When to avoid stayman

[pclayton]

Assume its matchpoints and you have at least 10 points.

 

Pard opens a 15-17 NT. When will you avoid the use of Stayman?

[han]

When I'm 4333.

[helene_t]

Soft cards should probably be a factor, too, but basically I Stayman if I have a doubleton.

[marcD]

Never . Actually I bid 3♦ with 4M333 and partner can relay if interested .

[jdonn]

I don't stayman if 4333, and I usually don't stayman with 4432 hands if my doubleton is very strong and my major is weak. The last one has given me fairly mixed results so far, sometimes better sometimes worse and usually not mattering much.

[MickyB]

Ditto Josh, plus stronger hands will be less likely to use Stayman - I'd just bid 3NT on a 4432 with Qx and 13+.

[cherdano]

I am more likely to stayman at matchpoints.

[ArtK78]

At matchpoints, I tend to avoid Stayman with 4333 and with above-minimum game forcing hands.

 

Normally, if the combined hands have about 29-31 HCP they will make the same number of tricks in notrump as in a suit contract if both hands are balanced.

 

At IMPs, I tend to Stayman all the time. There is always the chance that your doubleton will be opposite partner's doubleton or one of you will have a weak 3 card holding opposite shortness.

 

My response does not fall into any one of the poll categories neatly, so I did not vote.

[hrothgar]

Ditto Josh, plus stronger hands will be less likely to use Stayman - I'd just bid 3NT on a 4432 with Qx and 13+.

What he said:

 

The 4-4 major suit fit is much more likely to generate an extra trick if we have 24-26 HCP than 27-28.

[awm]

I am almost exactly the opposite of Art.

 

With two balanced hands, the 4-4 major fit will usually produce one additional trick.

 

At IMPs, when the major fit produces one more trick I'm happy to be in 3NT even though it plays a trick worse. So it's just between the "major equal or worse" and the "major produces two more tricks." This is fairly close; when I am 4333 or have a strong doubleton it is unlikely that the major produces two more tricks so I will just bid 3NT.

 

At MPs, the most frequent "4-4 fit one-trick better" situation means you have to bid 4M on these hands. So I would virtually always bid stayman.

[jdeegan]

:) The expert opinion (esp. at IMP's) is to avoid Stayman when the combined point count is 29 to 31 HCP, the major fit is potentially weak, and your hand is balanced - 4-3-3-3, or 4-4-3-2 shape. You avoid losing three unavoidable trump tricks plus another when you have most of the power.

 

Old fashioned lesson - read and learn.

[P_Marlowe]

At matchpoints, I tend to avoid Stayman with 4333 and with above-minimum game forcing hands.

 

Normally, if the combined hands have about 29-31 HCP they will make the same number of tricks in notrump as in a suit contract if both hands are balanced.

 

At IMPs, I tend to Stayman all the time. There is always the chance that your doubleton will be opposite partner's doubleton or one of you will have a weak 3 card holding opposite shortness.

 

My response does not fall into any one of the poll categories neatly, so I did not vote.

agree.

 

With kind regards

Marlowe

[Cascade]

I don't make much change in my style here between MPs and IMPs although I am occasionally willing to take more risks with 3NT at MPs.

 

I bid 3NT without investigating a 4-4 major fit when the hand fits 'enough' of the following hand types:

 

4-3-3-3

4-4-3-2 with a strong doubleton

a preponderance of Queens and Jacks over Aces and Kings

good spots - tens and nines

extra values

 

I tend not to bid 3NT without investigating a potential major fit with:

 

weak doubletons

two or more weak three-card suits

[pclayton]

I'm drawing up a lesson plan and my thinking was more along the lines of Art's: Don't use Stayman with extras, and a 4-3-3-3. This is what I've always been taught, and I haven't seen a lot of contradictory evidence.

 

I've also believed that soft cards and intermediates are a NT indicator, but perhaps that is just a tie-breaker.

 

4432 with a strong doubleton may or may not produce the extra trick. It is dependent on pard's holding opposite.

[TimG]

Edit: awm pointed out that my conclusions were faulty, I'm going to rerun the simulations with bigger sample sizes and (I hope) better address the MP angle.

[junyi_zhu]

Assume its matchpoints and you have at least 10 points.

 

Pard opens a 15-17 NT. When will you avoid the use of Stayman?

The system design should allow the responder to find 4-4 fit and still be able to stay in 3NT if opener has the right cards. Also, so many open 1NT with 5M, so you really don't want to miss 5-4 fit 4M, in that case, skipping stayman should be rather bad. So generally speaking, one should always try for stayman and the system design should allow him to choose 3NT or 4M when the major fit is discovered. Otherwise, many 5-4 fit and 4-4 fit suitable for 4M would be missed.

[cherdano]

Assume its matchpoints and you have at least 10 points.

 

Pard opens a 15-17 NT. When will you avoid the use of Stayman?

The system design should allow the responder to find 4-4 fit and still be able to stay in 3NT if opener has the right cards. Also, so many open 1NT with 5M, so you really don't want to miss 5-4 fit 4M, in that case, skipping stayman should be rather bad. So generally speaking, one should always try for stayman and the system design should allow him to choose 3NT or 4M when the major fit is discovered. Otherwise, many 5-4 fit and 4-4 fit suitable for 4M would be missed.

So that we can maximize the information available to defenders?

