[AL78]

smerriman, on 2022-June-09, 14:05, said:

More common is that 3♠ forces 3NT, with responder then able to show both 1 and 2 suited minor hands.

I'm used to playing 3♠ as showing five spades and four hearts, although I can't rememeber it ever coming up in the last decade, so using it en-route to investigating a minor suit slam sounds reasonable.

[pescetom]

Douglas43, on 2022-June-09, 11:44, said:

3♠ as Stayman for the minors is a reasonable option for seeking a minor suit fit.

smerriman, on 2022-June-09, 14:05, said:

More common is that 3♠ forces 3NT, with responder then able to show both 1 and 2 suited minor hands.

I prefer to have my cake and eat it, 3NT is the response with no 5 card minor over which responder is now able to show both 1 and 2 suited minor hands.

But isn't this a discussion in Novice and Beginner Forum?
If so then the reply of smerriman should be enough IMO, the rest is pretty esoteric.

[johnu]

DavidKok, on 2022-June-09, 02:49, said:

[*]Contrary to what others have said about e.g. 22-opposite-3 games, I will readily raise partner to 3NT with 3-4 HCP opposite 20-21. It's not that I expect it to make, but that I expect 2NT to fail most of the time anyway. We might as well risk the game bonus, since there's no way back to the safe 1NT. Similar to the previous point: if I have a garbage hand and partner has chosen to open 2NT, most of the time we are in a bad position. It's really not clear to me that passing (gambling on 2NT being the limit of the hands) is better than shooting for the stars.

2NT being a bad contract likely to go down in this situation has been a popular thought. There was even an idea that if 2NT was opened and responder passed that the opponent in pass out position should double to collect a bigger penalty since 2NT was likely going down, possibly a lot.

I've done some rough simulations. If responder has 3 HCP (opposite 20-21), you make 8 tricks ~50%, and 9 tricks about ~15% so game prospects are pretty dismal. If responder has 4HCP, you make 8 tricks about 68%, and 9 tricks ~35%. 2NT doesn't seem to be a bad contract opposite 3-4 HCP.

It seems clear that bidding 3NT at matchpoints is a losing strategy since you aren't close to 50% for 9 tricks. Also there is no reason to believe that the field will be in a better contract than 2NT, especially if you have a classic 2NT HCP and shape.

Playing IMPs with 4 HCP, your game prospects are closer to being good enough for game only if vulnerable.

[Cyberyeti]

AL78, on 2022-June-09, 14:14, said:

I'm used to playing 3♠ as showing five spades and four hearts, although I can't rememeber it ever coming up in the last decade, so using it en-route to investigating a minor suit slam sounds reasonable.

Depends on which version of 2N-3♣ you play.

The version we play is 2N-3♣-3M shows 5, 2N-3♣-3N is 2-2 or 2♠3♥ this means if you bid 3♦ over 3♣ there is a known 8 card fit if you're 5-4 as partner has either 3 spades or 4 hearts or both, hence no need for 3♠ to show 5-4.

We use 2N-3♠ to show a slam inv+ hand with one or both minors.

[johnu]

DavidKok, on 2022-June-09, 02:49, said:

[*]In line with the above two, I play any new call over 2NT as forcing to game (including the Jacoby transfers). I'm never contracting for exactly 9 tricks in a suit opposite a big balanced hand, so I prefer to salvage as much as possible of our slam auctions. It's just poor bridge to first open 2NT, preempting partner by two levels, and then try to land on a dime after. Even with my 20-21 range this too much of a gamble.

If you have a method to stop in a suit at the 3 level, you aren't contracting for exactly 9 tricks. You are contracting for 9 tricks or less and if you have less than 9 tricks, hoping that you'll go down less than if you stayed in 2NT.

For common methods, why wouldn't you just pass with a nothing HCP hand after a Jacoby transfer and partner just accepting the transfer? What is the point of bidding a terrible or hopeless game when you can stop a level lower?

[DavidKok]

johnu, on 2022-June-09, 15:08, said:

2NT being a bad contract likely to go down in this situation has been a popular thought. There was even an idea that if 2NT was opened and responder passed that the opponent in pass out position should double to collect a bigger penalty since 2NT was likely going down, possibly a lot.

