[straube]

We open a nebulous diamond. After 1D-1M we use a direct raise as a bad raise (possibly 3cd with shortness) and 1-under as a medium raise (also possibly 3 cd with shortness). We're devising our jump raise structure.

One idea (awm's actually) is...

1D-1H,
.....2S (unspecified void)
..........2N-asks void
.....2N-splinter-spade shortness
.....3C-5/5 minors
.....3D-splinter-diamond shortness
.....3H-splinter-club shortness

We're also concerned about helping the opponents unnecessarily. Partner suggests...

1D - 1H - 2S (unspecified shortness)

3C: Wants to bid game opposite high / mid short
3D: High short
3H: Mid short

3D: Wants to go opposite low shortness

3H: Sign off

Any thoughts?

[awm]

I think you somewhat overrate the value of this information to the defense. In particular, it's dummy revealing the shape and not declarer. And your method gives more information in a number of sequences. For example compare:

1♦ - 1♥
3♥(1) - Pass

(1) Club shortage

Here we don't know whether responder has a terrible hand, or has a decent hand with club wastage.

1♦ - 1♥
2♠(1) - 3♣(2)
3♥(3) - Pass

(1) Raise with shortage
(2) Game interest opposite some shortage
(3) Shows a specific shortage

Here we know responder has values, we know opener's shortage, and we likely know where responder "wanted" opener to have shortage.

1♦ - 1♥
2♠(1) - 3♥(2)

(1) Some shortage
(2) No game interest regardless

Here the defense knows declarer has a lousy hand, and might even find a speculative double.

I guess the point is, you exchange information in some sequences for information in others. But the method where opener shows shortage gives information about the dummy and discloses less about declarer's hand (which will be hidden in the play anyway). It also makes good use of bids (what else will you use your 3♥ jump raise for?)

[dake50]

Had this same discussion about mini-splinters vis-a-vis "full" splinters.
Decided raise has the "expected" mini-splinter.
An "other" short begins with catchall bid, then show if asked.
Delayed splinters begin 2/1 and show if non-minimum rebid from partner.
I think this is near optimum for blab/info.
Show most likely to promote partner's hand.
Let other shapes begin with GF: "do you want to show?"
"Or do you want 'no blab'?" input from partner.

[straube]

awm, good point about the difference between revealing declarer's and dummy's hand. I like your structure as written.

I read in your notes (hope you don't mind my discussing them here) that you use...

1D-1S, 2H-2N (GF)

.....3C-all with 3-cd support
.....3D-4, higher
.....3H-4, middle
.....3S-4, lower
.....3N-4, bal

I like the idea of using the 3C bid as balanced. This allows...
.....3D-RKC (which comes up a lot)
.....3H-higher shortness (responder showing his own shortness)
.....3S-middle shortness
.....3N-lower shortness

or possibly something even better. The thing is that the balanced hand occurs so frequently and we want to preserve space.

I know it's not optimum, but we use...

.....3C-bal
..........continuations as discussed
.....3D-3-1-4-5
.....3H-3-1-5-4
.....3S-3-4-1-5
.....3N-3-4-5-1
.....4C-4, higher
.....4D-4, middle
.....4H-4, lower

I noticed that you seldom inquire as to the full shape of the 3-cd support.

We continue queen point ask after the 3-level shapes.

After the 4-level shortness showing bids, we have only RKC or cue bids. I think this isn't that bad though because responder will have a very good idea as
to opener's strength and support.

Also noticed your rules as to when to make a simple raise, when to make a 1-under, and when to make a splinter bid. Could you translate that into Modern LTC for us?
I was thinking...

bad raise- 7 or more losers
good raise-6 or 6 1/2 losers
splinter raise-5 1/2 or fewer losers

That about right?

Finally, the only game try we're using is second suit game try.

After 1D-1S, 2H-3C shows a second suit and hopes to improve the contract from a possible 4-3 fit. 1D-1S, 2H-3C, 3D would suggest a 3-4-5-1.

I think this will seldom improve the contract, but I think I prefer it to short suit game tries. For us 1D-1H, 2H-2S is also a second suit GT and we can stop there.

[awm]

Our formal definition is something like "better than a balanced 13-count." You're roughly right on the loser counts (i.e. 7 or more for the single raise, 6 for the "good raise" and 5 for the jump raise) although this isn't precisely how we count our values. However, we tend to downgrade a bit (say about half a loser) when holding only three-card support.

There are two main reasons our responses to the asking bid are as described. First, our structure is symmetric with the jump raises, so for example 1♦-1♠-3♦ shows four card support with heart shortness, and 1♦-1♠-2♥-2NT-3♦ also shows four card support with heart shortness. This is nice for memory purposes. Second, the balanced hands are actually extremely rare, because the 2♥ good raise is "better than a balanced 13-count" ... occasionally you have some max balanced hand with two doubletons and/or extremely working values that's worth the upgrade, but these are sufficiently rare that we didn't prioritize them in our structure.