After opener's 2NT

[paul...@infi.net]

Club matchpoint pairs, South deals, N/S vulnerable
As East you have:

Q42
K32
AK10842
J

With N/S silent, your auction proceeds:
1C-1D
2NT-?

Your methods are fairly standard, a 1NT opening would have been 15-17 so you can expect 18-19 or an upgraded 17 form partner, and presumably no four-card major. Three of any suit by responder would be natural and forcing, no special gadgets here.

Your call?

[Player]

3D. Gee that was hard.

[David Goldfarb]

In article <a1df46d2-7609-4e56...@googlegroups.com>,

I bid 3D, and if partner bids 3NT I'll bid 4D. I'm hoping to get to 6D,
and I'm willing to get to 4NT if necessary.

--
David Goldfarb |"I weep for the death of the spirit and soul."
goldf...@gmail.com | "Hey, who doesn't? We'll be right back."
gold...@ocf.berkeley.edu | -- Mystery Science Theatre 3000

[Parson...@yahoo.com]

I bid 4C, which is RKC1430 for diamonds in my methods (according to Eddie Kantar's "Roman Keycard Blackwood, The Final Word"). This way, I learn about the A's and K's on the way to bidding either 5N or 6N. In my estimation, small slam in diamonds or NT has about the same high probability and since this is matchpoints, I'm favoring NT especially with the strong hand playing it.

[Lorne]

I would just bid 6N and hope that if we have 2+ losers and could have
found out by going slowly we gain more from making when the lack of
bidding makes it hard to find the killing lead than we lose when it goes
off and we could have got to a making spot.

IMP scoring I think it is worth exploring in case 7D is the spot and I
am less concerned at playing in 6D when 6N makes so I start with 3D.

[france...@googlemail.com]

4C, autosplinter for diamonds. But only if partner doesn't think it is Gerber.
Don't you want to get to 7D opposite AKx Ax QJx Axxxx but 6NT opposite AKx AQx Qxx K109xx ?

[paul...@infi.net]

On Wednesday, November 25, 2015 at 4:54:47 PM UTC-5, paul...@infi.net wrote:

For those suggesting specialized uses for 4C, interesting treatments. However, in all of my partnerships 4C would be Gerber, which we define as Jump Over Notrump Only (any jump to 4C over a 1NT or 2NT by partner, excluding Unusual Notrump.) Consistent with that 4NT directly over 2NT would be invitational, not any form of Blackwood. I'm sure there are better agreements but everyone plays with multiple partners so it's important to keep things simple.

[Barry Margolin]

In article <f70d8565-a435-4b32...@googlegroups.com>,

Parson...@yahoo.com wrote:

> On Wednesday, November 25, 2015 at 4:54:47 PM UTC-5, paul...@infi.net wrote:
> > Club matchpoint pairs, South deals, N/S vulnerable
> > As East you have:
> >
> > Q42
> > K32
> > AK10842
> > J
> >
> > With N/S silent, your auction proceeds:
> > 1C-1D
> > 2NT-?
> >
> > Your methods are fairly standard, a 1NT opening would have been 15-17 so
> > you can expect 18-19 or an upgraded 17 form partner, and presumably no
> > four-card major. Three of any suit by responder would be natural and
> > forcing, no special gadgets here.
> >
> > Your call?
>
> I bid 4C, which is RKC1430 for diamonds in my methods

The question says "your methods are fairly standard". Do you really
think that special agreement fits?

--
Barry Margolin
Arlington, MA

[paul...@infi.net]

I don't object to anyone posting their specialized methods, those are always food for thought. But of course the poster might also say what he would do lacking such agreements.

[Kenny McCormack]

In article <barmar-C68D89....@88-209-239-213.giganet.hu>,
Barry Margolin <bar...@alum.mit.edu> wrote:
...

>> I bid 4C, which is RKC1430 for diamonds in my methods
>
>The question says "your methods are fairly standard". Do you really
>think that special agreement fits?

Does it really matter?

Note that normally I would agree with the general thrust of your post, but
in this case, really, saying "I bid 4C which is RKC" is just a shorthand
for "You should bid whatever bid in your system is ace asking".

P.S. Myself, I would just bid 6D and hope for the best. I'm not worried
about "missing" 6N - I'll still get a good score for 6D (if it makes).

--
One of the best lines I've heard lately:

Obama could cure cancer tomorrow, and the Republicans would be
complaining that he had ruined the pharmaceutical business.

(Heard on Stephanie Miller = but the sad thing is that there is an awful lot
of direct truth in it. We've constructed an economy in which eliminating
cancer would be a horrible disaster. There are many other such examples.)

