[Zelandakh]

 fromageGB, on 2013-February-13, 11:30, said:

OK, thanks for that, but what I was getting at was what is opener supposed to show? I play at the moment that 1NT 2♣; 2♠ 3♥ is SID and nothing else. If it was also a "strong raise", would opener be bidding something like 3NT = 4333 shape, cue bid = other shape with slam-suitable hand, in case responder wants to go further than game? It is not something I have come across. I can see it could be very useful if you played that 1NT may include a 5 card major and/or a singleton!

I play this in the equivalent sequence for Puppet (1NT - 2♣; 2♦ - 2♥; 2♠) so that takes the 5 card major aspect out of it. I designed it originally for a system based on normal Stayman and I was assuming that in the earlier discussion. After 1NT - 2♣; 2♠ - 3♥ or 1NT - 2♣; 2♥ - 3♦, the simple way is to play
3M = 4333
3M+1 = not 4333, bad for slam
3M+2 and up = not 4333, good for slam, cue

Over 3M, 3NT is a sign-off and other bids are cues. Over spades it is better to invert the first 2 steps but consistency between the sequences is more important imho. You also lose Frivolous when Opener has a 4333 hand. On the other hand, getting the additional distributional information is often more useful.

[FrancesHinden]

 rhm, on 2013-February-10, 08:53, said:

The reason is if double dummy works as a good proxy for the expected result single dummy in general, there is no Bridge (or statistical) reason why the results should somehow be different when making some specific Bridge assumptions about distribution and HCP, but which are otherwise not related to the play of the hand itself.
What matters is only whether the number of deals in the simulation is high enough to give an approximation about the actual result for the total number of possible deals where this scenario can exist.

Much of what you have been stating as fact is your opinion, not a fact. I agree with gnasher and disagree with you about some of these so-called 'facts' but there's no way to prove many of them.
However, the above statement is demonstrably untrue. For one thing, there's absolutely no reason why it should be true, or indeed why you should even assume it to be true and no statistical reason whatsoever why you might choose to assume it.

The most obvious example is when someone (other than the opening leader) has a side suit void. Voids for the declaring side increase the number of single-dummy tricks for declarer, and a defender's void decreases them. Particularly in the slam zone. This is another opinion I am stating as a fact, but at least you should be able to see the logic to it: the opponents cannot "cash their ace and switch", they have to lead the right suit at trick 1. (Having an auction often decreases the SD vs DD difference here, of course).

Here are some other setups where specific assumptions on HCP and distribution are likely to skew the SD vs DD results:
- give opener a strong NT and responder an 9-count with a 6-card suit. My guess is that 3NT SD will strongly outperform DD because the lead is likely to be very important
- give opener a 4333 and responder a 3343 and a combined 36-count. DD will beat SD strongly on average.

[PhilKing]

 lalldonn, on 2013-February-13, 18:51, said:

That's a different situation since (for most people?) responder might not be balanced for transfer then 2NT. It would be nice to have a bid that says "I want to play 3NT if you are balanced" there, but there are other concerns and you can't do it all.

If only I could have persuaded you to play invitational transfers ...

The NT v M decisions become trivial.

[Fluffy]

 JLOGIC, on 2013-February-07, 14:31, said:

This type of hand shows why matchpoints is an incredibly difficult and awesome game imo, you have to balance tactics with not being too ridiculous, trade off scientific accuracy with keeping them in the dark, etc.

***** MPs, I went down today at 6♠ with this:

♠AJxx
♥KQ9x
♦AK9xx
♣-

♠K987
♥A10
♦Q10xx
♣Axx

just to find out that less than 30% of the field was playing slam O_o, obviously I did an awful job assesing the field's level, should had remembered that on round 2, 3♥+2 on 29 combined turned out to be a full average

At least it was on the end and didn't have to answer someone saying.... don't you know to test the suit with the king?

[Alik1974]

Very interesting question. OP has mentioned all the major reasons why one would prefer 3NT over 4M. Here is my personal view :

-------------------------------- MOSTLY BECAUSE ------------------------------
C- Soft values
G- rightsiding
A- 4333 shape
----------------------------------------- AND ALSO BECAUSE ----------------------------
I - Both hands are balanced
F- bad trumps
H- Concealing your hand

Also I agree with Fluffy about the side minor thing.

[lesh]

I remember that I read somewhere that the Blue Team played always 4M if they had to choose btw 3nt and 4M

[GreenMan]

 lesh, on 2013-February-21, 10:45, said:

I remember that I read somewhere that the Blue Team played always 4M if they had to choose btw 3nt and 4M

And when's the last time they won anything?

[PhilKing]

 lesh, on 2013-February-21, 10:45, said:

I remember that I read somewhere that the Blue Team played always 4M if they had to choose btw 3nt and 4M

They were probably just reducing variance - their edge in other areas was so large, playing in the eight card fit was a sound tactic, since that is what everyone else did.

[aguahombre]

 PhilKing, on 2013-February-21, 10:51, said:

They were probably just reducing variance - their edge in other areas was so large, playing in the eight card fit was a sound tactic, since that is what everyone else did.

