Join Bridge Winners
All comments by Richard Willey
You are ignoring the author of this comment. Click to temporarily show the comment.
> Fred opened 1♦ (Precision, could be a doubleton), and
> my RHO bid 2♣. This was alerted and explained as an
> overcall in either diamonds or hearts.

I have zero sympathy for a complaint that opposing pairs are using conventional defenses over a nebulous Diamond opening. I am quite sure that your own 1D opening is presenting every bit as much confusion to the non-Precision pairs that you are playing against. In my mind, the only difference is that “you” are the one playing the 1D opening and “they” are the ones overcalling.

I will also note that you've had decades to deal with pairs who are using similar types of defenses to your strong club opening. Maybe you were very very lucky and none of your opponents ever figured out that they could do the same over 1D. Now you know. Its probably time to shore up your meta agreements OR if this is too much bother, you always have the option NOT to play an artificial 1D opening.

However, you shouldn't be playing Precision without being willing to pay the piper…
Dec. 15, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> But when I showed up to play at
> the regional event described at
> the beginning, I didn't realize
> that I was going to play poker.
> I thought it was going to be
> duplicate bridge, a game of
> (mostly) skill.

Here's the thing John, despite your 20 years of experience, you don't actually know what “bridge” is…

You have your own weird little ideas but they have nothing to do with the Laws of the game, the regulations used in North America, or the game that I have been familiar with for the past 30 years.

I suggest that you either retreat to the backwater in which you have been playing or learn to accept that there are more things in heaven and earth than are dreamt of in your philosophy…
Dec. 15, 2016
Richard Willey edited this comment Dec. 15, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> In order to do that, they need systems allowed which
> will make the outcome of a board more random and
> luck dependent.

FWIW, I think that there is truth to this claim. I have long argued that the primary goal of convention regulations in the US is to suppress variance and - in doing so - reward the pros.

Here's where it gets amusing: I seem to recall Kit Woolsey writing some interesting stuff on whether or not to play for the drop or a finesse when missing five to the Queen. He claimed that a weaker team should consider playing for the drop and deliberately take a suboptimal line of play because the increase in variance was worth more to them than the decrease in the expected value…

I guess that that should only be legal when it comes to card play…
Dec. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> Bo-Yin Yang's analysis of its use in European and world
> championships (before it was prohibited) showed it gained
> more than 2 IMPs on average every time it was used.

There is an important caveat to this: The major wins for Wilkosz were occurring on hands where a Wilkosz 2D was opened at one table and 1M was opened at the other.

The pairs who were not playing Wilkosz were - in theory - playing relatively sound openings but could not bear to pass with an offensively oriented 5-5 hand and a sub-minimum strength. Later on, the wheels came off in their auction.

I personally believe that - as the world at large has gained more experience with opening 10 counts - that the expected gain for the Wilkosz 2D would decrease significantly.
Dec. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Fifteen or so years back, I was wanted to be able to play MOSCITO here in the US and worked very hard to get suggested defenses approved to the following opening scheme

1D = 4+ Hearts, might have a longer minor (~ 9 - 14 HCP)
1H = 4+ Spades, might have a longer minor (~ 9 - 14 HCP)
1S = unbalanced with 4+ Diamonds

The ACBL Conventions Committee ruled that this was far too complicated to ever allow at the Midchart level, refused to sanction any defense to this method, and does so to this day.

Of course, a couple years later, when Martel himself wanted to play “transfer Walsh” over his short club opening, that was approved immediately and was considered so innocuous that it wasn't necessary to provide any defense to the opponents.

And, of course, shortly there after, the ACBL pushed through regulations that short clubs are natural as to ban convention defenses to the 1C opening.

The only rhyme or reason to the convention regulations in the US is that methods that the US pros want to play will be legal and methods that they don't want to play against will not. Pretending that there is anything impacting this rather than influence and power is disingenuous.

FWIW, I found the whole blowup between the US team and the Spanish team this year extremely interesting. I don't think that the Spanish had nearly enough evidence to prove that Lall and Bathurst were playing a HUM and they way in which the Spanish team behaved did them no credit.

However, I suspect that someone a bit more competent might have been able to make something stick against any number of Precision pairs…
Dec. 14, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
If people wanted to play cards, nothing would seem to stop the club owners from running an unsanctioned game…

I suspect that the real problem is that a sufficient mass of players was playing in the Nationals that the club director decided not to both (and who knows, perhaps the directors wanted to play or were working)
Dec. 13, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> I did not cheat! I disclosed ACBL & Common Game hand
> record security issues.

Hi Art,

I'd be interested in understanding the time line here.

1. Can you provide a more detailed description of your exploit? (From the sounds of things, you were able to access an ACBL server that contained both hand records and personal information)

2. What information did you provide to the ACBL and how did you submit it?

3. What was the ACBL's response?

4. What were the next step of actions that you took? What type of disclosure did you provide?
Dec. 10, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I suspect that i would start by cashing the Ace of Hearts, exclude Spades and then exit with a Heart
Dec. 7, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Its unclear to me that they didn't

Please note: I have about as low an opinion about the ACBL as anyone on this list. In this case, I think that its unclear whether they reached the wrong decision.

