awm, on 2017-November-08, 02:37, said:

It seems like nullve's system will play 1NT on a lot of hands where 2m is better (i.e. 5314 11-count opposite 1345 6-count, to give an extreme example). This will bleed quite a few IMPs over time.

Are you sure about this? Assuming there's a two-trick difference between a notrump and a club contract, that LoTT holds and that opps will compete to 3

♦ over 3

♣ (to make things interesting), then e.g.

* if I get 5 tricks in 1N (-100 NV, -200 V), opps will get 10 tricks in 3

♦ (-130), so a gain/loss of 1/2 IMPs, respectively

* if I get 6 tricks in 1N (-50 NV, -100 V), opps will get 9 tricks in 3

♦ (-110), so a gain of 2/0 IMPs, respectively.

* If I get 7 tricks in 1N (+90), opps will get 8 tricks in 3

♦ (+50 NV, +100 V), so a gain of 1/0 IMPs, respectively.

* If I get 8 tricks in 1N (+120), opps will get 7 tricks in 3

♦ (+100 NV, +200 V), so a gain/loss of 1/2 IMP, respectively.

awm, on 2017-November-08, 02:37, said:

Further, you can't play in 2m really; after 1♠-1NT-2♦! you will have to go to the three-level in a minor, with the worst case being something like 5314 14-count opposite 1453 6-count (where you're at the three-level in a seven card fit). Another gradual bleed of IMPs.

Responder may pass 2

♦ with 6+ D.

You're right about 14 hcp, 5314 opposite 6 hcp, 1453 being a problem for me (and I know I'd pass with Responder's hand

). But problems would exist also if

* 1

♠-1N; 2

♣ = 4+ C or 5(32)3 and 1

♠-1N; 2

♣-2

♦ = NAT (how can Responder know it's right to pass/bid 2

♦?)

* 1

♠-1N; 2

♣ = 4+ C or 5(32)3 and 1

♠-1N; 2

♣-2

♦ = Bart or similar (Responder is forced to pass, possibly violating Burn's law)

* 1

♠-1N; 2

♣ = standard Gazzilli (Responder will rebid 2N).

awm, on 2017-November-08, 02:37, said:

Okay, what about the invitational hands? After 1♠-1NT-2♦, you need to be able to distinguish all of:

1. Lousy hand with a boatload of clubs

2. Lousy hand that just wants to play opener's minor.

3. Mildly forward-going hand with a boatload of clubs.

4. Mildly forward-going hand that wants to play opener's minor unless he's REALLY max (it seems like 2♦ is around 13-15, so this is like some 10-11 count).

5. Hands that are now game-forcing (i.e. 12-13 high) but cannot yet place the contract without knowing more about opener's shape.

That's a LOT of hand types and four of them need to be communicated by the time we reach 3♣. The only ways I can see to do it involve using 2♥ as an artificial force of some kind (and probably ALSO putting one of the "boatload of clubs" hands into an immediate jump shift). But this makes it a lot harder to find heart fits where responder has five and opener has three, which is easy in my methods (for example) because 1♠-1NT-2m-2♥ shows five hearts and constructive values.

Responder has no way of showing a mildly forward-going hand. So with 3. or 4. Responder must make an immediate decision whether to play game opposite MAX. (If 2

♦ doesn't include MIN, 5M5m, then 2

♦ already shows MAX.)

"Boatloads of clubs" hands are usually not a problem unless they contain 1- S. Then passing(!!) 2

♦ (instead of bidding 3

♣, P/C) with something like 5-9, 1336, hoping that Opener has diamonds, might be Responder's best bet, although it could fail spectacularly.

The 12-13 hands are hardly a problem if 2

♦ promises MAX, because there's enough space for almost everything after 1

♠-1N; 2

♦-2N(GF relay).

awm, on 2017-November-08, 02:37, said:

There are also hands where you really want to know opener's shortness/shape when he has a six-card major. For example, say I have some 12-count and the auction starts 1♠-1NT-2♥!; great I know we have game values now. But presumably 2♠ is to play, 2NT is like 10-11 (opposite the presumed 13-15), 3m might be a long suit of my own? It doesn't seem like I have an easy asking bid here. In my system these auctions time out a bit different and we're more likely to be able to communicate this information.

The idea (with a view to practical bridge where 3M-1 instead of 2M= is bad and information leakage costs tricks) is that Responder never invites after Opener's 2M-1 or 2M rebid, although I can't see why inviting or asking for a singleton should be more difficult after the 2M-1 rebid than after a standard 2M rebid, which has a wider range. Am I missing something?