nullve

http://www.infobridge.it/Convenzioni_Avarelli.htm "This convention is extremely useful when the asking partner holds a void. In its original form it was used in well-defined and precise bidding situations as established by the bidding system."

1♠-2♥; 2♠ is probably the most interesting start from Responder's point of view.

Yes. Nice trick question, Lovera!

Why not start by playing a low from dummy? RHO might not be able to calmly play low from Kx.

Some play Transfer lenbensohl over (2x[=PRE])-X-(P). Advancer can then show his suits more cheaply on slammish hands at the cost of no longer being able to invite in clubs.

Opps' bidding has revealed that partner has exactly 3 H and therefore most likely 4342 or 4351 distribution. With the latter he might be able to find another call (e.g. make an "action" double, if we play that) if I pass, so I'll assume he is 4342. Then there are 17 total trumps, but most likely only 16 total tricks due to mirror distribution. So Pass seems super-clear, at this and every other vulnerability.

Haha, yes.

If LoTT is accurate on this board, then with 17 total trumps (instead of 16, as here), 4♥ is only right if total tricks are divided 10-7, 12-5 or 13-4 in our favour. So P is likely best even then. Also, the 3433 shape and the club holding tend to bring the trick total down, so P is obvious.

It makes a huge difference whether 1♣-(X)-1♦-(1♠); 2♣ promises 6+ C or 5+ C, as anyone who has struggled with classical Precision's 2♣ opening (showing 11-15 hcp and either 6+ C or 5C4M) can attest. There simply has to be a way for partner to know whether 2♣ could be on a 5-1 fit (usually terrible) or a 6-1 fit (often best). Pass is not timid, since partner will know approximately what you have and be able to bid 1N himself, even without a spade stopper. Some even pass systemically with 12-14 bal., 2 H, here so that they can use 1N stay low with 18-19 bal., 2 H. If you play Walsh over 1♣-(X), then there's no need for 1♣-(X)-1♦-(1♠); X to show 4 H, since partner won't have 4 H unless he can can force to game. So playing X as support, or just plain takeout (which it should be by default anyway), seems reasonable. If you don't play Walsh, then you might have a choice between P and 1N, depending on how you play 1N (12-14 vs. 18-19). (2♣ would again promise 6+ C.)

3♥ with only 4 H hardly makes sense unless it's forcing and XX denied 5+ H.

Your main point, which I believe was , is valid and I can't imagine we really disagree on anything here.

Which includes the auction and the play so far. That's why your probabilities must be taken with a huge grain of salt. For example, is there a 5 % chance that RHO has a diamond void? I know GiB's overcalling style vs. 1N is extremely conservative in some ways, but against human experts, at least, I'm sure that would almost never happen.

Those are a priori probabilities. And opps didn't bid.

Agree with Mr Ace, but ♠K at trick 2, followed by the ♦K if it holds, is also interesting.

Partner has just contradicted himself, so it could be anything.

A CC from 2016: http://www.ecatsbridge.com/documents/files/BPinfo/OpenTeams/Norway/NOROT_Brogeland-Lindqvist.pdf

Recently I've been toying with something similar against 2-level preempts, e.g. (2♠)-?: X = takeout, but 15-17 or 21+ (instead of 18/19+) if bal. without tolerance for all unbid suits 2N = bad C overcall / 18-20 bal. 3♣ = good C overcall / bad D overcall 3♦ = good D overcall / bad H overcall 3♥ = good H overcall / ? (....), which makes the overcall ranges more manageable as long as Advancer can cope with not knowing partner's suit with absolute certainty to begin with.

That's more or less correct. Some pairs even play two kinds of CoG over 1N-[2M-1]; 2M, don't they? Maybe one with 5M(332) and one with 5M4m22, or might they also bid this way with a small singleton? Yes, looks like the most logical choice in an otherwise standard structure. Yes. Yes. Looks like a clear improvement.

That's a good point which makes me regret the title of this thread. The idea was not so much to free up the 3N rebids themselves as freeing up bidding space by removing the hand types associated with those bids in an otherwise standard rebid structure. In a structure with 2nd round transfers, maybe the 3♠ rebids ("transfer to notrump"?) would be affected (if not entirely freed up) instead. Since I certainly didn't want to rule out non-standard rebid structures, I shouldn't have been talking about freeing up 3N rebids at all. I realise that my first paragraph sounded like the whole purpose of my Puppet 2N variation is to free up those 3N rebids. But it can also be played for the usual reasons, e.g as a low information alternative to regular Stayman. I don't have anythng very concrete in mind, but I've been trying to come up with decent relay structures, both over 1N-2♣; 2x and 1N-[2M-1]; 2M, using all of Responder's rebids starting from 3♣1. Since available bidding space is scarce, it really helps to be able to respond 2N with on more hands than I, at least, have been doing so far, which have just been balanced hands on which Responder might have some interest in 4M. I like Zelandakh's suggestions in an otherwise standard rebid structure. 1 14.12.16: Ok, here's something semi-concrete:

