Join Bridge Winners
All comments by Maaijke Mevius
You are ignoring the author of this comment. Click to temporarily show the comment.
I believe the investigation committee did a similar but independent study, I have no information about the details.
Feb. 17
You are ignoring the author of this comment. Click to temporarily show the comment.
@Oleg. The real question is: The experts had NO knowledge how the bidding cards were placed and did not fully understand the details of the Polish club system. Yet they were able to predict the gap between the bidding cards for a large fraction of the bids, ONLY by looking at the bid sequence and the cards. How is that possible?

If the people that evaluated the bids did not understand anything about the bidding, you would expect that they got it right only about 50% of the time. The same is true if there was no correlation between the hand strength and bidding cards placement.

To me the answer to this question is that either the experts did have knowledge about the the gap between the bid cards and were possibly biased or a bias was introduced in the selection of the hands. And then the third possibility is of course that there really was a correlation between the strength of the hand and the position of the bidding cards… The only point the scientists here are trying to make is that the exclusion of the 3 experts solely based on their lack of knowledge of the Polish club system as it was stated in the verdict, was plain wrong.
Feb. 17
You are ignoring the author of this comment. Click to temporarily show the comment.
I assume that the experts did not have knowledge about the placement of the bidding cards when evaluating the bids (this is in contrast to Jassem when he did his evaluation of a couple of boards), otherwise their evaluation was indeed probably biased.

If the hand evaluation was indeed done independently and without prior knowledge, the fact that the experts did not fully understand Polish club would only have weakened the correlation by introducing more noise. Including their testimonies could only have been in favour of the defendants as far as statistics is concerned…
Feb. 14
You are ignoring the author of this comment. Click to temporarily show the comment.
Appendix C lists only the 6 groups plus selected bids, no videos. In group 6 it says:

Russia
26. 11 z 2 Monaco
27. 3 z 2


This is the missing carriage return I was talking about. As far as I checked the listed videos match the selected bids. I agree there are only two bids from the Russia match (another one in group 3), which is odd indeed.
Nov. 9, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Kit, in this case the only “transcription error” was a missing carriage return after board 26 of group 6, which confused Nicolas.
Nov. 9, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I did not measure anything, I just used the table on page 5. I believe the measurements come from Rui Marques. I had the same impression as you on the wide bids, but there is not enough data here to confirm this.
Nov. 3, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Yes, it is natural to place your bidding cards wider on other bids,such that the previous bid is still visible. I only wanted to state that it is not straightforward to come with a single distribution for narrow and wide gaps.
Nov. 3, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I believe there is a difference between cards placed on pass/x cards and cards placed on other bidding cards. Also I believe there is a difference for B and Z (hence the 8 examples provided to the observers). Given the statistics only a hint can be found from the given numbers for narrow bids, but this is what I calculated from the table (1 sigma errors, assuming normal distribution which is disputable):

Z: narrow on pass/x: 0.91 +- 0.04 inch
narrow on other: 1.02 +- 0.06 inch
B: narrow on pass/x: 1.00 +- 0.06 inch
narrow on other: 1.14 +- 0.05 inch

Maybe there is data available for all hands.
Nov. 3, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I agree. You need a bit more statistics on trigger-signal. Nicolas, I assume you have also looked at the hands where Z is not on lead, or where there was some bidding by BZ?
Oct. 23, 2015
Maaijke Mevius edited this comment Oct. 23, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Make sure you have a good supply of popcorn if you decide to watch the Israel-Germany match. It is a great show.
Oct. 20, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Ok, thanks for clarifying that.
Oct. 16, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Sorry, I do not really understand this discussion. First of all, why would, as Boye stated somewhere above, in the FN case, Ish have to repeatedly inform Kit that the code was cracked? I originally sent the message about FN not only to Boye, but, in order to increase the probability that someone would actually read it, also to Kit. Since he replied within minutes that he would check it, and since the pattern was well explained and not that difficult to spot, it being a close to 100% match, I cannot imagine this is correct.

But also, who cares about credits? I have absolutely no doubt that hadn't it be me, someone else would have seen a pattern in FN leads, and maybe some did. I personally regret that I allowed Boye to publish my name, I hadn't anticipated the media attention it would trigger. Credits should go to the person who initiated the crowd sourcing, and maybe to all who spend hours collecting material and watching videos, whether or not they “found something”.

I do understand that, being seen by many (not me) as an “unbiased authority”, Kit's repeated criticism of the bidding gap theory doesn't help the BZ case, so I guess this is what this is all about…
Oct. 16, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
What about: KQJTx - KT9xxx Tx ?
Oct. 9, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Interesting. If you correct those two boards, remove the wrong board 31 and add board 22 as a +2 (I can prove that the passes are spaced far apart, but please correct me if this is not a maximum pass), you get a correlation coefficient of 0.89. Still doesn't mean much given the limited statistics.
Oct. 8, 2015
Maaijke Mevius edited this comment Oct. 8, 2015
.

Bottom Home Top