Re: [computer-go] Mogo: nakade
I used --19 --time 30 --nbThreads 2 DL -Original Message- From: Sylvain Gelly [EMAIL PROTECTED] To: computer-go computer-go@computer-go.org Sent: Fri, 5 Oct 2007 5:48 am Subject: Re: [computer-go] Mogo: nakade I set 30 second playing time, but often it takes up to couple minutes for a move. That is not normal, could you post the command line you use? Cheers, Sylvain ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/ Email and AIM finally together. You've gotta check out free AOL Mail! - http://mail.aol.com ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] Mogo: nakade
Hi, ok I understand. Unfortunately MoGo is not multithreaded on windows due to the bad performance of pthread. So it actually use only 1 thread. However a side effect of --nbThreads x is that the time per move is multiplied by x, because the time is computed as CPU time. In your setting that means that MoGo actually uses twice as much time, ie 60 s. Sorry for the confusion for you, and for other people who may have run into it. Cheers, Sylvain 2007/10/6, [EMAIL PROTECTED] [EMAIL PROTECTED]: I used --19 --time 30 --nbThreads 2 DL -Original Message- From: Sylvain Gelly [EMAIL PROTECTED] To: computer-go computer-go@computer-go.org Sent: Fri, 5 Oct 2007 5:48 am Subject: Re: [computer-go] Mogo: nakade I set 30 second playing time, but often it takes up to couple minutes for a move. That is not normal, could you post the command line you use? Cheers, Sylvain ___ computer-go mailing list [EMAIL PROTECTED]://www.computer-go.org/mailman/listinfo/computer-go/ -- Email and AIM finally together. You've gotta check out free AOL Mailhttp://o.aolcdn.com/cdn.webmail.aol.com/mailtour/aol/en-us/index.htm?ncid=AOLAOF0002000970 ! ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/ ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] Mogo: nakade
I set 30 second playing time, but often it takes up to couple minutes for a move. That is not normal, could you post the command line you use? Cheers, Sylvain ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] Mogo: nakade
I played quite a few games with Mogo to recognizing the game pattern. I'm one or two levels higher than Mogo in 19X19. For the first 15 moves or so Mogo playing moves in the center(except the database forced plays in the corner). Probably its search concluded the center is the best place to surround territory. Once center is covered, Mogo begin to play attack moves. Probably its search concluded that from now on simply covering the territory won't lead to win.?For human players this point is started somewhat earlier. Up to now and in next few moves Mogo didn't show marked weakness. However, one major weakness shows up in moves following. It always try to kill a weak group. Good Go players all know that the?attack is not always (in most cases it's not) to kill that group. Mogo is so obsessed with the?killing effort, it often left several weaknesses in its position. I wonder what search mechanism biased Mogo towards such a behavior? I used a AMD 1.6GHz Turino dual core and the windows version of Mogo. At the end of game Mogo's time used is about an hour. I set 30 second playing time, but often it takes up to couple minutes for a move. It's ok, I would like it to think longer to come up better moves. DL? Email and AIM finally together. You've gotta check out free AOL Mail! - http://mail.aol.com ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] Mogo: nakade
Hi Darren, Thank you for playing (I hope you enjoyed :)) and your report. Sylvain, can you explain more about why it has this particular weakness? What you describe is exactly the issue I describe in the FAQ. From my analysis this come from the fact that the simulations are more likely to make the dead group to live than to die. It happens only on some situations, but when it happens, and you are in a part which is not explored by the tree, the probability for the group to die is low so MoGo simply consider that there is more important part of the board to analyse. Of course you can be more agressive on exploration and then get rid of this problem, but what I want is the best playing level in average, not solve some particular positions. I guess an overall fix of the simulation would be the best way to go for that. Also, is there a plan to fix it? Not by me at least ;). As I said earlier I am not working on it anymore... Cheers, Sylvain ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/