Sunday, July 22, 2007

Inference from Redundancy, Part III

So far, we have discussed the nuances of two forms of "shape bash" bids. The first was a "12-14" delayed splinter, through 3C. The second was a direct splinter in immediate response to a major opening. The latter (direct splinter) could be used for voids only. Alternatively, the delayed splinter (3C...stiff) and the direct splinter each show three of four holdings (or more):

1. Stiff is not a stiff honor
2. Minimal trump honors, top three (delayed = one top honor, direct = two top trump honors)
3. Suit for normal 2/1 (minor) has only one of the top three honors
4. Fourth suit has no 1st/2nd-round control

The 3NT Direct Raise -- Another Shape Bash?

The next issue is the 1M-P-3NT auction.

This auction has been troubling for me for some time. The typical definition of an auction 1M-P-3NT has a HCP range for 3NT, sometimes the light variety (maybe 13-15) and sometimes the strong variety (maybe 16-18), but no definition of hand type is provided other than "balanced." No discussion of the honors expected is done.

Consider, for instance, the wild difference between these two "13-15" hands, in response to a One Spade opening:
.
1. Axx QJx QJxx QJx
2. Jxx Axx Axxx Axx
.
These two hands seem to be wildly different, of course. Imagine a partner holding KQxxx Kxxxx Kx x. Opposite the first, partner cannot make more than 4S. Opposite the second, the slam is very strong odds to make. How, precisely, does Opener explore his options, if 3NT says nothing about strength?

I have forced partners to pick one or the other -- Aces and spaces OR quacks -- as the "honor expectancy" of this call. Either might make sense, but playing that you might have one, the other, or a blend of these seems unplayable.

In any event, whatever you select tailors cuebidding that does not elect that option. Discuss partner's preference, and then elect a style.

If 3NT shows Aces-and-spaces, then cuebidding sequences will not ever show a 4333 3-card range with Aces-and-spaces. If 3NT shows Quacks, then cuebidding sequences will not show a balanced quack hand.

No comments: