Re: Double-dummy solving

On Jul 31, 2:06 pm, moja...@xxxxxxxxxx (Everett M. Greene) wrote:
As a thought problem, I've been contemplating what's involved
in developing a double-dummy solver. I was thinking that a
brute-force, try-everything approach would work but then
realized that the "best" results would be when idiotic plays
are made by the opponents. 7N would (nearly) always be the
"optimum" contract by this approach so it would solve nothing.

Has there ever been anything written regarding this problem
and the approach(es) that could be used?

Bear in mind that double dummy assumes error-free play on both sides.
Your "best" results will never happen because the opponents are not
allowed to make "idiotic" plays.

The double dummy solvers currently available do use brute force, along
with search tree trimming techniques, rather than apply heuristics.
Search for Ginsberg and "partition search" in this group.

Cheers,

Carl

www.carlritner.com <- books, bridge, cheap, lots of them



.