[Barry Margolin]

In article <n378ps$vc5$1...@news.xmission.com>,

gaz...@shell.xmission.com (Kenny McCormack) wrote:

> In article <barmar-C68D89....@88-209-239-213.giganet.hu>,
> Barry Margolin <bar...@alum.mit.edu> wrote:
> ...
> >> I bid 4C, which is RKC1430 for diamonds in my methods
> >
> >The question says "your methods are fairly standard". Do you really
> >think that special agreement fits?
>
> Does it really matter?
>
> Note that normally I would agree with the general thrust of your post, but
> in this case, really, saying "I bid 4C which is RKC" is just a shorthand
> for "You should bid whatever bid in your system is ace asking".

But in this case, the ace-asking bid has a special space-saving feature
that most other ace-asking conventions don't. That makes it very useful
when the suit is a minor, and he might not have the same advice when
using "standard" methods.

[Kenny McCormack]

In article <barmar-C495CF....@88-209-239-213.giganet.hu>,

Barry Margolin <bar...@alum.mit.edu> wrote:
>In article <n378ps$vc5$1...@news.xmission.com>,
> gaz...@shell.xmission.com (Kenny McCormack) wrote:
>
>> In article <barmar-C68D89....@88-209-239-213.giganet.hu>,
>> Barry Margolin <bar...@alum.mit.edu> wrote:
>> ...
>> >> I bid 4C, which is RKC1430 for diamonds in my methods
>> >
>> >The question says "your methods are fairly standard". Do you really
>> >think that special agreement fits?
>>
>> Does it really matter?
>>
>> Note that normally I would agree with the general thrust of your post, but
>> in this case, really, saying "I bid 4C which is RKC" is just a shorthand
>> for "You should bid whatever bid in your system is ace asking".
>
>But in this case, the ace-asking bid has a special space-saving feature
>that most other ace-asking conventions don't. That makes it very useful
>when the suit is a minor, and he might not have the same advice when
>using "standard" methods.

Isn't 4C ace-asking here in standard methods?

(Assuming "standard methods" includes Gerber... Which it certainy did in
the days when I was learning "standard methods")

--
Mike Huckabee has yet to consciously uncouple from Josh Duggar.

[Parson...@yahoo.com]

On Thursday, November 26, 2015 at 9:58:50 AM UTC-5, paul...@infi.net wrote:
> On Wednesday, November 25, 2015 at 4:54:47 PM UTC-5, paul...@infi.net wrote:
> > Club matchpoint pairs, South deals, N/S vulnerable
> > As East you have:
> >
> > Q42
> > K32
> > AK10842
> > J
> >
> > With N/S silent, your auction proceeds:
> > 1C-1D
> > 2NT-?
> >
> > Your methods are fairly standard, a 1NT opening would have been 15-17 so you can expect 18-19 or an upgraded 17 form partner, and presumably no four-card major. Three of any suit by responder would be natural and forcing, no special gadgets here.
> >
> > Your call?

Yes, 4C is Gerber and it is RKC Gerber. This is fairly standard at my club for those playing RKC (and who doesn't?), per Eddie Kantar's definitive book published eons ago. It's just as standard as the auction after a Jacoby Transfer bid:

1N P 2D P
2H P 4C

which is a RKC ask for hearts. You don't play this way in the UK?

[Player]

Standard beginner includes Gerber. Standard better player + does not. Good players don't need baby food.

[Player]

This is definitely not expert standard. This shows an autosplinter for most experts; certainly not Gerber.

[Parson...@yahoo.com]

Interesting -- I had never heard of using 4C as an autosplinter before. Thanks for bringing it to my attention.

[france...@googlemail.com]

I suggested 4C on the basis that was standard, at least for certain categories of players. If I were playing for the first time with an expert in the UK, I would be fairly comfortable bidding 4C and assuming it was a splinter.

Lacking any such agreements I would bid 4D.

[paul...@infi.net]

On Wednesday, November 25, 2015 at 4:54:47 PM UTC-5, paul...@infi.net wrote:

OK, we had two votes for 3D, one each for 6D and 6NT, one for 4C as RKCG and another for 4C as auto-splinter. Bridge World Standard defines 4C as Gerber, and I have that agreement with all my partners; I would be inclined to bid that rather than blasting into slam possibly off two Aces, but of course you might steal a slam if they misdefend. Gerber isn't as useful as RKCG here since the diamond Queen is a vital card.

Let's suppose you settle on 3D, forcing. Opener comes back with 3S. Now what?