1. I think that the primary consideration needs to be what set of policies get put in place moving forward.

2. While I would like to see harsher punishments for what happened in the past, I'm not sure whether I would prioritize “punishment” ahead of “avoid the possibility of costly litigation”'

From what I can tell, the big open question is how do you handle individuals who want to forfeit titles. And, as an interesting extension, if you chose to forfeit a title, does this mean that you are also relinquishing seeding points and master points. (Arguably, the two should not be separable)
Dec. 7, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
I can point the folks at work at this.

The might find the problem interesting enough to play with…
Dec. 7, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> If the order of boards was randomly permuted
> after the set was generated, the cracker would
> take a lot longer to work.

Why do you believe that randomly permuting the board ordering is more challenging for the cracker than increasing the set of hands that the dealer is capable of generating?

Once again, it you are adding additional entropy, this makes life harder

If you are re-distributing entropy from hand generation to hand permutation, this shouldn't matter with a decent implementation.
Dec. 7, 2016
Richard Willey edited this comment Dec. 7, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> Someone had a good idea the other day: having
> dealt 24 or so deals you should randomize the
> order in which they appear in the hand records.
> Would this be of any use?

Entropy is entropy. It really doesn't matter if you are using that entropy to do a better job generating hands or permuting the order of the ones that you are generating. Either way, it pretty much is going to come out the same.

And, for better or worse, this is pretty much moot because the flawed Dealer program that the ACBL was using has already been swapped out for major events and is in the process of being swapped out for all events.

At this point in time, the work that Nic is doing is (hopefully) just a historical curiosity to help characterize how bad it might have been…
Dec. 6, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Nice job! (Sorry we didn't have longer to talk in Orlando)
Dec. 6, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> Please explain (how 2♦ = ♠ and a red suit is GCC-legal).

Under Competitive section, the following is explicitly sanctioned

7. DEFENSE TO:
a) conventional calls (except see #10 RESPONSES and REBIDS
above and #7 under DISALLOWED below)

While the ACBL treats one specific type of short club opening as “natural”, not all Short Club openings are natural. Since this opening is defined as a short club rather than the specific natural case, the overcall is a defense to a conventional opening and thereby allowed.
Dec. 6, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> Perhaps it is an optical illusion or hallucination on my part.
> That's possible. But I tend to be more objective than most.
> Just sayin'!!

You're welcome to audit the source code.

You'll need to find a period to the function used to assign random numbers and then show that this period is also present in the PRNG than is being used to generate the deals…

(Before you go and do so, I suggest that you look at any number of studies about how good people are in projecting patterns onto random data)
Dec. 5, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Yes, thanks. Cut and paste issue. I'll correct
Dec. 5, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Here's a quote from the Big Deal documentation. It (pretty much) precludes the type of issue that you are suggesting… You'd need some kind of weird synchronization between the output of the PRNG and the second function that is being referenced. Even if this were to happen for one part of the sample space, this shouldn't have any bearing on the rest of the sample space unless you're using a space filling design or some such rather than a decent source of entropy.

“The number of possible bridge deals is just a bit under 2^96, as shown in the reference section. This means that it should be possible, at least in theory, to design a pair of functions, one of which can convert any bridge deal into a 96-bit number, and the other can convert that 96-bit number back into the same bridge deal. Jeroen Kuipers has indeed designed and implemented this pair of functions, and he proved the correctness of his implementation in a document provided with the software. We use the second function of the pair in the program.

So the algorithm to convert the stream of 160-bit pseudo-random numbers into bridge deals runs as follows:

First reduce the 160-bit number to a 96-bit number by just throwing away 64 bits. Since the original 160-bit number is pseudo-random this 96-bit number will also be pseudo-random.
The number of bridge deals is somewhat less than 2^96, so if this 96-bit number is too big throw away the whole number and go back to step 1 with the next 160-bit pseudo-random number.
Convert the 96-bit number into a bridge deal. Given the pseudo-randomness of the original numbers this should be able to generate each possible bridge deal, with equal probability.
This in effect finishes the innovative parts of the program.”

I suspect that what you are experiencing is a variant of the birthday paradox.
Dec. 5, 2016
Richard Willey edited this comment Dec. 5, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Folks have looked into the impact of taking a deck of cards that has been played and - thereby - sorted and subjecting this to an incomplete shuffle.

The resulting deal is biased towards being artificially flat
Dec. 4, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
> True, Richard - but all I've been trying to say all along is
> that they aren't always synonymous either.

Then why are you choosing a construct that places them in opposition?
Dec. 4, 2016
You are ignoring the author of this comment. Click to temporarily show the comment.
Good pun (if there is such a thing)

With this said and done, I wasn't suggesting that. Rather, I was implying that legal is not the opposite of “unsportsmanlike”.
Dec. 4, 2016
.

Bottom Home Top