked server on the internet.

2. Hands for each round are generated at the start of the round, seeded using a physical source of entropy.

3. All North's are physically segregated in one room. All South's are physically segregated in a second. East's in a third. West's in a fourth.

4. Participants bid, play, and communicate using tablets provided by the event organizers. This information can be broadcast to the internet with a time delay.

With this said and done, I am not demanding that this exact system gets implemented. What I do consider to be essential is that the powers that be commit to implementing some equivalent system AND they kick off discussion to design this.

If the WBF or whomever wanted to hire me to design them a system, I'd be happy to do so. However, they'd need to hire me away from my current gig doing this same type of work for Akamai and I don't think they're going to match my salary.




Sept. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
One reason that I have been stressing physical separation for members of the same pair is to force the use of more powerful devices that will be easier to catch via a “sweep”
Sept. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
This Is Why We Can't Have Nice Things

http://www.imp-bridge.nl/sites/default/files/Fantunes.jpg
Sept. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Video tapes work great if the cheaters are stupid enough to rely on primitive communications channels like the placement of cards or bidding boxes or what have you.

People have been using electronic devices to cheat at chess, backgammon, blackjack, and the like for decades.

Why do you believe that bridge players are less capable in the way that they cheat?

The only cheats that you are catching are the stupid ones.

Sept. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Could you also label F+N
Sept. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Since you were asking for predictions, I would expect to see

1. A positive correlation between skill as a declarer and skill as a defender

2. F-S scoring significantly higher on defense than on offense
Sept. 13, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Hi Nic

Any chance that you could (easily) generate the following:

First: Examine ONLY those cases in which the same contract was declared at both tables.

Next, for each pair, calculate their IMPS score when they are defending and when they are declaring.

Generate a scatter plot showing where each pair is located in this space.

Last, but not least, label the point that corresponds to Fisher Schwarz. If the data set appears to show a strong linear trend, label a few of the outliers.

Thanks!

Sept. 13, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Out of curiousity, what happens to that information after the game? Does it get sent to the NBO and stored in a readily accessible database? Alternatively, does it all get deleted as soon as folks get their masterpoints?

I, personally, wonder about the accuracy of that information if it is being input by “volunteers”. (By which I mean players at the table who don't understand why this information is so valuable)





Sept. 12, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I think that its worth flagging (in a good sort of way) Mark's second comment.

I am of the opinion statistics based on comparing a pair of values may very well provide better information that statistics based on looking at one value in isolation (hence my interest in comparing scores when declaring with scores when defending).

If we have the luxury to compare and individual's opening leads within a given partnership with opening leads across partnerships this could provide some very useful information.

Sept. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Parity leads / PRISM signals
Sept. 10, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
What Hanan is describing is what is known as a polyalphabetic cipher. People have known how to break these since that days of Babbage. If anyone cares about the details, I recommend consulting Karn about cracking the Vigenère cipher. (None of which detracts from the advantages of breaking things into a test set and a validation set)

The real problem that we need to grapple with is pairs that are using electronics rather than primitive channels like coughing or board placement.

We know that individuals have been caught using electronics to cheat at chess. I don't understand why we belevve that bridge players aren't doing the same.

Sept. 10, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Few quick comments

1. Video does no good if a pair is using electronics to cheat. We KNOW that this is already happening in chess. We have no knowledge about bridge because we haven't thought to look.

2. I agree that old fashioned “cards” might be more enjoyable to spectate for those who have the luxury to travel to Chennai and watch a card game. As for the rest of us, an electronic playing environment is enormously better because it increases the number of tables covered by orders of magnitude.

3. An electronic playing environment makes it easy to collate comprehensive and accurate records. If there's one thing that this scandal has demonstrated, its the value of records.


Sept. 10, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Lycanthrope! Lycanthrope!
Sept. 9, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Silly aside, but when I read that the first thing that came to mind was

“Even a man who is pure in heart and says his prayers by night, may become a wolf when the wolfbane blooms and the autumn moon is bright.”
Sept. 9, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
FWIW, I suggested something similar a couple weeks back.

I wanted to come up with a way to use outlier detection methods to identify pairs that might be engaged in something shady. The best option that I was able to come up with was to group board results into two sets.

The first set was the set of boards that I thought that covert signaling would have the least impact.

The second set was the the set of boards where I thought that covert signaling would have the least impact.

In an idea world, I'd love to be able to look at board results for a few tournaments compare the ratio of results for set A with the results for set B for a decent number of pairs.

I don't expect pairs to be equally skilled at all aspects of the game of bridge. However, if I find an extremely successful pair whose phenomenal results only occur in sub disciplines which might benefit from (hypothetical) covert signaling, this might cause me to take notice.

It seems reasonable to look at declarer play in “par” contracts for set 1. (I understand that bidding can have some impact on declarer play, but you need to start some where)

“Defense” felt like a good choice for the second set.

I'm off to a friend's wedding for a few days, but maybe I'll try to look into this when I get back home next week.



Sept. 9, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Hi Kit

I hate to keep beating a dead horse, however, the process that you are describing is flawed. You describe a process in which

1. You examined the set of hands in the Spingold
2. You formulated a hypothesis based on behavior exhibited in these hands
3. You used the identical set of hands to test your hypothesis

In this case, your examination of the set of hands used in the Spingold lead you to decide that F-S signals involved a defensive shift. You then tested these hands to see whether this is true.

I'm guessing that you are now going to state that I don't understand what you're doing and your methods control for this. I'm sorry. I think that you're wrong. The way in which you are framing your discussion of the CDF of a binomial distribution makes it very clear that you are misrepresenting likelihood estimates.

The results that you are generating are critical for the discussions that are taking place. If you continue to make this type of mistake, it risks discrediting a lot of very good work. I am begging you. In the future, please separate the hands that you are analyzing into a training set and a validation set set.

Develop your hypothesis using the training set. Check your hypothesis with your test set. It will make the statisticians here much more happy.










Sept. 9, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Perhaps a jury is the wrong model…

If I go into a casino and play blackjack too well, the casino is able to ban my ass. They don't need to prove that I was counting cards or using a computer or had an accomplice. They simply need to decide that they don't want me to play blackjack. And that same casino can also share this information with other casinos and they can refuse to let me play as well.


Sept. 8, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Hi Josh

I think that you are overstating the difficulty that rotating the encoding would cause a serious analyst. From my perspective, the critical discovery in this process was the initial recognition that Fisher and Schwartz were varying the position in which they were placing the bidding tray. Once this was picked up, it was only a matter of time before someone cracked the code. (If anyone is interested in this topic, I recommend that you look at techniques for cracking polyalphabetic cipers like a Vigenère cipher)

With this said and done, I think that you're dead on when you identify the problem inherent in insisting that people are able to identity “how” a pair is cheating. As you note, this pretty much guarantees that the only cheats that you catch are the dumb ones.

Indeed, I am reminded of the title of a post of mine ffrom a year and half back "Elinescu-Wladow were stupid. You haven't found the smart cheats…



Sept. 8, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
A more interesting question is whether F-S wouldbe able to launch a future claim regarding the right to be forgotten. That wouldn't apply to a contemporaneous discussion, but could come into play in the future.
Sept. 6, 2015
.

Bottom Home Top