[undoubling]

In my experience, whether or not to open 1NT with a 5-card Major is dependent on the OTHER major suit holding. IF that other major is a worthless holding, then I'll open the Major suit. If the other Major has working cards, then I'll open in NT.

[pescetom]

 mycroft, on 2020-October-27, 17:56, said:

The assumption that they will never bid over strong NT died thirty years ago

As said, not considering interference over 1NT was a lazy assumption for a script in a language I do not yet know fully. It makes little difference if they win the auction when they would have done the same over 1♠, or if they push us to 2NT making, but it does change things if they help us find 2♠ instead of 1NT, as nullve pointed out. I'll try to incorporate at least some estimate of that in the simulation.

 mycroft, on 2020-October-27, 17:56, said:

the assumption that after 1♠-2♠, they will never bid is dying; frankly, it's more likely that they will pass out 1♠ than 2.

Competent opps with a 9-card ♥ fit will go to 3-level of course, but that is not going to change the proportions much: seems to me opps are almost as likely to find such fit over 1NT as over 1♠.

 mycroft, on 2020-October-27, 17:56, said:

What you've shown is that a trump fit tends to play a trick better than in NT, if extra strength isn't there.

Well no, the number of tricks in ♠ vs NT has been analysed here before and I can calculate it with a 1 line mod to this script too. What I wanted to see instead is how often this results in a better/worse/same score on the same hand, which is of course what counts in MP terms.

Of course I knew beforehand that 1NT was going to cost here, and agree that there are significant benefits elsewhere to offset that cost. But I like to know how big such costs are, especially when going against most of the room (as I do when I open 1NT here).

[mycroft]

I agree with almost all of that (although competent opps over 1♠-2♠ will only let you play it when they can see it's bad for you or for them, not just when they have a 9-carder. "I'd rather be -1100 than -110", Larry Cohen quote from some VuGraph years ago).

My issue is that if you're comparing whether 1NT-AP will beat 1♠-2♠-AP or 1♠-1NT(forcing); 2x-2♠-AP; and using that to determine whether 1NT on 5♠332 scores better or worse than 1♠ when partner has a weak hand, then you're generalizing a lot. These auctions, especially on the hands where there is a fit, get competed against aggressively.

What if the auction would have gone 1NT-2!D on a 4-count (or a zero-count) that would pass 1♠?

As I showed, removing transfer responder hands from the equation boosts the chance that you have a spade fit enormously. Obviously, spades will play better than NT if you're more likely to have a 5=4 fit than a 5=2 fit.

If this is just step 1 of a long set of tests, then sure. But I don't think you can generalize much, except by saying "opening 1NT inhibits finding major fits when responder is weak", which is well known (if worse here, certainly, when we add 5-3 and some 5-4 fits to the 4-4 ones).

[bluenikki]

 mycroft, on 2020-October-28, 15:01, said:

I agree with almost all of that (although competent opps over 1♠-2♠ will only let you play it when they can see it's bad for you or for them, not just when they have a 9-carder. "I'd rather be -1100 than -110", Larry Cohen quote from some VuGraph years ago).

But even Larry Cohen would agree it's better to be -170 than -420.

[Douglas43]

Further to bluenikki's post, thought you might like an angle on "garbage Stayman" (sorry bluenikki) from weak NT-land:

"Supposing partner bids 1 No Trump and I hold:

Q10xx
Jxxx
Jxxx
x

I want to bid 2 Clubs on it, and I don't want partner to bid 2 No Trumps to show his silly maximum."

S J Simon, Design for Bidding. Published 1949 (quote from page 258)

[mycroft]

So, you, too, are assuming that the opponents won't find game unless we help them (or, you're saying they're smart enough to determine if 17 opposite 6 works in this case, but only if they're pushed. Which, I will grant, is the problem with balancing against weaker players - opener says "I have a little more, I'll push", and then responder says "I have a little more, I'll raise to game", and it's cold). If you play in fields where that is right, then sure. But if you play in fields where this is right, then those pairs are unlikely to be wondering whether to bid 1NT with 5 spades.

If you need the opponents to balance to get to those games, too, then I think there's bigger things to work on than "should I open 1NT on 5332 in case we have game in spades but not enough to Stayman?"

For me, if I'm trying to decide how best to play the game to win C, it's I who has the problem, not my bidding system. For others, different priorities apply (and I'm not denigrating them. "Be the best player you can be" is a laudable goal. I won't ever be a world champion, that shouldn't stop me from being the best I can. It's just that that's "win local A" to me, not "win C").

[Vampyr]

 Douglas43, on 2020-October-29, 09:27, said:

Further to bluenikki's post, thought you might like an angle on "garbage Stayman" (sorry bluenikki) from weak NT-land:

"Supposing partner bids 1 No Trump and I hold:

Q10xx
Jxxx
Jxxx
x

I want to bid 2 Clubs on it, and I don't want partner to bid 2 No Trumps to show his silly maximum."

