[whereagles]

Hi there,

Was thinking.. anyone playing Keri over 1NT rebids, as a substitute for checkback 'n stuff? It could be something like keeping Keri active, and making inconsequent bids as natural. Example

1C 1H
1NT ...

2C = Keri relay
2D = transf
2H = transf. Must be 44 maj invitational, otherwise reply wouldn't be 1H
2S = range inquiry. Subsequent non-heart 1-suiter bids become natural, not RKCB
2NT = transf
3x = splinter, with 3D (originally heart splinter) now changed to natural, i.e. a 55 invitational, inv since a gf 55 would be bid via transf.

Keri seems superior to checkback, although one could perhaps simplify the structure considerably.

[1eyedjack]

Dunno about Keri, but I agree that there are better alternatives to checkback.

I play that 2D rebid by responder is an artificial one round force showing one of 3 hand types:
(1) the first move on any invitational hand,
(2) a game forcing hand whose only interest is identifying major suit fits for the purposes of choosing between 3N/4M,
(3) the first move on a balanced slam try relay sequence.

This is quite efficient because openers' rebids generally provide useful information opposite any of the 3 hand types that responder might hold. The only exception is that opener distinguishes between min v max when this distinction is of no interest to responder with hand type (2). But even that is not all bad news. An opponent with a marginal double of the final game contract might wish to double if responder had hand type (1) but not if he had hand type (2), and he never gets to find out which type responder has, so he is doubling in the dark. Furthermore, the approach to defence on opening lead (ie active v passive) may well depend on which of hand types (1) (2) or (3) are held by responder, and usually there remains ambiguity.

This use of 2D is not suitable if you are playing a wide range 1NT rebid, but that I think is a flawed method anyway, and the inability to use this continuation adds (in my opinion) a further nail in the coffin of the wide range rebid.

Most of the other rebids by responder are puppets, that contain either weak or GF hand types. It is particularly useful to reserve 2C as a puppet rather than an enquiry bid, because the number of continuations that end in 3NT but start with 2C are so much greater than via other starting points (Fibonacci and all that).

This approach enables a GF responder to show his precise shape, with virtually any shape, at a reasonable level.

[EricK]

This is a reasonable idea.

I would suggest you tailor your responses based on the answers to these two question.

How often will partner raise to 2M on 3 card support instead of bidding 1NT? (For some pairs the answer is never, for others it is whenever he has a small doubleton, for others it is if he has no tenaces to protect, and so on)

How often will partner rebid 1NT with a singleton in your suit? (Again some pairs will never do this, others will do it sometimes, others will do it frequently).

Depending on your answers, you might decide you have little need to rebid a 5 card suit (if partner is unlikley to have 3) or unwise to do so (if partner may have a singleton), so the second round transfer would suggest a six card suit.

For the second question (or possibly in any case), you might want to have a bid to show that your suit is weak, so he shouldn't rely on it to be a stop in 3NT - maybe you could use the "impossible" splinter (eg 1♣ 1♥ 1NT 3♦) to show that. I don't know whether telling opps about this is a good idea though!

Eric

[whereagles]

I cooked up a complete scheme for Keri over 1NT rebids. You basically use the Keri structure, allowing for some logical deviations. Here's a couple examples:

1m 1M
1N ...

2C = keri relay. Followed by 3m is invite, but doesn't need 6 cards now
2D/H = transfers. Followed by 2NT is 54, followed by 3m is 55 forcing (as in normal Keri)
2S = range ask. Followed by 3m/3M is RKCB for m/M (as in normal Keri)
2NT = transfer to clubs. Multipurpose
3x = splinter on suit above. Splinter to M is natural 55 invitational

Coping with redundancies after 1H:

1m 1H
1NT 2C
2D 2H = invite, 5 hearts. May have 4 side spades if scattered values

1m 1H
1NT 2C
2D 2S = invite with 44. Concentrated values

1m 1H
1NT 2D
2H 2S = invite with 45. Concentrated values

1m 1H
1NT 2H (xfer spades) = invite with 44, scattered values. Non-forcing

Redundancies after 1S:

1m 1S
1NT 2C
2D 2H = invite with 54, non-forcing

1m 1S
1NT 2D (xfer hearts) = 55 major. Followed by 2S is NF

As you see, using the extra information of the 1M reply to 1M allows for more precise invites.

