Discussion:
Rock, Paper, Scissors
(too old to reply)
P***@yahoo.com
2016-09-16 07:55:39 UTC
Permalink
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.

I ran across an interesting case of a hand, where the "optimum contract" is problematic. Envision perfect partnership communication in a hand that results in the following probabilities of making eleven or twelve tricks (depending upon what opponents have):

Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.

Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.

With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.

Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
Co Wiersma
2016-09-16 08:41:32 UTC
Permalink
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
I do not get you
Why is it optimum to play 6NT if the field is in a contract with better
chance?
If all are in six of a suit, then bidding six of a suit will get us 50%
all the time while bidding 6NT will get us a zero 55% of the time which
makes up for an average of 45%?

Co Wiersma
P***@yahoo.com
2016-09-16 09:50:29 UTC
Permalink
Post by Co Wiersma
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
I do not get you
Why is it optimum to play 6NT if the field is in a contract with better
chance?
If all are in six of a suit, then bidding six of a suit will get us 50%
all the time while bidding 6NT will get us a zero 55% of the time which
makes up for an average of 45%?
Co Wiersma
If everyone else plays in six of a suit,

6NT will get us:

15% of the time => zero matchpoints
40% of the time => 50% matchpoints (everyone will go down one)
45% of the time => 100% matchpoints
Bruce Evans
2016-09-16 10:31:13 UTC
Permalink
Post by P***@yahoo.com
Post by Co Wiersma
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding
problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Post by Co Wiersma
Post by P***@yahoo.com
I ran across an interesting case of a hand, where the "optimum
contract" is problematic. Envision perfect partnership communication in
a hand that results in the following probabilities of making eleven or
Post by Co Wiersma
Post by P***@yahoo.com
Six of Suit Contract = 60% probability of making twelve tricks / 40%
probability of making eleven tricks.
Post by Co Wiersma
Post by P***@yahoo.com
Six Notrump = 45% chance of making twelve tricks / 55% probability
of making eleven tricks.
Post by Co Wiersma
Post by P***@yahoo.com
With those probabilities and at matchpoints, slam in a suit contract
is optimal if the field plays below slam level. Slam in notrump is
optimal if the field plays at the slam level. And a 3N contract is
optimal if the field plays in a slam at notrump.
Post by Co Wiersma
Post by P***@yahoo.com
Rock, paper, scissors! Perhaps an example of why IMP scoring is so
much more logical?
Post by Co Wiersma
I do not get you
Why is it optimum to play 6NT if the field is in a contract with better
chance?
If all are in six of a suit, then bidding six of a suit will get us 50%
all the time while bidding 6NT will get us a zero 55% of the time which
makes up for an average of 45%?
If everyone else plays in six of a suit,
15% of the time => zero matchpoints
40% of the time => 50% matchpoints (everyone will go down one)
45% of the time => 100% matchpoints
6NTX will get you 0% matchpoints 40% more of the time (everyone will go
down 1).

So bidding 6NT is only right if the field plays at the same level AND
your opponents don't double enough. They will rarely have enough
authorized information to double, but might be trigger-happy. This
example shows that they should be trigger-happy if the field plays
in slam.

Similarly for games vs partscores, except slams are easier to double.
It is right to bid 3NT with a 40+ to 50- percent chance of making in
preference to a 50+ to 60- percent 4M, if the field plays in game,
but only if the opponents don't double.

Another consideration is that in the slam case, you need a long auction
that gets to at least 5NT in order to full investigate the slam chances.
This might find the par 6S or the matchpoint lottery 6N. But to be that
accurate, it must often disclose the information needed for the opponents
to double 6NT without guessing or auto-triggering.

Bruce
Barry Margolin
2016-09-16 19:30:03 UTC
Permalink
Post by P***@yahoo.com
If everyone else plays in six of a suit,
15% of the time => zero matchpoints
40% of the time => 50% matchpoints (everyone will go down one)
45% of the time => 100% matchpoints
Doesn't this assume that there's no correlation between the suit and NT
success probabilities?

E.g. if both contracts require a particular finesse, then you can ignore
that trick when comparing the contracts.
--
Barry Margolin
Arlington, MA
jogs
2016-09-16 21:03:44 UTC
Permalink
Post by P***@yahoo.com
If everyone else plays in six of a suit,
15% of the time => zero matchpoints
40% of the time => 50% matchpoints (everyone will go down one)
45% of the time => 100% matchpoints
Study the hand record histories. It is presumptuous of any of us to assume we have any clue what the field is doing. The field is random.
Lorne Anderson
2016-09-16 11:34:09 UTC
Permalink
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
Judging what the field will do is critical in MP's. My general rule is
that if you are bidding a slam on less than about 31 points go for the
safest one because many will stop in game but if you have 32+ usually go
for the highest scoring option as most others will do the same.

If the field is high quality however then adjust the way you work out
what the field will do for weaker hand with good shape an a fit, and
never take any risk in a low quality field. I once bid a 24 point grand
in a weak field, it was 90% to make and did make but was a terrible
contract as nobody else bid even the small slam and several people were
in a part score. The near 100% small slam which was making even if
trumps were 5-0 was the best contact in that field.
P***@yahoo.com
2016-09-16 12:41:06 UTC
Permalink
Post by Lorne Anderson
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
Judging what the field will do is critical in MP's. My general rule is
that if you are bidding a slam on less than about 31 points go for the
safest one because many will stop in game but if you have 32+ usually go
for the highest scoring option as most others will do the same.
If the field is high quality however then adjust the way you work out
what the field will do for weaker hand with good shape an a fit, and
never take any risk in a low quality field. I once bid a 24 point grand
in a weak field, it was 90% to make and did make but was a terrible
contract as nobody else bid even the small slam and several people were
in a part score. The near 100% small slam which was making even if
trumps were 5-0 was the best contact in that field.
Excellent advice! I love it. Thanks, Lorne.
jogs
2016-09-16 14:21:11 UTC
Permalink
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
During the auction we are in no position to estimate the odds of any contract with that degree of accuracy.
p***@infi.net
2016-09-16 20:07:27 UTC
Permalink
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
But the field is not likely to be all in one contract on such a hand. I would set up a spreadsheet with percentages of the field in the various contracts, compute matchpoints, and see what happens as the percentages shift.
Charles Brenner
2016-09-16 20:53:55 UTC
Permalink
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
My Bridge World article "Rock paper scissors" (neither I nor editor Rubens noticed that Marshall Miles had previously used the title) of a several years ago illustrated exactly this theme: A cycle of three contracts each superior to one of the others. That was at IMP scoring, so no, this phenomenon is not limited to MP and therefore does is not an argument that IMPs is more logical.

It's not even limited to bidding. I also gave in the article a cycle of dummy play strategies for a certain 3NT contract.

Two side points -- I meant the article to be illustrated by an example I'd invented some decades previous, which I couldn't find but tried to reconstruct. Jeff wasn't satisfied with my construction so he came up with a nice simple one of his own which we used. As for the title, I noticed later that Marshall Miles had earlier published a BW article with the same title.
Charles Brenner
2016-09-16 20:58:21 UTC
Permalink
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
My Bridge World article "Rock paper scissors" of a several years ago illustrated exactly this theme: A cycle of three contracts each superior to one of the others. That was at IMP scoring, so no, this phenomenon is not limited to MP and therefore does is not an argument that IMPs is more logical.

It's not even limited to bidding. I also presented a cycle of dummy play strategies for a certain 3NT contract.

Two side points -- I meant the bidding cycle example to be one I invented some decades previous, which I couldn't find but tried to reconstruct. Jeff wasn't satisfied with my construction so he came up with a nice simple one of his own which we used. As for the title, apparently we both overlooked that Marshall Miles had earlier published a BW article with the same title.
P***@yahoo.com
2016-09-17 00:01:20 UTC
Permalink
Post by Charles Brenner
Post by P***@yahoo.com
As I've been working on a thorough scientific analysis of a bidding problem, I'm cataloguing hands into "optimum contracts" at matchpoints.
Six of Suit Contract = 60% probability of making twelve tricks / 40% probability of making eleven tricks.
Six Notrump = 45% chance of making twelve tricks / 55% probability of making eleven tricks.
With those probabilities and at matchpoints, slam in a suit contract is optimal if the field plays below slam level. Slam in notrump is optimal if the field plays at the slam level. And a 3N contract is optimal if the field plays in a slam at notrump.
Rock, paper, scissors! Perhaps an example of why IMP scoring is so much more logical?
My Bridge World article "Rock paper scissors" of a several years ago illustrated exactly this theme: A cycle of three contracts each superior to one of the others. That was at IMP scoring, so no, this phenomenon is not limited to MP and therefore does is not an argument that IMPs is more logical.
It's not even limited to bidding. I also presented a cycle of dummy play strategies for a certain 3NT contract.
Two side points -- I meant the bidding cycle example to be one I invented some decades previous, which I couldn't find but tried to reconstruct. Jeff wasn't satisfied with my construction so he came up with a nice simple one of his own which we used. As for the title, apparently we both overlooked that Marshall Miles had earlier published a BW article with the same title.
Fascinating -- I hadn't heard of your article, but would be interested in it.
Steve Willner
2016-10-11 23:33:58 UTC
Permalink
Post by Charles Brenner
My Bridge World article "Rock paper scissors" of a several years ago
2010 May according to the index at
http://www.bridgeworld.com/indexphp.php?page=/pages/backissues/indexbyauthor.html
Post by Charles Brenner
illustrated exactly this theme: A cycle of three contracts each
superior to one of the others.
...
Post by Charles Brenner
As for the title, apparently we both overlooked that Marshall Miles
had earlier published a BW article with the same title.
Close but maybe not quite the same. I couldn't find the original, but
"Rock-Scissors-Paper" was published as a "Classic Rewind" in 1999 Sep.
Charles Brenner
2016-10-15 11:23:24 UTC
Permalink
Post by Steve Willner
Post by Charles Brenner
My Bridge World article "Rock paper scissors" of a several years ago
2010 May according to the index at
http://www.bridgeworld.com/indexphp.php?page=/pages/backissues/indexbyauthor.html
Post by Charles Brenner
illustrated exactly this theme: A cycle of three contracts each
superior to one of the others.
...
Post by Charles Brenner
As for the title, apparently we both overlooked that Marshall Miles
had earlier published a BW article with the same title.
Close but maybe not quite the same. I couldn't find the original, but
"Rock-Scissors-Paper" was published as a "Classic Rewind" in 1999 Sep.
Thanks; it's reassuring that I didn't copy Marshall's title. Also good to note that (punctuation aside) we have 4 permutations available for future work.
Continue reading on narkive:
Loading...