[awm]

Suppose you have a pair of hands with game values. There is a suit in which you have xx opposite xxx in the two hands, but otherwise you have nine top tricks for notrump.

Alternatively, you can play five of a minor. This is on a two-way finesse in a side suit.

It's easy enough to compute a priori odds -- the weak suit breaking 4-4 is less likely than a finesse. But these odds all must be modified slightly because if you bid 1NT-3NT, opponents might not lead the weak suit. On the other hand, good declarers will get a 2-way finesse right more than half the time as well, and there is even a remote chance that opponents lead the two-way finesse suit for you.

As a side question, would you feel differently about this if the side suit were xx opposite xx instead of xx opposite xxx?

[han]

I'd slightly prefer 1NT-3NT I think, but with xx vs xx I'd much prefer 5m.

[ArtK78]

han, on Apr 29 2008, 12:16 PM, said:

I'd slightly prefer 1NT-3NT I think, but with xx vs xx I'd much prefer 5m.

It is hard to argue against this position.

By the way, a blockage in the unguarded suit is another possible way to succeed in 3NT. I ran into an annoying blockage in a suit against a 4♠ contract the other night. I had the singleton ♦A and partner had KJxxx. Partner had no entry so even after my lead of the ♦A we could not get a second diamond trick before declarer got rid of his diamonds on dummy's long club suit.

So, blockages happen, even in suit contracts. And, of course, sometimes the opponents find a way to block the suit on their own.

[Apollo81]

You definitely want to play 5m with xx opposite xx, and I think you definitely want to play 3NT on 1NT-3NT with xx opp xxx if that is a minor suit.

[pclayton]

han, on Apr 29 2008, 09:16 AM, said:

I'd slightly prefer 1NT-3NT I think, but with xx vs xx I'd much prefer 5m.

My thoughts exactly. Not only do we win with 4-4 - we will frequently win with 5-3 and the 3 card suit on lead.

With xx opposite xx its a bigger problem. Now, both opponents can and frequently will lead the suit. Unless you get some blockage due to inept defending, you are down off the bat. Your only real hope is that they split 6-3 with the 3 on lead.

[hrothgar]

For what its worth, I think that this might very well be a problem that is well suited for BRBR.

Look at the set of all hands that bid

1N - (P) - 3N (or some variant there off)

Find hands where the declaring side has a holding of xxx versus xx or some such.

Determine the frequency with which the opponent's hit the weak suit on opening lead... (You probably need to take 4-4 splits into consideration)

For the set of hands where the opponents made the wrong lead, determine how often the good guys can cash nine tricks...

[cherdano]

Unless we are playing we are playing vs gnome, the odds for 3N are fairly easy to compute approximately. We are going down if opening leader has 5 or more of the suit, which is roughly 34%. (I will ignore the case of him being 5-5 and leading the other suit, since our combined hands are fairly balanced under the assumptions, and since his other 5-card suit is very likely of worse quality.)
If opening leader sometimes doesn't lead his longest suit, probably the odds go up.

[pclayton]

cherdano, on Apr 29 2008, 10:25 AM, said:

Unless we are playing we are playing vs gnome, the odds for 3N are fairly easy to compute approximately. We are going down if opening leader has 5 or more of the suit, which is roughly 34%. (I will ignore the case of him being 5-5 and leading the other suit, since our combined hands are fairly balanced under the assumptions, and since his other 5-card suit is very likely of worse quality.)
If opening leader sometimes doesn't lead his longest suit, probably the odds go up.

The problem is Gnome will try to lead his void. If this isn't available, he'll lead the card closest to his left thumb.

[gnasher]

The way that the question is phrased doesn't consider another disadvantage of investigating for weak suits: you may determine that you don't have a wide-open suit and therefore that 3NT is the right contract, but in doing so may have made he lead or subsequent defence easier for the opponents.

[han]

Jlall, on Apr 29 2008, 01:56 PM, said:

I prefer the roughly 90 % contract

90% chance of going down when I'm defending against you?

[whereagles]

Apollo81, on Apr 29 2008, 05:34 PM, said:

You definitely want to play 5m with xx opposite xx, and I think you definitely want to play 3NT on 1NT-3NT with xx opp xxx if that is a minor suit.