I also devised a scheme after 2NT rebids. I'll post it here if someone wants it.

[luke warm]

1m 1H
1NT 2C
2D 2H = invite, 5 hearts. May have 4 side spades if scattered values

1m 1H
1NT 2C
2D 2S = invite with 44. Concentrated values

1m 1H
1NT 2D
2H 2S = invite with 45. Concentrated values

1m 1S
1NT 2C
2D 2H = invite with 54, non-forcing

i don't think these are any different from 2 way ckback, are they?

1m 1H
1NT 2H (xfer spades) = invite with 44, scattered values. Non-forcing

i don't like this, but that's because i prefer the 2D/1NT bid to be a game force (no xfer to hearts for a pass)

this is sorta like the thread on best defensive treatments over a 1NT opening in that the only real way to compare is to actually post 2 or 3 hands

[laughter]

Keri can be readily adapted to 1NT rebids.
I have tried it in my spare time, it works out ok.
But you'd better clarify your continuations in the following sequences:
1) 1m-1M; 1NT-2C; 2D-2NT:
Imo, reasonable to play it as balanced FG, probably slam interest (similiar to Keri).
However, you get to redesign the rebids as opener cannot have 5 card major. Since responder is balanced and may be interested in a minor, plus opener has already described some of his minor suit distribution by his choice of opening, I suggest the rebids to put more focuses on minors.
2) 1m-1H; 1NT-2C; 2H:
Are you sure that you want to play it as NF transfer to S, showing 4-4 and scattered values? It is more than 3-leg parlay (4S+4H+inv+scattered values) and its occurance must be rare. Much better, imo, to play it as transfer to spades, denying fifth H, yet unlimited (inv+). The advantage is you can use it to create FG (and try for slam) with 4=4-(3-2), allowing the partnership to uncover minor fit at 3 level if necessary. Btw, if you would force with 2H with 4S, then the Keri 2NT should deny interest in spades and 3S by both partners can be artificially employed.
3) 1m-1M; 1NT-2S:
You can't have a hand that want to ask for Keycard in other suits.

[whereagles]

1. Indeed. A possible way out is

1m 1M
1NT 2C
2D 2NT. The 2NT is usually a minor suit slam try or probing for best game with a good 5-card major
3C = 4432 or 5332, min or max (need to stay low here, in case resp has 5 in major)
3D = 4333 min
3H = 4333 max
3S = nat 4 spades and 2 hearts, max

But, of course, you can try many other schemes here.


2. In the auction

1m 1H
1NT 2H

I proposed 2H as 44 invitational but NF so as to allow opener to play in the moysian 2H fit if he has a 2344 or 2335. But yes, you can use it to show a forcing 44 if you prefer.


3. On most natural systems, the 1m opener is on 3 cards if m=clubs, 4 cards if diamonds (with 4432 shape opener would have supported hearts). Thus

1m 1M
1NT 2S
2NT 3m = can be used as RKCB for the minor. P.ex. opposite 1C opener with Axx KQxx x Axxxx

1m 1M
1NT 2S
2NT 3-other minor or 3-other major can be used as suit asking bid with M as trumps. Later responder can ask for aces with RKCB 3NT or 4NT

[1eyedjack]

It just seems to me to be fundamentally unlikely that a system that is designed (whether successfully or otherwise) to be optimised for responding to a 1NT opening bid should also by coincidence be optimised for continuing the auction after a 1NT rebid after both partners have bid a suit prior to the occasion of the 1NT bid.

[EricK]

1eyedjack, on Aug 30 2004, 06:43 PM, said:

It just seems to me to be fundamentally unlikely that a system that is designed (whether successfully or otherwise) to be optimised for responding to a 1NT opening bid should also by coincidence be optimised for continuing the auction after a 1NT rebid after both partners have bid a suit prior to the occasion of the 1NT bid.

But an interesting phenomenon I have observed is that many partnerships have better, more complete, agreements after 1NT than they do after 1x 1y 1NT.

If those partnerships were at least to use there exact same methods after 1x 1y 1NT as they do over 1NT then there bidding would automatically improve.

Eric

[whereagles]

That is actually why I thought of using Keri on various circumstances, not just after 1NT openers