I'm not really fanatic about this either way. If I knew how to easily "fix" this, I probably would provide a mode to make it work either way.
I see it as aesthetically MORE pleasing when the program appears to be playing stupid, but then you realize that YOU are the one that doesn't understand. When I first started studying chess as a little boy, I was completely annoyed that the grandmasters never played the game out but resigned. I realize now that it is not aesthetically pleasing to pretend you don't realize the game is over. I feel this way a little bit here (but not enough to lose any sleep over.) It's almost ugly (to me) for it to pretend to fight for something that doesn't matter - almost childishly. But of course I agree that from a stylistic point of view you can see the cleanup phase as just a continuation of good habit. - Don On Sun, 2007-04-08 at 11:02 -0600, Arend Bayer wrote: > > > On 4/7/07, Don Dailey <[EMAIL PROTECTED]> wrote: > On Sat, 2007-04-07 at 14:36 -0400, Matt Kingston wrote: > > What I want from a commercial go playing program is one that > I can use > > to learn to be a better go player. This brings up two > important > > deficiencies in the "win by 0.5" strategy. If I'm always > loosing by > > half a point, It's difficult for me to see when I'm playing > well and > > when I'm playing poorly. If the computer doesn't exploit my > weaknesses > > because it knows that it will win anyway, then I won't learn > to defend > > properly. I'll never know if I "almost won" or if the > computer was > > just toying with me the whole way. The feedback from "the > computer > > just killed everything" can help me play better. > > Hi Matt, > > I have heard this argument before, but I personally do not > subscribe to it. Please let me explain why. > > Of what learning purpose is it if you are losing the game and > the computer gets you to focus on a dramatic battle somewhere > that is totally irelevant? What are you being taught? I > think you would develop the wrong sense of the game. Instead > of thinking and planning for a win, you would be thinking and > planning for local battles - and as UCT and MC have shown us, > this is the WRONG way to think about the game. > > I am mostly with Matt on this one. Take endgame: If you only ever > fight for every point in the endgame when it actually matters, then > you get very few opportunities to learn a precise endgame. 0.5 leads > before the endgame are just extremely rare unless the players are at > professional level. Also, in a typical 19x19 endgame, when you are > down by 3 points and playing against an amateur, your best chance is > just to fight for every single point and hope your opponent makes 2-3 > small slips that add up to 3-4 points. I.e. in a realistic game, > maximizing your winning percentage is often equivalent to just playing > for a small loss for now; so the UCT all-or-nothing approach is just > almost never the right thing to do for humans (unless the position is > truly hopeless). > > There are also aesthetical reasons why I prefer professional handling > of lost games over UCT's, but that's another email. > > Arend > > > _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
