Join Bridge Winners
All comments by Richard Willey
You are ignoring the author of this comment. Click to temporarily show the comment.
I must admit, when I first saw the quote, I was wondering whether it was a Fawlty Towers reference
Sept. 8
You are ignoring the author of this comment. Click to temporarily show the comment.
Would you mind providing the name of the bridge club that you play at?

It will be interesting to see what things look like in 5 years time…
Sept. 8
You are ignoring the author of this comment. Click to temporarily show the comment.
I stopped playing any kind of serious bridge in ACBL Land because of Convention Chart idiocy…
Sept. 7
You are ignoring the author of this comment. Click to temporarily show the comment.
Mike, you clearly don't understand the issue at hand…

It doesn't matter is multi is a great convention or a waste of a bid. Rather, a regulatory environment that bans people from experimenting with a multi or MOSCITO or whatever won't ever be able to attract younger players who want to discover for themselves whether the multi is good or not.

Not you might go around and so “What about all those geriatric ACBL members… Won't this drive them away from the game?” To which I say “Who cares. They're gonna be dead in five years anyway… Or, at the very least too old to continue playing.”

Honestly, its probably too late. The bulk of the ACBL membership is too old and too set in their ways to make the changes that the organization will need to survive. Your own attitude of “I don't care” is a perfect example of why the ACBL is circling the drain.
Sept. 7
You are ignoring the author of this comment. Click to temporarily show the comment.
almost certainly not
Sept. 6
You are ignoring the author of this comment. Click to temporarily show the comment.
> that algorithm can be translated into something that
> can be grasped by anyone who can grasp how to play bridge

Let me guess, you have discovered a truly marvelous proof of this, however, this margin is too narrow to contain it…
Sept. 6
You are ignoring the author of this comment. Click to temporarily show the comment.
FWIW, I have long wondered whether it would be possible to flip a lot of this on its head. (I doubt that this would be practical any time soon, but…)

Imaging a world in which disclosure consisted of providing the opponents with a corpus of hands that are consistent with bid foo. (Perhaps some metadata as well) In theory, this might be providing from either

1. A database of hands that this partnership had bid in the past
2. A dealer script or some such

Each pair would have their own “convention card”, however, rather than showing what they are playing, the convention card translate the set of hands that the opponents provide into something that this pair is comfortable understand. So, if I am a walrus and insist that he only think that I care about is HCPs, then the bi would be described using HCPs.
Sept. 6
You are ignoring the author of this comment. Click to temporarily show the comment.
In most events, your opponent can simple note to the director that the regulations mandate that English is used for disclosure.

I suspect that you are well aware of this fact.
Sept. 6
You are ignoring the author of this comment. Click to temporarily show the comment.
FWIW, when I posted this specific example, I wasn't trying to claim that there is a need to disclose precise percentages or the like.

Rather, there seemed to be some disagreement about whether an entire range should be disclosed OR whether it is better to trim the tails of the distribution. I thought that providing a precise example might help people better frame the various trade offs that might be involved.
Sept. 6
You are ignoring the author of this comment. Click to temporarily show the comment.
I actually play 2-3 ACBL tournaments a year, largely when folks from Akamai are desperate to fill out a team…

I play MOSCITO on BBO on Sundays with Atul versus BrianM
Sept. 5
You are ignoring the author of this comment. Click to temporarily show the comment.
I very much agree that having some mechanism to collect data is critical.

I have my own preferences on how this might be done, however, in some ways I consider this to be secondary to collecting and collating data.

I find it extremely depressing that the power that be don't seem to understand the importance of this sort of issue.
Sept. 5
You are ignoring the author of this comment. Click to temporarily show the comment.
With respect to #2, it would only seem to be of interest either

A. Around the boundaries
B. If you are playing a mixed strategy
Sept. 4
You are ignoring the author of this comment. Click to temporarily show the comment.
I agree that the mechanics seem strange…

I would expect that there would be a single penalty for playing more than one system.

It would be interesting to see precisely how the regulation in question is worded.
Sept. 4
You are ignoring the author of this comment. Click to temporarily show the comment.
I will note in passing that when I described various opening bids in my MOSCITO notes, I very deliberately included sections that show stuff like

1. Here are 10 or so hands that I consider to be bare minimum 1 openings
2. Here are 10 or so hands that I consider to be maximal 1 openings

Ultimately, I thought that this was a more effective way to convey information about range to a dedicated reader than relying on some arbitrary metric.
Sept. 4
Richard Willey edited this comment Sept. 4
You are ignoring the author of this comment. Click to temporarily show the comment.
> How does Richard's algorithm determine which hands
> are better or worse than other hands? As far as I
> am aware no such metric exists.

We are getting far afield here, however, in theory one might do something like the following:

Use some generic method like ZAR Points or BUM RAM or even HCP to identify the left hand tail of the distribution.

For each hand in this tail, generate a 1000 or so different deals, then run them through a double dummy solver and see how many tricks this hand produces when declaring in its best fit.

Identify the 10 worst hands…
Sept. 4
You are ignoring the author of this comment. Click to temporarily show the comment.
I have a dealer script that I can use to generate hands that I deem to be consistent with a MOSCITO opening of type “foo”.

I used said script to generate 100,000 1H openers and then calculated the frequency with which these openings had 7/8/9/…/18 HCPs

I agree that, in a perfect world, it would be better if I either

A. handed folks the script and allowed them to read it
B. handed folks 100K examples of 1H openings and let them draw their own conclusions

However, in the world that I live in, it is mandated that I

1. Use a convention card to describe my openings
2. Use Work HCPs to describe the strength of these openings

So, my question is

Given this set of constraints, what should be done, especially wrt boundary conditions?
Sept. 3
You are ignoring the author of this comment. Click to temporarily show the comment.
I think that the scripts provide an accurate representation of the actual bidding choices that I make.

It should be noted that MOSCITO is a high variance system to begin with. As such, I don't see the need to make “state of the match” decisions
Sept. 3
You are ignoring the author of this comment. Click to temporarily show the comment.
> So on the 8 point or 15 point hands do more than 20%
> of 8 or 15 point hands open the 1♥? I'd guess no, but
> I think this is the more useful question.

I can easily run these numbers once I get home.

Of course, the rub is that the reason that I can generate these numbers is that I have a detailed set of scripts that I use to study stuff like this. Not sure how many people are in the same position. And, regretfully, we rarelyhave good records regarding players behaviour at the table.
Sept. 3
You are ignoring the author of this comment. Click to temporarily show the comment.
>For example, suppose that I open 1♠ with all hands
>with five spades that are in the 11-22 range.
>Analysis shows that I will have a 22-count only 0.5%
>of the time. That's a very small number; should I
>therefore say that the opening is 10-21? Obviously not.

I don't consider your conclusion at all obvious

Based on the results of the pole, yours appears to be very much a minority opinion
Sept. 3
You are ignoring the author of this comment. Click to temporarily show the comment.
Not that i suspect that anyone cares much about MOSCITO style openings, however, the general point made here might be mildly illustrative.

MOSCITO requires 6 slam points for a constructive opening (5 if the hand contains 10 cards in the two longest suits). As such, if we're looking at a 5-4-3-1 pattern or some such

Axxxx
Kxxx
Qxx
x

would qualify as a minimum strength opening

Axxxx
Qxxx
QJx
x

would not

Both hands contain nice HCP, however, the decision to open isn't based on Work HCP
Sept. 3
.

Bottom Home Top