I've done some rough simulations. If responder has 3 HCP (opposite 20-21), you make 8 tricks ~50%, and 9 tricks about ~15% so game prospects are pretty dismal. If responder has 4HCP, you make 8 tricks about 68%, and 9 tricks ~35%. 2NT doesn't seem to be a bad contract opposite 3-4 HCP.

It seems clear that bidding 3NT at matchpoints is a losing strategy since you aren't close to 50% for 9 tricks. Also there is no reason to believe that the field will be in a better contract than 2NT, especially if you have a classic 2NT HCP and shape.

Playing IMPs with 4 HCP, your game prospects are closer to being good enough for game only if vulnerable.

Defending NT is not that easy, I expect that the blind lead alone lifts these numbers by enough to make game a reasonable gamble, though maybe not at matchpoints. I don't subscribe to the penalty double theory.
The relevant question is (still) not about the 'classic 2NT HCP and shape' hands, but about the hands that get upgraded into the range, or when you play a wider range than I do. My main claim is that that is a losing strategy.

[DavidKok]

johnu, on 2022-June-10, 00:29, said:

If you have a method to stop in a suit at the 3 level, you aren't contracting for exactly 9 tricks. You are contracting for 9 tricks or less and if you have less than 9 tricks, hoping that you'll go down less than if you stayed in 2NT.

For common methods, why wouldn't you just pass with a nothing HCP hand after a Jacoby transfer and partner just accepting the transfer? What is the point of bidding a terrible or hopeless game when you can stop a level lower?

DavidKok, on 2022-June-09, 03:18, said:

I think having a bidding system that caters to getting out at 3M, even at MPs, is a loser in the long run. Kind of like having a system to get out at 3M/4m over 2♣ - if that's where you wanted to go, you should have opened on the 1-level. It wins in the specific cases that you can't make anything (or can make 9 tricks in a major) [..]. Weak-opposite-big-balanced is not frequent enough to disrupt my responding system at the 3-level, and when it does come up it doesn't win enough (even at MPs, there is always the possibility of a defensive slip-up or favourable card position).

I don't play 'Jacoby transfer and partner just accepting the transfer'. Since the transfers are game forcing, partner accepts with a fit, superaccepts with the better half of all fit hands, and bids 3NT on hands with at most a doubleton support. This extra step is very valuable for slam auctions, for example when responder has a non-minimum two-suited hand there is no ambiguity about which suit is trumps.
It loses on hands where we do not have a game. It wins on hands where we have a COG, or possible slam. The latter is, in my experience, more important for the score.

[thepossum]

TylerE, on 2022-June-08, 17:27, said:

I haven't seen 20-22 in anything published in the last 50 years.

They are just ranges I see on some online pages but seem rather high to me

I remember reading somewhere that 2NT is thought by some as a slam killing bid - I think I prefer sometimes 2C followed by 2NT

[AL78]

johnu, on 2022-June-09, 15:08, said:

2NT being a bad contract likely to go down in this situation has been a popular thought. There was even an idea that if 2NT was opened and responder passed that the opponent in pass out position should double to collect a bigger penalty since 2NT was likely going down, possibly a lot.

I've done some rough simulations. If responder has 3 HCP (opposite 20-21), you make 8 tricks ~50%, and 9 tricks about ~15% so game prospects are pretty dismal. If responder has 4HCP, you make 8 tricks about 68%, and 9 tricks ~35%. 2NT doesn't seem to be a bad contract opposite 3-4 HCP.

It seems clear that bidding 3NT at matchpoints is a losing strategy since you aren't close to 50% for 9 tricks. Also there is no reason to believe that the field will be in a better contract than 2NT, especially if you have a classic 2NT HCP and shape.

Playing IMPs with 4 HCP, your game prospects are closer to being good enough for game only if vulnerable.

Do these simulations assume double dummy defence so no-one ever blows a trick leading into the strong hand?

[johnu]

AL78, on 2022-June-10, 13:18, said:

Do these simulations assume double dummy defence so no-one ever blows a trick leading into the strong hand?

The simulations also assume double dummy declarer play, so declarer never misguesses a finesse, always knows the way every suit splits, and in the endgame, always knows exactly what the opponents are holding for possible squeezes and endplays. Same for the defense.