This is probably true, but it creates another interesting strategy question. We know Justin believes very strongly that 3NT should be the choice in those cases were there is an informed choice. So, he and his partner/teamates are going to go with what they think is right and not be concerned with variation strategy against top competition.

But, if he were in an earlier round against a team of us sheep, would he go for the reduction in variance or stick with what he prefers?

[WellSpyder]

 aguahombre, on 2013-February-21, 11:14, said:

would he go for the reduction in variance or stick with what he prefers?

If you always try to do what your opponents do to reduce variance, then you won't have an edge any more!

[aguahombre]

 WellSpyder, on 2013-February-21, 11:21, said:

If you always try to do what your opponents do to reduce variance, then you won't have an edge any more!

Did someone say always? We are talking about a strategy. If your opinion is that top pairs should not consider that strategy, fine. I was interested in whether and to what extent top pairs do use it and asked my question to possibly get an answer. Don't assume what you think I would always or never do.

[PhilKing]

 WellSpyder, on 2013-February-21, 11:21, said:

If you always try to do what your opponents do to reduce variance, then you won't have an edge any more!

This is a strategic are where one approach will work, for arguments sake 10% more often than another. In most situation, the better player will be doing something that is always or overwhelmingly better.

I've looked through every available Vanderbilt, Spingold and USBC swing 3NT hand post year 2000 from semi finals and onwards over the last year and I have seen one striking example of a pair reducing variance: Meckwell play a very ABC 4th highest lead style versus 3NT versus most teams, and I strongly suspect this is a case in point. They defend better than almost any other pair, so reduce trick 1 variance. The edge for making a better theoretical lead is perhaps less and certainly more variable than the edge on subsequent defence.

I might have another look to see if they vary it up a bit against top European teams who are more inclined to lead from weak holdings.

[cherdano]

Or maybe Meckwell think leading a straightforward 4th best against 3NT is just normally the best lead?

[JLOGIC]

You really think Meckwell are purposely making leads they think are less likely to defeat a contract to reduce variance? Come on.

[JLOGIC]

It's interesting that you chose a pair who plays an anti field system, complete with anti field NT ranges, anti field transfers in competitive auctions that will lead to playing hands from different sides, open very light (anti field action), and one of their players is well known for doing very weird/abnormal things in the bidding more than any other top american player.

Perhaps normal leads including leading your longest suit are good leads, or perhaps they think so.

[jogs]

 benlessard, on 2013-February-13, 18:34, said:

Had a MP hand today with a student.

Axx
KQx
Axx
KJxx

1Nt-2D
2H--2NT (inv)
??

Partner had a 3532 and even if both hands had no spot and we had great trumps, 3Nt was better than 4M in MP, so in Imps it wouldnt have been close at all.

What people do here with a (3334) bid an artificial 3m ?

This is a probability game. With 4333 after
partner opens 1NT, the 4-4 fit makes 0.5 tricks
more than NT. 4M requires one full extra trick.
Just bid 3NT.
Of course, opponents can easily run five tricks
off the top in some suit. In the long run you
win more imps by going with the odds.

[Mbodell]

 benlessard, on 2013-February-13, 18:34, said:

Had a MP hand today with a student.

Axx
KQx
Axx
KJxx

1Nt-2D
2H--2NT (inv)
??

Partner had a 3532 and even if both hands had no spot and we had great trumps, 3Nt was better than 4M in MP, so in Imps it wouldnt have been close at all.

What people do here with a (3334) bid an artificial 3m ?

This is a topic and hand that illustrates some of Keri's strengths. The auction would go 1nt-2♣-2♦-2♥ and now partner would have shown an invite with 4 or 5 hearts. This has the advantage that if you decline the invite you can still play 2♥. On this hand where you are a maximum and accepting the continuation is 3♦ which shows you have a 4333 pattern. Partner can bid 3♥ if partner wants to play in 4♥ if you are 3=4=3=3 (and you'll bid 3nt with any other 4 card suit), or partner can bid 3nt to say they don't want to play in 4♥, even when you are 3=4=3=3. Generally, the hands that bid 3nt are the 5332 and the 3=4=3=3, while other shapes like 4432 would look for a 4♥ game if the 4-4 fit is found.

I think we'd all agree that with mirrored flat shapes 4333 you belong in 3nt. With 5332 opposite 4333, at least if not 5-4 fit, most would prefer 3nt. This thread shows many think even with a 5-4 fit, you'd prefer 3nt - although you can bid 3♥ to look for the 9 card fit if you disagree. There are systems to try and diagnose and deal with this issue.

[jogs]

 Mbodell, on 2013-February-21, 19:59, said:

Generally, the hands that bid 3nt are the 5332 and the 3=4=3=3, while other shapes like 4432 would look for a 4♥ game if the 4-4 fit is found.

3433 facing 4432.
The 4-4 heart fit wins about 0.5 more tricks than 3NT.
In the long run 3NT is the winner.

There needs to be doubletons in two different suits
before 4-4 fit wins one full trick more than NT.
3424 facing 4432.