[Player]

4h. 3s was a cue bid of course 😊, wasn't it?

[france...@googlemail.com]

You missed my answer which was a vote for 4D, natural slam try, if 4C was Gerber. 3D is murky

[paul...@infi.net]

Didn't miss it, just wanted to explore the continuations over 3D first, which might well illustrate the wisdom of bidding something else. 4C autosplinter looks ideal on this hand if you can get back to 4NT. Over the autosplinter, would 4NT by opener be to play? If opener bids anything else, would 4NT by either player now be RKCB?

With 4C not autosplinter, how do we continue over 4D? Would 4NT be to play, 4H or 4S control bids, and after a control bid 4NT is RKCB?

[jogs]

This is the type of board which is well suited for DD analysis.
One side holds most of the power. The number of possible hand types is manageable.
Column 1 is the trial number.
Column 2 is tricks made in 3NT.
Column 3 is tricks made in 5NT or 6NT. 5NT when missing 2 aces else 6NT.
Column 4 is the difference of scores between col 3 and col 4.
When in 5NT, 3NT wins or ties.
When in 6NT, 3NT wins when 6NT fails.

[Parson...@yahoo.com]

Pretty easy to run with the simulator. Out of 10,000 hands,

Number of times in 5N, missing 2 aces = 154

Col1 Col2 Col3 Col4
10000 9831 8370 1443

Also of interest is 6D, where Col3 = 8574 and Col4 = 1257.

[jogs]

Shocking!! Missing 2 aces only 1.54% of the time.

Maybe I didn't explain it correctly.
Of the remaining 10,000 - 154
a) How often did 6NT make?
b) How often did 6NT fail?

If a) greater than b).

If so, then it is right to Blackwood.
If not, rebid 3NT is best.

[Parson...@yahoo.com]

a) is about 8220
b) is about 1630

[jogs]

That pretty well clarifies that one can ask for aces on the way to 6NT.

Now next question. Missing no aces. How often does NT make 13 tricks?
How often does NT make 12 or fewer tricks?

And another question. Missing no aces. Does diamonds make 13 tricks appreciable more often than NT?

[Parson...@yahoo.com]

Missing no aces, NT makes 13 tricks 48.7% of the time and makes 12 tricks 39.3% of the time.

Missing no aces, Diamonds makes 13 tricks 55.8% of the time and makes 12 tricks 38.3% of the time.

[jogs]

> >
> > That pretty well clarifies that one can ask for aces on the way to 6NT.
> >
> > Now next question. Missing no aces. How often does NT make 13 tricks?
> > How often does NT make 12 or fewer tricks?
> >
> > And another question. Missing no aces. Does diamonds make 13 tricks appreciable more often than NT?
>
> Missing no aces, NT makes 13 tricks 48.7% of the time and makes 12 tricks 39.3% of the time.
>
> Missing no aces, Diamonds makes 13 tricks 55.8% of the time and makes 12 tricks 38.3% of the time.

Q42
K32
AK10842
J

This diamond suit demonstrates the difference between DD analysis and 'real bridge'. Let's give opener 953 of diamonds.

AKT842

953

DD analysis is deterministic/perfect information. Each side knows the location of all 52 cards. In game theory each side uses optimal strategy. They will find the expected value(EV) outcome. Only one outcome for each board setup. A declarer using DD analysis only fails to pick up diamonds with no losers when diamonds break 1-3 or 0-4 offsides.
'Real bridge' is non-deterministic/imperfect information. The declarer knows the location of his 13 cards and the dummy's 13 cards. Each defender knows the location of his own 13 cards and dummy's 13 cards. Many players will not be using optimal strategy. Multiple outcomes while playing in the same strain is not uncommon.
A 'real bridge' declarer will play the DA on the first play of diamonds. If diamonds break QJxx -- _ or QJx -- x DD declarers will be able to avoid diamond losers. Real bridge declarers will not.

J76 // Q
Q76 // J
76 // QJ

DD declarers will pickup these combinations for no losers. Real declarers, playing restricted choice, will lose a trick to 76 // QJ.

[Parson...@yahoo.com]

Yes, this is all rather obvious. But it does not invalidate conclusions that are guided by double dummy analysis. Published studies on this include Pavlicek's "Study 9X29: Actual Play Vs. Double-Dummy" at

http://www.rpbridge.net/9x29.htm

where he shows (in Lawrence Diamond's words):

<quoting>
Pavlicek has shown (using Vugraph data from 1996-2013) that at the part-score level, actual play took 8.08 tricks per hand versus 7.96 double dummy. Only at slam level, was the average double-dummy result better than actual declarer play (11.78 tricks vs 11.72). Pavlicek's results were confirmed using the database of hands for this (Lawrence Diamond's) book. Thus it seems reasonable to accept the premise that double-dummy data is satisfactory for making such decisions if the averages are collected from a sufficient number of hands.
<end quote>

That's good enough for me, but I'm afraid will never be good enough for you for some reason, Jogs.

