magnus.persson@xxxxxxxxxxxxxxxxx wrote:
And there is no joseki-library. The idea is that the program should be
prepared for any move you play against it. A joseki-library is the last
thing I will implement in my program - that is: never since there
Thanks for the information on Viking. I am wondering how many other
programs adopt this approach of no joseki library.
Interestingly, a related topic came up in the book on computer checkers I
am reading now. The checkers expert on the project was resisting building
a stronger opening library (yes, I guess checkers has openings), since it
would, he claimed, reduce the program to a mere database lookup, etc.
etc. The author was in favor of a better library because it would, um,
make the program stronger.
Anyway, I can easily imagine the myriad difficulties in getting a joseki
library to work in a go program, but still all in all isn't this a
relatively easy way to make a program stronger? Even if the program is
smart enough to figure out the right joseki move "on its own" (i.e.,
without reference to a book), having the information available in the form
of a library should be able to chop huge pieces out of the search space
and let the program spend its time analyzing more interesting parts of the
game tree. But what's more, in most cases the program would _not_ be able
to figure out the right line on its own -- even in the game of checkers, a
game who knows how many orders of magnitude less complex than go, a
computer searching ahead 30 moves would still miss lines that were
correctly documented in opening books authored by humans.
--
Bob Myers <rtm@xxxxxxxxxxxxxxxxx>