[CSGibson]

I never stayman with 4-3-3-3, I will always stayman with shortness somewhere, even if it's AK.

 

My partnerships often open 1N with a 5 card major, but I don't think that will make a difference; I'm never ruffing in the short hand if I'm playing in a trump contract, so missing a 5-4 fit isn't the end of the world, while hiding info from the defense on the 19 hands out of 20 where we don't have a 9 card trump fit always helps.

[Finch]

Assume its matchpoints and you have at least 10 points.

 

Pard opens a 15-17 NT. When will you avoid the use of Stayman?

The system design should allow the responder to find 4-4 fit and still be able to stay in 3NT if opener has the right cards. Also, so many open 1NT with 5M, so you really don't want to miss 5-4 fit 4M, in that case, skipping stayman should be rather bad. So generally speaking, one should always try for stayman and the system design should allow him to choose 3NT or 4M when the major fit is discovered. Otherwise, many 5-4 fit and 4-4 fit suitable for 4M would be missed.

So that we can maximize the information available to defenders?

While concealing opener's distribution is worth something, I've noticed that whenever I get to 3NT with no stop in one of the suits in the auction 1NT - 3NT, the opponents have a horrible tendency either to lead the suit, or to switch to it.

 

If you don't have 9 cashing tricks, it's often wrong to play in 3NT even if you give the opponents a blind lead. Maybe this is taking a good idea too far, but in a way it might be in the long term be superior to raise 1NT to 3NT looking at a 2=6=3=2 distribution than a 3=4=3=3 hand.

 

I agree that system design should allow responder to discover if there is a 4-4 fit, offer the choice between the fit and 3NT, and keep declarer's hand type and range concealed. Funnily enough, in one partnership that makes looking for a fit with spades more tempting than looking with hearts, because more of opener's distribution is kept confidential.

[georgi]

would you make stayman bid with

 

♠Jxxx ♥Jxxx ♦xxxx ♣x

 

If yes, what you are going to call after

 

1nt ( partner ) - pas - 2♣(you), 3♣(LHO) - X ( partner) - pas - ?

[awm]

Assume its matchpoints... when will you avoid the use of stayman?

 

Seems like there is widespread agreement here!

 

When I'm 4333.

basically I Stayman if I have a doubleton

I don't stayman if 4333

Ditto Josh

At matchpoints, I tend to avoid Stayman with 4333

agree

I don't make much change in my style here between MPs and IMPs although I am occasionally willing to take more risks with 3NT at MPs.

I'm drawing up a lesson plan and my thinking was more along the lines of Art's.

 

There are only a small number of dissenters:

 

I am more likely to stayman at matchpoints.

At MPs, the most frequent "4-4 fit one-trick better" situation means you have to bid 4M on these hands. So I would virtually always bid stayman.

 

Yet the simulation results imply that the dissenters were right!:

 

When opener has 4432 with 4 spades and 15 HCP and responder has 4333 with 4 spades and 10 HCP, spades produces 9.31 tricks and NT produces 8.47 tricks, a difference of 0.84 tricks. Game in spade makes 40.6% and game in NT makes 48.8%.

 

Note that the average tricks in spades are substantially higher, meaning that you will very often get a better matchpoint score in 4♠ rather than 3NT even with 4432 opposite 4333. At IMPs however, avoiding stayman is best. Nonetheless the original question was about matchpoints! Even the guy who ran the simulation gets this wrong:

 

This suggests to me that even with a barely GF 4333, skipping Stayman is a good idea.

[TimG]

Nonetheless the original question was about matchpoints! Even the guy who ran the simulation gets this wrong.

You are correct: I did not read the data properly. I'll edit the post in a bit so that the wrong conclusions don't accompany the data for all eternity.

[MickyB]

Yes, the sim results say, all else being equal, that you want to be in 4M when you are 4333; However, all else isn't equal. By Staymaning, you are giving away info on hands where you'll always land up in 3NT; It's also quite possible that double-dummy analysis is biased against 3NT, where the opening lead is probably more important.

[TimG]

Sorry about the bad data/conclusions. This time I have done:

 

south 4(432) with 15 HCP

north 4333 with 10 HCP

1000 deals

 

average tricks in spades: 9.305

average tricks in NT: 8.383

difference: 0.922 tricks

 

For each of the 1000 deals, I compared the double dummy result in 3N to the double dummy result in 4S and assigned 1 point for a better score in 3N, 1/2 point for a tie, and 0 points for a better result in 4S. The total was 459. I hope this means that if you play in 3N while the field is playing 4S (and everyone plays double dummy) you will score 45.9% in the long run.

 

As before: One must consider that these are double dummy results and that declarer's real world advantage will be greater when the auction is 1N-3N than when the auction is 1N-2C-2D-3N, 1N-2C-2M-3N, or 1N-2C-2M-4M. So, you must not only consider the relative merits of NT vs 44 when there is a fit, but the cost of investigation both when there is a fit and when there is no fit.

 

Thanks, Adam, for pointing out my error.

[ArtK78]

I don't doubt that the 4-4 fit will often produce more tricks than NT. However, it is also very true that when you do not find a 4-4 fit, by providing distributional information to the opponents you will have given up some tricks.

 

Furthermore, even when you find the 4-4 fit, the opponents will have more information than if the auction had gone 1NT-3NT. So, even with the large plus for playing in the 4-4 fit (as clearly demonstrated by the simulation), there will also be a minus in that the defenders will have more information.