[pescetom]

I never thought it was a sound system decision to open 1NT containing a 5cM while retaining a traditional 4cM Stayman Ask.
But if we must, surely the opener should exercise some judgement before playing 3NT with majors 5-3 and xx in a minor?
Curious to know if a significant number simulations really supported this choice (and if the East hand was excluded because GiB would bid Cappelletti).

[johnu]

Well you win some by concealing a 5 card major and not looking for one, and lose some like in your example. In real life, a lot of top experts have added some sort of 5 card major ask.

Opposite a responding unbalanced hand with a 5-3 fit, my quick and dirty simulations show that the 5-3 fit is substantially better than NT. Opposite a balanced hand with a 5-3 fit, the 5-3 fit is a little better at IMPs, and substantially better at MPs.

[mycroft]

Yes, a lot have embraced 3♣ GF 5-card ask - with no distinction between "no 5 card" and "no 4 card". And they use it almost systematically, so that the information passed about dummy's hand is low as well.

I have such an ask in my non-Stayman system as well; I should use it more just so the information passed when we use it as GF 4-card Stayman isn't as high (even though that partner is almost allergic to 5cM in 1NT).

[pescetom]

mycroft, on 2026-February-16, 09:16, said:

Yes, a lot have embraced 3♣ GF 5-card ask - with no distinction between "no 5 card" and "no 4 card". And they use it almost systematically, so that the information passed about dummy's hand is low as well.

I have such an ask in my non-Stayman system as well; I should use it more just so the information passed when we use it as GF 4-card Stayman isn't as high (even though that partner is almost allergic to 5cM in 1NT).

Plenty of other ways of skinning the 5 card cat too. I play 2♣ as a pure 5 card ask, 2♦ denies... minimal leakage about both hands (and we might already be exactly where Responder wanted to play). Many others here play 2♣ as an inv+ 5 card ask with 2♦ promising one or both 4cM and 2NT denying, which is more generous to opponents.
The important thing is to have some method, it cannot be wise to have only a 4 card ask when 1NT systematically contains 5M332, although I guess you can get away with it if most of the field is similarly hampered.

Be that as it may, my question is more about whether the robot did well to pick 3NT here: it's method of "thinking" by simulation should lead to a sound decision, but this one looks odd to me.

[mycroft]

But again, the issue is not "look how much we're losing when we open 1NT with 5M332"; it's "look how much *more* we're winning when we open 1M which will *not be* 5M332 in NT range". How much better 1♥-1♠; 1NT (or whatever other call you overload with 16 HCP 5M332s) auctions are. How much easier it is to respond to 1♠-1NT Forcing. Or 1♠-2m.

And I played 2♣ Puppet for a few years when I first started. The losses we got from that, specifically on Garbage Stayman and weak 5-4M hands were just brutal. The bonuses we got when we *did* find the 5-3 fit, sure. But it wasn't worth it. Now with the 3♣ "check for 5cM" bid, we at least reduce our losses to partscore hands (which at IMPs is a big source-of-loss reducer; at MPs, well, a serious minority at least).

The day we invent a perfect bidding system I will cheer. And then probably give up the game as boring. Of course, I expect to do that the week after Hamman gives up playing after his perfect session.

[pescetom]

mycroft, on 2026-February-16, 10:01, said:

But again, the issue is not "look how much we're losing when we open 1NT with 5M332"; it's "look how much *more* we're winning when we open 1M which will *not be* 5M332 in NT range". How much better 1♥-1♠; 1NT (or whatever other call you overload with 16 HCP 5M332s) auctions are. How much easier it is to respond to 1♠-1NT Forcing. Or 1♠-2m.

Sure, did anyone question it? You don't need to convince me, at any rate: I simulated it to death a few years ago and convinced myself that it was an all round winner. And I play enough of both (unfortunately) to be able to compare.

mycroft, on 2026-February-16, 10:01, said:

And I played 2♣ Puppet for a few years when I first started. The losses we got from that, specifically on Garbage Stayman and weak 5-4M hands were just brutal. The bonuses we got when we *did* find the 5-3 fit, sure. But it wasn't worth it. Now with the 3♣ "check for 5cM" bid, we at least reduce our losses to partscore hands (which at IMPs is a big source-of-loss reducer; at MPs, well, a serious minority at least).

I imagine that by "Puppet" you mean the convention that I outlined others here playing, basically a transposition of what most people play as "Puppet" over a 2NT opening, right? I agree that is poor, for a wealth of reasons. Because I really struggle to believe you got brutal losses from playing a true Puppet Stayman, where 2♦ denies a 5cM (if Opener can have one) and says nothing else. Specifically, it offers a much *more* effective Garbage pass, which offsets the losses with a weak 5-4 forced to play in the wrong 2M if the competition is playing Crawling (admittedly, not much of the competition here does).

[johnu]

pescetom, on 2026-February-16, 09:47, said:

Plenty of other ways of skinning the 5 card cat too. I play 2♣ as a pure 5 card ask, 2♦ denies... minimal leakage about both hands (and we might already be exactly where Responder wanted to play).

This is the original method written up by Kit Woolsey in the Bridge World.

[johnu]

mycroft, on 2026-February-16, 10:01, said:

And I played 2♣ Puppet for a few years when I first started. The losses we got from that, specifically on Garbage Stayman and weak 5-4M hands were just brutal.

For those who haven't played Puppet Stayman, it has its own version of Garbage Stayman. Basically just pass the 2♦, 2♥ and 2♠ responses. You need 5+ diamonds, 2+ in both majors so you are in at least a 5-2 fit, and could have better.

With a weak major 5-4, just transfer to the 5 card major. You'll miss a potential 4-4 (or even a 5-4) major suit fit, but you'll get to a 5-2 or 5-3 fit instead of maybe a 4-3 fit.