[gwnn]

not sure what that stat means.....

If I am a top expert and I am exercising my judgement and also another top expert is exercising his/her judgement and we get to a different conclusion, a priori that one of us rates to be right who chose 4M.

So.. umm...?

If I consider myself a random top expert, then I should tweak my game in such a way that from now on I favour 4M a little bit more often than I used to on 4-4 fits and I favour 3NT a little bit more than I used to on 5-3 fits.

[jdonn]

gwnn, on Apr 7 2010, 07:21 PM, said:

not sure what that stat means.....

If I am a top expert and I am exercising my judgement and also another top expert is exercising his/her judgement and we get to a different conclusion, a priori that one of us rates to be right who chose 4M.

So.. umm...?

If I consider myself a random top expert, than I should tweak my game in such a way that from now on I favour 4M a little bit more often than I used to on 4-4 fits and I favour 3NT a little bit more than I used to on 5-3 fits.

In fairness it doesn't give any statistics for 3NT when one side bids it directly and the other side investigates something on the way.

[gwnn]

I think it would be hard to make any small modification to those statistics to make them useful in practice since the basic set-up seems to be that on some set of hands experts disagreed as to the optimal strategy and some of the numerous strategies prove to be slightly more successful than others. This does not mean that some strategies are inherently flawed, only that top experts, on average, seem to choose some strategies too often.

The way this kind of statistics could be useful if there were some definite schools suppose N Ossetian experts told you "only bid stayman if you have a notrumpiness of at most 6.7 on a 10 scale" and S Ossetian experts told you "oh no that's stupid, the correct limit is at 6.3". Now you could look up all the Interossetian championship hands and decide which limit is more appropriate (after eliminating random noise inherent of differences in skill level etc, for example N Ossetians don't like to underlead J8xx).

[cherdanno]

I think Pavlicek's statistic isn't very useful as it contains competitive auctions. If one partnership doesn't find their 4-4 major suit fit in a competitive auction and the other one does, well of course the one that finds the fit will come out on top.

[Fluffy]

IMO with 4333 it is best to use stayman if you find your fit since it probably puts you in a better place, however it is not worth the info you give when you fail.

If using 1NT-3♥/♠ as specific stayman I think I'd use it with most 4333s.

[Mbodell]

Klinger in his book on Keri talks about 3nt being usually better if you are 4333 opposite 4333 (in the 4-4 fit) and if you are 4333 opposite 5332 (in the 5-3 fit) but that 4333 opposite 4432 is usually better in the 4-4 fit. In a couple of partnerships I play keri and thus can investigate what shape my partner is if I want, but I still sometimes skip the science on 4333 hands for fear of information leakage. But our NT shapes are 4333, 4432, 5332, 5m422, and some 6m322 so the odds of a good 4-4 fit when I have 4=3=3=3 or 3=4=3=3 aren't that great.

I do play 3M as game forcing asking for 4 card support in the unbid major in a couple of partnerships, but there I usually don't make that call with 4333 either, only with 4432 or 5m4M22.

[fred]

David Stern's letter appears on page 44 of the February 2010 issue of The Bridge World.

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

[ulven]

fred, on Apr 7 2010, 08:02 PM, said:

David Stern's letter appears on page 44 of the February 2010 issue of The Bridge World.

Was initially published in San Remo Open Europeans bulletins last summer.

[Mbodell]

And in fact if you look here (warning, PDF) you will see the 4 part article on the superiority of 3nt.

[chasetb]

P_Marlowe, on Apr 7 2010, 04:18 PM, said:

Regarding the 4333 Stayman issue - a possible agreement is
to play a 3H / 3S response as showing 4333 , the 4 card the
other major.

This would ensure, that you play 3NT if partner is also 4333,
otherwise you play the major.This assumes of course, that you don't use 3H / 3S for something else.

This convention was posted in one of the ACBL Bridge Bulletins under the "Bidding Lab". Using Stayman with a 4=3=3=3, and the auction going 1NT - 2♣ - 2♠,
3♥ tells partner you have a 4=3=3=3 GF hand, and to decide on either 3NT or 4♠. With a 4=3=3=3 or a hand with values NOT in ♠, opener bids 3NT; otherwise bid 4♠.

If you are 3=4=3=3 and the auction goes 1NT - 2♣ - 2♥, 3♠ does the same thing.