[Phil]

Jexa_ issued an Arena challenge to me this morning and there was an interesting hand from the set. Let's start off with a bidding problem:

♠A2♥AKxx♦AJ9xx♣AK

I assume you agree with my choice of 2♣, followed by 2N? Partner now transfers to 3♠ and rebids 3N. Probably right to sit?

The book situation for pulling 3N with a doubleton in partner's suit is when you have AK. Marshall Miles discussed this in one of his books (All 52 Cards probably). the reasoning is that you will have communication problems in NT with this holding, but not in the 5-2 fit.

Partner's most likely patterns are 5(332) and perhaps 52(42) or possibly a weak 5143. I thought partner was an underdog to have three diamonds, making 3N pretty awkward. We have a timing problem in 3N with either round suit lead, so in spite of my putrid spade holding I tried 4♠.

Partner had ♠QJ98x ♥x ♦Q87x ♣J9x.

On a club lead, I played sA, s. The dT came back so I hopped, played hA, ruffed a heart, gave up a diamond and made 5. Trumps were 3-3 and the pointed Kings were split.

Jexa_ made 3N (I think 4 can be made but it risks the contract) so I won the board.

The reason I post this is, maybe there's some rationale of playing the 5-2 not just when we have the magical AK or KQ, but anytime one hand is very strong and the other is weak.

Does someone want to run a sim for this? I'd suggest the parameters are that Opener is 20+, the partnership has a known 5-2 fit, (and not an 8 card fit in the other major or a 10 card fit in a minor) and the combined assets are 24-28 HCP.

[Fluffy]

I kinda feel that the weaker responder's hand the more you want to play, so it wiill not suffice with 20+, you will most likely want to run simulations for 20, 21, 22, 23, 24 all separatedly to come with wich holdings it is better.

[MrAce]

 Phil, on 2017-February-15, 11:01, said:

The reason I post this is, maybe there's some rationale of playing the 5-2 not just when we have the magical AK or KQ, but anytime one hand is very strong and the other is weak.

Does someone want to run a sim for this? I'd suggest the parameters are that Opener is 20+, the partnership has a known 5-2 fit, (and not an 8 card fit in the other major or a 10 card fit in a minor) and the combined assets are 24-28 HCP.

Problem is you can not be sure that your combined hcp are 24-28. Because it is actually 24-31 hcp.
Each time Marshall tried this we got a field zero. Found me with 6-8 hcp and bailed too many spade tricks. Or not more tricks than NT would take. Unless you find an illegal way of communicating how weak pd bid 3 NT, his idea of AK-KQ trump holding should play 4M sucks big time.
Oh and I forgot, but you can imagine, you are definitely contrasting your contract from the field. That is something Marshall always loved to do. Results are random. Not everyone likes this. Good players actually hates it.
But let me tell you something funny, he actually had amazing talent to play these 5-2 and 4-3 and 4-2 fit hands compared to his talents of playing 8 or 9 or 10 card fit hands.Once you start to bid like him, you get good at those fits and feel stranger to better fits!

[cherdano]

Maybe it's responder who should insist on playing 4M with a weak hand that has no entry outside trumps? Your 2C-2N opener will usually have Hx support.

[rhm]

 Phil, on 2017-February-15, 11:01, said:

Does someone want to run a sim for this? I'd suggest the parameters are that Opener is 20+, the partnership has a known 5-2 fit, (and not an 8 card fit in the other major or a 10 card fit in a minor) and the combined assets are 24-28 HCP.

I did this according to the above specification plus 2 in addition.
But before reporting results I would like to mention what made your takeout from 3NT to 4M successful.
I do not criticize the bidding, which is due to system constraints.
However fact is:

openers hand is not notrump suitable:

He got 5422 distribution which is more suited for a trump contract
It has no less than 10 controls and no queens. In fact the only non control honor card is a single jack.

What made the 5-2 fit good was that repsonders 5 card suit had "body". QJ98x is an extremely lucky buy for such a risk.

Result with the above suggested parameters (1000 deals on each simulation):

4M made on 551 deals
3NT made on 665 deals

Average number of tricks in 4M 9.57
Average number of tricks in 3NT 9.06

In practice 3NT would have an even bigger advantage single dummy. (wrong opening lead etc.)

So the basic answer is no, but your specifications may not have been good enough. For example with xx in the major opener will not disturb 3NT.

Another sim, this time point range between 24-31 and opener has at least one of the top 3 honors in the major, rest of the specifications as above:

Result:

4M made on 684 deals
3NT made on 751 deals

Average number of tricks in 4M 10.10
Average number of tricks in 3NT 9.71

Still no good
One reason surely is that bad trump breaks are of course more likely once the combined trumps you have yourself is reduced. Few 5-2 fits will play well once opponents trumps break 5-1

Last try: What happens if we leave all as on the last sim, but insist opener to have AK tight.
This is the worst for 3NT since you are totally blocked in responders major, because you can not even overtake this holding in 3NT and openers resources in the major mean he has less outside in HCP:

Result:

4M made on 662 deals
3NT made on 696 deals

Average number of tricks in 4M 9.99
Average number of tricks in 3NT 9.44

So even the last simulation is negative for the 5-2 fit.
It is my believe you need at least an 0.7 trick advantage on average, before you should prefer 4M to 3NT at the table.

The only way you could cater sensibly for good 5-2 fits would be to be more discriminating with your 5 card suits, whether to transfer or use Stayman.
Say you have enough values for 3NT, you could in principle agree only to transfer if you would not mind if opener takes you back on a very strong doubleton.

Of course you would end in 3NT with some 5-3 major suit fits, not the end of the world when repsonder is balanced too and his 5 card suit is weak.
Trouble is what to do with responder hands which are distributional, but have a mediocre 5 card major.
Now you probably prefer 5-3 major fit at almost any cost. Stayman will not do.

Rainer Herrmann