Idea: Use a Puppet Stayman-like 2N response to 1N * with 5 M in a (semi)balanced hand to free up Responder's 3N rebid over 1N-[2M-1](Jacoby xfer); 2M; * with 4S4H to free up 3N over 1N-2♣(Stayman); 2♦1 ; * to find the 5-3 OM fit when Opener has 2M5OM and Responder has GF values and 5M3OM(32). 1 3N over 1N-2♣; 2♦-3N shows 4S4H the way many play (regular) Puppet 2N/♣, because they would have used Puppet 2N/3♣ with 4M3-OM and they wouldn't have used Stayman at all with 3-S3-H. [Edit: I'm being a bit stupid here, because the reason they don't use Puppet with 4S4H is right-siding issues after e.g. 1N-3♣(Puppet); 3♦-3♥ = 4 S (likely wrong-siding 4♥ if Responder has 4S4H).] What I've come up with so far: 1N-2N = GF with a (semi)balanced hand suitable for either Puppet Stayman (2N or 3♣) or 1N-2[M-1]; 2M-3N playing normal methods / ? 1N-2N; ?: Either a) (too much information leakage?) 3♣ = 4-S4-H, likely ruffing value ...3M-1 = 4+M3-OM ......3M = 3M, likely ruffing value .........3N = 4 M, NF ............P = 3 M ............4M = 4 M .........(...) .........4M = 5 M, to play .........(...) ......3N = 2 M ......4M-1 = 4 M, likely ruffing value, wants p to declare 4M ......4M = 4 M, likely ruffing value, wants to declare 4M ...3♠ = 3-S3-H / ? ...3N = 4S4H, NF ......4M-2 = wants p to play 4M ......4M = wants to declare 4M ...(...) 3M-1 = 5 M ...3M = 2M5OM ......3N = 2 OM or no likely ruffing value ......4OM = to play ......others: undefined ...(...) ...3N = to play ...(...) ...4M-1 = wants p to play 4M ...4M = wants to declare 4M ...(...) 3N = 4-S4-H, no likely ruffing value ohters: undefined or b) (less information leakage, changes from a) highlighted in blue) 3♣ = 4-S4-H, likely ruffing value / [meaning shared with 3N:] 4-S4-H, no likely ruffing value ...3M-1 = 4+M3-OM ......3M = 3M, likely ruffing value / [meaning shared with 4M:] 4 M, likely ruffing value, wants to declare 4M .........3N = 4 M, NF ............P = 3 M ............4M = 4 M .........(...) .........4M = 5 M, to play .........(...) ......3N = 2 M, or 3 M and no likely ruffing value ......4M-1 = 4 M, likely ruffing value, wants p to declare 4M ......4M = [meaning shared with 3M:] 4 M, likely ruffing value, wants to declare 4M ...3♠ = 3-S3-H / ? ...3N = 4S4H, NF ......4M-2 = wants p to play 4M ......4M = to play ...(...) 3M-1 = 5 M ...3M = 2M5OM ......3N = 2 OM or no likely ruffing value ......4OM = to play ......others: undefined ...(...) ...3N = to play ...(...) ...4M-1 = wants p to declare 4M ...4M = wants to declare 4M ...(...) 3N = [meaning shared with 3♣:] 4-S4-H, no likely ruffing value ohters: undefined. The added idea here is that Opener will try to randomise (in a real or fake way, subject to full disclosure) between * 3♣ and 3N over 1N-2N, with no likely ruffing value * 3M and 4M over 1N-2N; 3♣-[3M-1], with 4 M, a likely ruffing value and willingness to declare 4M, to keep opps as much in the dark about his M length as possible after 1N-2N; 3♣-[3M-1]; 3M/3N.

If unbalanced hands have to meet the rule of 31 to qualify for a 2♣ opening, then the vast majority of hands that qualify will either be balanced or fall into a 3-point range for any given shape. My suggestion is to try to take advantage of this by pretending that stronger unbalanced hands don't exist, as in the following scheme over 2♣-2♦: 2♥ = Kokish: rules of 31-33, 5+ H, unbal. / 25+ bal. ...2♠ = expected ......2N = 25+ bal. ......3♣+ = rules of 31-33, 5+ H, unbal., nat. 2♠ = rules of 31-33, 5+ S, unbal. 2N = 22-24 bal. 3♣+ = rules of 31-33, otherwise standard (On unbalanced rule of 34+ hands Opener will have to improvise. But at least he didn't have to open either 1-of-a-suit, 2N or 3N, as in EHAA!) The point is that since Opener's strength will soon enough basically be known to within a 3-point range, there's hardly any need for Responder to say anything about his strength in response to 2♣.

A transfer to clubs, I think. (Can be read in two ways, I know.)

I wonder: Since * 1N-3♠ = GF, 1-S5+D5+C * 1N-3♦ = GF, 1-S3-H6+D4-C * 1N-2N; 3♣-3♠ = GF, 13(54), does this mean that * 1N-2♣; 2♠-3♦ = GF, 1-S4H5+D instead of just 1N-2♣; 2♠-3♦ = GF, 1-S4-H5+D? If so, then it seems to me (right now) that you can play 1N-2♣; ?: 2♦ = no major 2♥ = hearts, denies 4 S 2♠ = spades, does NOT deny 4 H 2N+ = undefined and still never lose a 4-4 major fit. As a bonus you can sign off in C via 1N-2♣; 2x-2N.

You respond 2N to Stayman with 4S4H just so you won't lose a 4-4 S fit because 3♠ over 1N-2♣; 2♥-3♣; 3♦-3♥ is Last Train for D, right?