[straube]

So we open an artificial 1D and we play support doubles. I'm thinking, however, that takeout doubles are better.

1D P 1H (2D)

and I hold

Axx Kxx Kxx Qxxx

not sure why I want to show support here. Pass it around and partner might have

KQxx Axxx xx Kxx and he can double back in and we're at an awful 2H anyway. But with

KQxx Axxx xxx Kx maybe he's passing because he knows that I have diamonds.

But if my double is takeout and not support and I have

Axxx Kx xx KQxxx I can double with this showing shortness, 2 or 3 hearts, 3 or 4 spades and 4+ clubs and partner with

KQxx Axxx xxx xx

will bid 2S.

Obviously, we'll pick up on more penalty doubles, too, when partner passes the double and he ought to be passing with more
frequency than the support double which is made both with and without enemy suit length.

My example hands might be poor, but interested in thoughts on the basic idea.

[Siegmund]

When I was playing polish club, I used 1C-(p)-1M-(overall)-X as negative rather than support. Sounds like you may be in a similar position with your nebulous diamond here.

[awm]

I play this with many partners. It's not that different from "optional support doubles" though -- personally I was never enamored of being forced to make a support double on a 4333 hand or a hand with substantial length in the suit the opponents bid, as we will often trade a plus score on defense for a minus score when declaring. Another advantage is that it clarifies a number of follow-up sequences. Playing support doubles I am often left wondering if auctions like 1♦ - Pass - 1♥ - 2♦ - Dbl - Pass - 2♠ are supposed to be forcing, whether they guarantee five hearts, deny five hearts, etc. You could probably have agreements about these (although I suspect it may be supposed to depend on the level and the sequence) but playing takeout double clarifies basically all of them. It also helps you find your 4-4 spade fits in some auctions when otherwise you have no chance.

You do occasionally get into some trouble when opponents raise, for example 1♦ - Pass - 1♠ - 2♥ - Pass - 3♥ - ??? might leave responder poorly placed if he holds five or six spades and a poor hand (i.e. double might be risky) whereas if opener had made a support double he can compete to 3♠. But in general I find that wins from this style outweigh the losses.

[straube]

I'm thinking that at low levels (up through 3m) that it would pay for opener to stretch in order to show shape.

For instance, in your 1D P 1S (2H) example, if opener will always compete to 3C with a 5/5 (even a poor 5/5) then after
P (3H) when responder is looking at AQxxxx xxx Kx xx (and if he trusts the opponents to have an 8+ card fit) then he will know that
partner must have 2-2-(54) shape.

That sort of thing.

If the bidding goes 1D P 1H (2C) I think with 4S/6D I ought to bid 2S.

Partner and I have already noticed that it pays to go slow when holding opponent's suit. For instance, after 1D (1S) we pass with Kxxx Axxxx xx xx. We don't want to endplay partner into bidding 2C. We can likely find hearts later.