A number of studies have shown that double dummy simulations fairly closely match actual in person results. Feel free to disregard any double dummy simulations. However, if you don't want to rely on simulations, how do you evaluate trick potential?

[AL78]

johnu, on 2022-June-10, 14:24, said:

A number of studies have shown that double dummy simulations fairly closely match actual in person results. Feel free to disregard any double dummy simulations. However, if you don't want to rely on simulations, how do you evaluate trick potential?

I'm a bit sceptical of the application of double dummy simulations, but I don't dismiss them entirely, I'm just careful not to take them too literally. For example, don't use them to give partner an ear bashing. As for their apparent alignment with real time results, I'd be very interested to test that myself using my primary club. There have been times when the double dummy result would have been well below average but I don't know what would happen if I looked at a large number of hands. I might try it out and see over a number of sessions, if a theoretical NS and EW pair got the best double dummy result possible, what would their final score be.

[smerriman]

Richard Pavlicek's data here suggests that actual play results in slightly better than double dummy for everything except slams (and the biggest advantage is in NT contracts), precisely because of that lead factor.

But not enough to boost those percentages for 3NT sufficiently.

[johnu]

DavidKok, on 2022-June-10, 03:58, said:

I don't play 'Jacoby transfer and partner just accepting the transfer'. Since the transfers are game forcing, partner accepts with a fit, superaccepts with the better half of all fit hands, and bids 3NT on hands with at most a doubleton support. This extra step is very valuable for slam auctions, for example when responder has a non-minimum two-suited hand there is no ambiguity about which suit is trumps.
It loses on hands where we do not have a game. It wins on hands where we have a COG, or possible slam. The latter is, in my experience, more important for the score.

If you are going to bid like that, I would accept the transfer with 2 card support, and not accept the transfer with more support. This still gives responder the option of passing the transfer accept with a really bad hand opposite 2 card support. More support for weak responder's suit means more potential tricks, and there is a big difference in expected tricks for 2 card support compared to 3 (and 4) card support.

[AL78]

AL78, on 2022-June-10, 15:17, said:

I'm a bit sceptical of the application of double dummy simulations, but I don't dismiss them entirely, I'm just careful not to take them too literally. For example, don't use them to give partner an ear bashing. As for their apparent alignment with real time results, I'd be very interested to test that myself using my primary club. There have been times when the double dummy result would have been well below average but I don't know what would happen if I looked at a large number of hands. I might try it out and see over a number of sessions, if a theoretical NS and EW pair got the best double dummy result possible, what would their final score be.

I've just had a quick go with one recent session and compared the best double dummy NS result with the other pairs, and this double dummy player would have finished with a little under 47%. The two EW grand slams available which no-one else is bidding, and the EW game on a 4-3 fit which makes because trumps break 3-3 and again, no-one is finding, are some of the boards which really harm the double dummy score.

[smerriman]

AL78, on 2022-June-10, 15:58, said:

The two EW grand slams available which no-one else is bidding, and the EW game on a 4-3 fit which makes because trumps break 3-3 and again, no-one is finding, are some of the boards which really harm the double dummy score.

How would these harm the double dummy score? You're measuring the difference between what you'd score if you got the number of tricks double dummy says you can, and what you actually scored. If your opponents the only ones bidding a grand that makes, the difference is 0 in this case - both double and single dummy give you the same score - so it doesn't harm or help.

(I think you may be mixing it up with the *par* score which is something completely different.)

[AL78]

smerriman, on 2022-June-10, 17:08, said:

How would these harm the double dummy score? You're measuring the difference between what you'd score if you got the number of tricks double dummy says you can, and what you actually scored. If your opponents the only ones bidding a grand that makes, the difference is 0 in this case - both double and single dummy give you the same score - so it doesn't harm or help.

(I think you may be mixing it up with the *par* score which is something completely different.)

I was calculating the best MP score that NS could get given double dummy play both ways on every board. Thus if EW have a vulnerable grand on and the best NS can do is sacrifice for -1400, that is the comparison with the rest of the field, which will be a bottom when no-one else is finding the grand.

Another analysis is to do what I think you thought I was trying to do, which is take a NS player, given what happened at their table, work out what is the best double dummy score they could get.