On Tue, 2008-11-18 at 17:31 -0500, Michael Williams wrote: > Well, "vastly" when k is small. > > The only way to find a good Komi would be testing and guesstimating. > > I think MCTS would be well suited to this variant because you still have the > problem of difficulty in finding a good evaluation function and MCTS solves > that. > > Computers would probably be stronger than humans for sufficiently small > values of k. With the small branching factor, the computer would be able to > build a > very deep tree.
I keep thinking about this. My first idea was that alpha/beta might be a better choice with a low branching factor but then you have the evaluation issue. But does MCTS handle that as well as we think? I'm not completely sure of this. It seems like that each move would require more precision to be right. It may be that a lot of move are very obvious. I think a heavier playout would be required. This is just guesswork of course. - Don > > > Michael Williams wrote: > > I think computers would be much better at this game (than they are at > > Go) because you have vastly reduced the branching factor of the game. > > > > Ingo Althöfer wrote: > >> Hello, > >> > >> one of the basic problems of go newbies > >> is their tendency to place the next stone near to the latest stone of > >> the opponent. > >> Sometimes this is called the "2-inch heuristic > >> of beginners". > >> > >> What do you think about a formalized variant > >> of Go with one-sided distance-k rule? > >> > >>> Let k be some natural number. > >>> The normal rules of Go hold, except for one thing: > >>> When possible, White has to place his next stone > >>> within distance k (in city-block metric) of the latest > >>> stone of Black. If there is no feasible move of this type > >>> the stone has to be placed within the smallest > >>> possible city-block distance of the latest stone of > >>> Black. White may pass at any time. Example: > >>> On 19x19 board k=36 would mean no restriction at all.) > >> > >> * What should be fair values of komi(k) or fair numbers > >> of handicap stones? > >> > >> * Main question: How strong would MCTS-based programs be in this > >> variant(s)? > >> > >> * Would computers be stronger than humans for certain values of k? > >> > >> Ingo. > >> > > > > > > _______________________________________________ > computer-go mailing list > computer-go@computer-go.org > http://www.computer-go.org/mailman/listinfo/computer-go/
signature.asc
Description: This is a digitally signed message part
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