S J Simon, Design for Bidding. Published 1949 (quote from page 258)

Crowhurst may have been popular in 1959, but today it is not. So there is no danger that partner will bid 2NT or anything higher.

[bluenikki]

 Douglas43, on 2020-October-29, 09:27, said:

Further to bluenikki's post, thought you might like an angle on "garbage Stayman" (sorry bluenikki) from weak NT-land:

"Supposing partner bids 1 No Trump and I hold:

Q10xx
Jxxx
Jxxx
x

I want to bid 2 Clubs on it, and I don't want partner to bid 2 No Trumps to show his silly maximum."

S J Simon, Design for Bidding. Published 1949 (quote from page 258)

You are certainly bidding 2♣ whether you play GS or not.

[bluenikki]

 mycroft, on 2020-October-29, 09:58, said:

So, you, too, are assuming that the opponents won't find game unless we help them (or, you're saying they're smart enough to determine if 17 opposite 6 works in this case, but only if they're pushed. Which, I will grant, is the problem with balancing against weaker players - opener says "I have a little more, I'll push", and then responder says "I have a little more, I'll raise to game", and it's cold). If you play in fields where that is right, then sure. But if you play in fields where this is right, then those pairs are unlikely to be wondering whether to bid 1NT with 5 spades.

If you need the opponents to balance to get to those games, too, then I think there's bigger things to work on than "should I open 1NT on 5332 in case we have game in spades but not enough to Stayman?"

For me, if I'm trying to decide how best to play the game to win C, it's I who has the problem, not my bidding system. For others, different priorities apply (and I'm not denigrating them. "Be the best player you can be" is a laudable goal. I won't ever be a world champion, that shouldn't stop me from being the best I can. It's just that that's "win local A" to me, not "win C").

Sorry to keep on with off-topic of (1♠) - P - (2♠) - P ; (P) - ?

I take it for granted that with 11 hcp and fewer than 3 spades, you would already have strained to act over 1♠. Maybe weaker NV. So. given that you are weaker than that, to hope to get out with a small loss at the 3-level, you are presuming significant high cards opposite. And no no no partner is not marked with any such thing.

[Douglas43]

By way of background, in S J Simon's day, the 2C enquiry was still in its infancy. Wikipedia cites a Bridge World article by Sam Stayman in 1945. Simon was criticising the idea, then current, that it should only be used on invitational values or stronger . What is currently being called "garbage stayman" is actually just a fairly recent name for an idea that has been around practically since the convention was invented.

There are other examples, e.g. a recommendation on page 135 of Jeremy Flint's Tiger Bridge published in 1970 to use Stayman with 10xxx, xxxx, Jxx, xx (again as part of Flint-Pender which had a weak NT). Certainly when I learned to play in the 1970's using Stayman as an escape mechanism was considered routine.

Good point from Vampyr about the difference between Stayman and Crowhurst/checkback which follows 1 suit - 1 suit - 1NT; hence there is no need for an escape mechanism and Crowhurst/checkback is always constructive.

Should have added thanks to original poster for opening the thread, you obviously did quite a bit of work on the modelling.

For what it's worth, I often open 1NT with 5 hearts but very seldom with 5 spades. I'd be interested in any model that looked at the difference between the two.

[mycroft]

 bluenikki, on 2020-October-29, 11:09, said:

Sorry to keep on with off-topic of (1♠) - P - (2♠) - P ; (P) - ?

I take it for granted that with 11 hcp and fewer than 3 spades, you would already have strained to act over 1♠. Maybe weaker NV.

Well, of course. Although there may be several hands that just don't have a call over 1♠ (granted, few of them have two or fewer spades).

Quote

So. given that you are weaker than that, to hope to get out with a small loss at the 3-level, you are presuming significant high cards opposite. And no no no partner is not marked with any such thing.

Every hand has 40 HCP, no more, no less. So if partner doesn't have them, who does? Were they dropped on the floor?

Unless you're playing against people who don't look for game with 25-ish points, when responder limits her hand and opener passes, every card I don't have, partner does. I will bid game with any 9 that isn't an 8 opposite a strong NT, and 17 should definitely be inviting opposite the standard 6-9 and known fit, 16 probably should. If they don't invite, *they don't have it*, or *they're bad*.

Similarly, the Law of Total Tricks says, effectively, "if they have a fit, you have a fit". So not only does partner have cards, there's a place to play.

Okay, I'm committing to the 3 level with somewhere between 16 and 23 points; almost always somewhere between 18 and 22. That seems reasonable.

I believe this so strongly, I play OBAR BIDS(*) - so 1♠-p-2♠-X doesn't show values; just shortness in spades and looking for a fit. That's even if they have game values! Because even if they do, they usually can't get 3 of our fit for enough to make it worth going for penalty, and in exchange for trading the trump break to declarer, we get a good handle on an opening lead. I won't try to convince you of the value of this, but I will say that it works, even against National level players, at least as well as not doing it does.

(*) Opponents Bid And Raise, we Balance In Direct Seat.

[pescetom]

 Douglas43, on 2020-October-29, 14:10, said:

By way of background, in S J Simon's day, the 2C enquiry was still in its infancy. Wikipedia cites a Bridge World article by Sam Stayman in 1945. Simon was criticising the idea, then current, that it should only be used on invitational values or stronger . What is currently being called "garbage stayman" is actually just a fairly recent name for an idea that has been around practically since the convention was invented.

As I recall, Simon was a voice in the desert crying for a 3-card reply, whereas Marx used 2NT to show max and Rape'e to show both majors. Of course in this Simon was well ahead of his time - 2NT logically hit the bucket early in weak NT land, but over here a set of replies showing both min/max and one or both majors still survive and were the norm until the last decade.

 Douglas43, on 2020-October-29, 14:10, said:

For what it's worth, I often open 1NT with 5 hearts but very seldom with 5 spades. I'd be interested in any model that looked at the difference between the two.

I agree it would be interesting, although not easy to model. We'll see.

[bluenikki]

 mycroft, on 2020-October-29, 16:02, said:

Well, of course. Although there may be several hands that just don't have a call over 1♠ (granted, few of them have two or fewer spades).

Every hand has 40 HCP, no more, no less. So if partner doesn't have them, who does? Were they dropped on the floor?

Unless you're playing against people who don't look for game with 25-ish points, when responder limits her hand and opener passes, every card I don't have, partner does. I will bid game with any 9 that isn't an 8 opposite a strong NT, and 17 should definitely be inviting opposite the standard 6-9 and known fit, 16 probably should. If they don't invite, *they don't have it*, or *they're bad*.

Similarly, the Law of Total Tricks says, effectively, "if they have a fit, you have a fit". So not only does partner have cards, there's a place to play.

Okay, I'm committing to the 3 level with somewhere between 16 and 23 points; almost always somewhere between 18 and 22. That seems reasonable.

I believe this so strongly, I play OBAR BIDS(*) - so 1♠-p-2♠-X doesn't show values; just shortness in spades and looking for a fit. That's even if they have game values! Because even if they do, they usually can't get 3 of our fit for enough to make it worth going for penalty, and in exchange for trading the trump break to declarer, we get a good handle on an opening lead. I won't try to convince you of the value of this, but I will say that it works, even against National level players, at least as well as not doing it does.

(*) Opponents Bid And Raise, we Balance In Direct Seat.

So they have 23. You have 9....

By the way, the real danger of an adverse game is when responder suppressed a fair 6-card suit to raise the major.

Doubling the 2♠ response directly with singleton is a completely different matter. If I have a singleton in the suit opened, I don't much look at anything else. But I don't wait for opener to limit himself.

[Zelandakh]

 bluenikki, on 2020-October-29, 10:57, said:

You are certainly bidding 2♣ whether you play GS or not.

 Vampyr, on 2020-October-29, 10:12, said:

Crowhurst may have been popular in 1959, but today it is not. So there is no danger that partner will bid 2NT or anything higher.

Maybe not where you live but in some areas (eg France and Germany) it is absolutely standard for the majority of players to respond 2NT with both 4 card majors.

[mycroft]

 bluenikki, on 2020-October-29, 20:24, said:

So they have 23. You have 9....

Partner has 8. We have 17, and a fit somewhere, and all their spade honours aren't worth much. WTP? And it's just as likely that instead of a flat 14 vs 9, it's a 5431 11 vs 6, and partner has 14 with 3 or 4 spades - you want him to double with 4=2=3=4?

Quote

By the way, the real danger of an adverse game is when responder suppressed a fair 6-card suit to raise the major.

Bridge is a game of probabilities. Balancing is dangerous. Sometimes it won't work, sometimes it won't work spectacularly. But letting 1♠-2♠ play is a 35% expectation. Even if you win as often as you lose, you win 60% and lose 30%.

At IMPs, you're expecting -3 or so on average (somewhere between -2 and -6). Now, the push them to game loses 10 and so does 800. So you have to win twice as often as you lose. So you're more careful.

Sure, you can ignore me - what do I know? After all, those than can, do; those that can't, teach; those that can't teach, direct. I just listen to people like Larry Cohen. All I'm saying for this thread (where we're comparing playing 2♠ vs 1NT with 15-17 opposite 4-7 (of course, opener knows that and never invites after 2♠)) is that so many people also listen to Larry that you just won't be playing 2♠ very often.

