Hmm. It will take quite a while to run a few thousand games with 200K playouts, and I might need a stronger opponent, but I'm trying it now.
From: computer-go-boun...@computer-go.org [mailto:computer-go-boun...@computer-go.org] On Behalf Of Olivier Teytaud Sent: Saturday, August 29, 2009 12:17 PM To: computer-go Subject: Re: [computer-go] fill board not effective in Many Faces If none, fill_board: pick a random empty point. If it is not on the 1st or 2nd line and there are no stones in 8 adjacent points, play it We repeat this several times, but this does not explain why it does not work for just a single time. The main weakness in your experimental setup is that you consider a small number of simulations per move, 10k, for which in our publication we have shown that there was no improvement - the improvement is for large numbers: 78.4% for 200 000 simulations per move. I hope it will be better now :-)
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
