Quoting David Fotland <[EMAIL PROTECTED]>:
7x7 is pretty much solved (by people, not exhaustive search), so it's a much
easier game than chess.
Correct komi is 9 points. http://senseis.xmp.net/?7x7BestPlay
Here are some random comments about 7x7 and Valkyria. I think the
complexity of
7x7 is much less than chess simply because the games are so short. In
chess you
can move around pieces for a long time but in 7x7 the critical fight
often lasts
just 10-15 ply.
Valkyria is good enough to beat me most of time if I play the side that is not
favored by the komi. Also if I experiment with moves that I know is not
part of
the solution then I am in deep trouble as well.
Thinking time matters a lot. I often give Valkyria very long thinking times so
the above might not apply for fast games.
The hard part is as always the opening. I found that everytime I change
something in the simulation part of Valkyria, then the moves selected in the
opening also change a lot. Mostly I have improved fighting strength, but this
does not seem to be helpful in the opening so Valkyria tend to play
random good
moves. But a good move is often a losing move in 7x7. Valkyria plays stronger
though when it comes closer to the end of the game. And if you make one single
mistake then the program will crush you, unless you can reach a position with
seki or something that it does not understand properly.
Right now the opening book of Valkyria is just a handful of moves for
the first
1-3 ply, but I guess if I implented all the moves in the solution that
was sent
around in an sgf file it will be really hard to beat. It played a handful of
games of KGS and won almost all of them. But it was clear that the opponents
knew very little about 7x7 opening strategy so it was not a tough test
compared
to what Crazystone did.
Magnus
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