That has been studied by french theoreticians and the outcome is you should play 5m with x/xxx but stick to 3NT with xx/xxx or even xx/xx. Of course, this is on abstract, with opps passing and you being careful not to tip off the weakness...

[han]

The French? Don't they use the metric system?

[whereagles]

metric system pwnz joo!

[hrothgar]

whereagles, on Apr 30 2008, 12:51 PM, said:

Apollo81, on Apr 29 2008, 05:34 PM, said:

You definitely want to play 5m with xx opposite xx, and I think you definitely want to play 3NT on 1NT-3NT with xx opp xxx if that is a minor suit.

That has been studied by french theoreticians and the outcome is you should play 5m with x/xxx but stick to 3NT with xx/xxx or even xx/xx. Of course, this is on abstract, with opps passing and you being careful not to tip off the weakness...

In general, these types of citations aren't particularly useful

I couldn't care less what "french theoreticians" have to say... I'm qually bored with random anecdotes from the Germans, the Irish, or the Yanks for that matter.

I would be quite interested in more specific information... For example:

Who are these "french theoreticians" ?
Where did they publish their results ?
What methodology did they use to verify their claim?

You might note that this comment is very similar to the one that I made about your simulations where I noted the danger in taking random information at face value without any means of evaluation...

[whereagles]

> Who are these "french theoreticians" ?

Mark Kerlero and other people at the Université de Bridge. I think Mark is a former world junior champion (not sure, though).


> Where did they publish their results ?

In Mark Kerlero's book "Enchéres, mode d'emploi."


> What methodology did they use to verify their claim?

I don't know.


I believe these guys because they invented and polished up french standard (SEF), a bidding system that is, in my opinion, a masterpiece of systematization. If you want more references, I suggest you try to contact them. Not sure if Mark is still playing, though.

[awm]

Note that this is not a general question "should I bid 3NT with xx opposite xxx" but rather a specific question about the odds on this holding versus 5m on a two-way guess.

There are probably a sizable number of hands where 5m has no play whatsoever and you want to be in 3NT on xx opposite xxx because it's more than 50% (counting the possibility of a wrong lead). It does not necessarily follow that 3NT is the right answer here -- suppose I told you that 5m was completely cold whereas 3NT needs opponents not to cash five tricks in a suit that's xx opposite xxx, wouldn't you want to be in 5m then?

My personal preference is to play in 5m. I suspect that my odds of making the two contracts are actually pretty similar -- if I need to guess a two-way finesse in a side suit where I have time to get count, I suspect I can get it right close to 2/3, which is also about the odds for them not cashing five tricks in my xx opposite xxx suit (assuming they always lead this suit from five-plus and never from three-minus). The reason I prefer 5m is that I like to feel like I'm in control of the result. By bidding the contract that's double-dummy cold I can always make it if I guess right. Bidding 3NT puts me at the mercy of the opponents, a less comfortable position (for me anyway) when the odds are close to equal.

[gnasher]

whereagles, on Apr 30 2008, 06:29 PM, said:

I believe these guys because they invented and polished up french standard (SEF), a bidding system that is, in my opinion, a masterpiece of systematization.

Is it their methodology that you admire, or the resulting system, or both?

Also, is the official documentation of SEF available online? I've found these:

http://usf.bridge.free.fr/bridge/sef/
http://p.gourinel.fr...e/Encheres.html
http://pagesperso-or...ard/frame23.htm

but I'm not sure if any of these is the original system, rather than someone else's interpretation.

[MickyB]

I'm not sure the answer to this question is terribly relevant.

With xx opposite xx in the scenario presented, I'd rather be in 5m. However, I can't imagine getting there terribly often - it would require giving too much info to opponents on hands that will lead to 3NT anyway. Plus, how do you propose to diagnose that we aren't missing an ace on the side as well? Admittedly 3NT may well be off then, even if they don't lead their suit they'll usually have a chance to switch to it.

A related question that I've been considering is "After 1N:2♣, 2M:??, should a (43)(42) with xx in an outside suit investigate alternatives to 3NT?", however that is more aimed at playing 4M in a 4-3 than playing 5m.