[jogs]

That is not the proper way to compare results. One should compare first differences. For each separate trial the difference between DD and actual play.
Average tricks per can be equal when every trial produced different results.

> Only at slam level, was the average double-dummy result better than actual declarer play (11.78 tricks vs 11.72). Pavlicek's results were confirmed using the database of hands for this (Lawrence Diamond's) book. Thus it seems reasonable to accept the premise that double-dummy data is satisfactory for making such decisions if the averages are collected from a sufficient number of hands.
> <end quote>
>
> That's good enough for me, but I'm afraid will never be good enough for you for some reason, Jogs.

The errors don't balance out if a NT contract is compared with a suit contract. And as Kurt told you it depends on the assumptions on the conditions. A minor deviation on the assumptions can lead to divergent conclusions.
I have seen Pavlicek's work. He has routinely shown poor management of assumptions. And he has drawn incorrect conclusions.

[patmp...@gmail.com]

I agree with that. Its like the old lawyer said, "When I was young I lost some cases I should have won. Later, I won some cases I should have lost. On the average, justice was done."

[paul...@infi.net]

On Wednesday, November 25, 2015 at 4:54:47 PM UTC-5, paul...@infi.net wrote:
> Club matchpoint pairs, South deals, N/S vulnerable
> As East you have:
>
> Q42
> K32
> AK10842
> J
>
> With N/S silent, your auction proceeds:
> 1C-1D
> 2NT-?
>
> Your methods are fairly standard, a 1NT opening would have been 15-17 so you can expect 18-19 or an upgraded 17 form partner, and presumably no four-card major. Three of any suit by responder would be natural and forcing, no special gadgets here.
>
> Your call?

753
98654
J
10976
A86 Q42
Q7 K32
Q7 AK10842
AKQ543 J
KJ109
AJ10
9653
82

Partner (East) settled for 3NT. I thought perhaps 1C-1D; 2NT-3D; 3S-4NT etc. might be reasonable.

[Lorne]

On 07/12/2015 02:43, jogs wrote:
>> <quoting> Pavlicek has shown (using Vugraph data from 1996-2013)
>> that at the part-score level, actual play took 8.08 tricks per hand
>> versus 7.96 double dummy.
>
> That is not the proper way to compare results. One should compare
> first differences. For each separate trial the difference between DD
> and actual play. Average tricks per can be equal when every trial
> produced different results.
>

If you are using DD results to validate a bidding decision I think it is
correct. If DD says you are right to do something 70% of the time and
real life analysis says DD average results are very close to real life
it does not matter if the 70 cases that DD gets right vary significantly
from the 70 or so that real declarers gets right. What matters is that
if you make that decision you will be right 70% of the time.

[imo.c...@gmail.com]

paul...@infi.net於 2015年11月26日星期四 UTC+8上午5時54分47秒寫道：

> Club matchpoint pairs, South deals, N/S vulnerable
> As East you have:
>
> Q42
> K32
> AK10842
> J
>
> With N/S silent, your auction proceeds:
> 1C-1D
> 2NT-?
>
> Your methods are fairly standard, a 1NT opening would have been 15-17 so you can expect 18-19 or an upgraded 17 form partner, and presumably no four-card major. Three of any suit by responder would be natural and forcing, no special gadgets here.
>
> Your call?

I start with 3D. 4C is going to cause a lot of confusion - feel like my p is taking that as GF clubs most of the time.

[jogs]

> If you are using DD results to validate a bidding decision I think it is
> correct. If DD says you are right to do something 70% of the time and
> real life analysis says DD average results are very close to real life
> it does not matter if the 70 cases that DD gets right vary significantly
> from the 70 or so that real declarers gets right. What matters is that
> if you make that decision you will be right 70% of the time.

I don't agree with your statement. You can't actually know if you're right 70% of the time. If you can consistently score 60%, you will win a lot of club games.
It depends on how you define 'right'. Bridge is a non-deterministic/imperfect information game. In game theory 'right' or best strategy is the strategy which maximizes your chance of winning. That does not mean you will win in every case or in this case.
Often DD gives us insight. Please don't elevate higher. It doesn't pass a strict scientific test. Also you must thoroughly explain every assumption you have made.