[Vampyr]

 Zelandakh, on 2020-October-29, 20:36, said:

Maybe not where you live but in some areas (eg France and Germany) it is absolutely standard for the majority of players to respond 2NT with both 4 card majors.

You mean opener’s rebid? Well, you can play Stayman as promising an invitational hand, but I think you are losing a lot. Especially if you play weak NT.

[Zelandakh]

 Vampyr, on 2020-October-29, 23:29, said:

You mean opener’s rebid? Well, you can play Stayman as promising an invitational hand, but I think you are losing a lot. Especially if you play weak NT.

I agree completely...but nonetheless that is what the locals like to do and not promising an invitational hand with a 4 card major is considered unusual.

[nullve]

#
# A similar test but
# * limited to the case when both sides are vulnerable (for simplicity)
# * ignoring cases where opps would do better DD by bidding over 2S (the
#   '(not owdbDDbbo2S)' part of the condition)
# 

nt1517= hcp(north)>=15 and hcp(north)<=17
weak= hcp(south)>=4 and hcp(south)<=7
xfer= spades(south)>=5 or hearts(south)>=5 or clubs(south)>=6 or diamonds(south)>=5
s5 = spades(north)==5 and shape(north, any 5332)
x1N_score = score(vul,x1N,tricks(north,notrumps))
x2S_score = score(vul,x2S,tricks(north,spades))
owdbDDbbo2S = 
(
score(vul,x2N,tricks(east,notrump))>(-1)*x2S_score
or 
score(vul,x2N,tricks(west,notrump))>(-1)*x2S_score
or 
score(vul,x3C,tricks(east,clubs))>(-1)*x2S_score 
or 
score(vul,x3C,tricks(west,clubs))>(-1)*x2S_score 
or 
score(vul,x3D,tricks(east,diamonds))>(-1)*x2S_score 
or 
score(vul,x3D,tricks(west,diamonds))>(-1)*x2S_score 
or 
score(vul,x3H,tricks(east,hearts))>(-1)*x2S_score
or 
score(vul,x3H,tricks(west,hearts))>(-1)*x2S_score
)
produce 100
condition nt1517 and s5 and weak and (not xfer) and (not owdbDDbbo2S)
outcome = x1N_score == x2S_score ? 0 : (x1N_score > x2S_score ? 1 : -1)
action frequency(outcome,-1,1)

10 runs:

Spoiler

1st:

Frequency :
   -1	      44
    0	      12
    1	      44
Generated 232751 hands
Produced 100 hands
Initial random seed 1604044796
Time needed  385.264 sec

2nd:

Frequency :
   -1	      52
    0	       8
    1	      40
Generated 263173 hands
Produced 100 hands
Initial random seed 1604045529
Time needed  406.991 sec

3rd:

Frequency :
   -1	      48
    0	       6
    1	      46
Generated 282961 hands
Produced 100 hands
Initial random seed 1604045964
Time needed  341.173 sec

4th:

Frequency :
   -1	      36
    0	      18
    1	      46
Generated 272414 hands
Produced 100 hands
Initial random seed 1604046425
Time needed  378.308 sec

5th:

Frequency :
   -1	      45
    0	      10
    1	      45
Generated 269294 hands
Produced 100 hands
Initial random seed 1604046887
Time needed  388.281 sec

6th:

Frequency :
   -1	      33
    0	      20
    1	      47
Generated 254589 hands
Produced 100 hands
Initial random seed 1604047312
Time needed  343.974 sec

7th:

Frequency :
   -1	      29
    0	      15
    1	      56
Generated 306356 hands
Produced 100 hands
Initial random seed 1604047675
Time needed  418.061 sec

8th:

Frequency :
   -1	      36
    0	      13
    1	      51
Generated 227907 hands
Produced 100 hands
Initial random seed 1604048114
Time needed  434.995 sec

9th:

Frequency :
   -1	      48
    0	      11
    1	      41
Generated 299655 hands
Produced 100 hands
Initial random seed 1604048632
Time needed  353.994 sec

10th:

Frequency :
   -1	      33
    0	      15
    1	      52
Generated 256778 hands
Produced 100 hands
Initial random seed 1604049024
Time needed  394.311 sec

[pescetom]

 Zelandakh, on 2020-October-30, 00:42, said:

I agree completely...but nonetheless that is what the locals like to do and not promising an invitational hand with a 4 card major is considered unusual.

In italy too... you need to announce that 2C does not promise an invitational hand, whereas they do not need to announce that 2H denies spades. There are also people in the world who play that 2H denies spades and 2S does not deny hearts, come to that: Goren even recommended this in good ol' standard American.