On Sat, Dec 20, 2014 at 2:57 PM, Aja Huang <ajahu...@gmail.com> wrote: >> If so, do you have an idea how much >> extra performance that provides compared to only the last or not using >> it at all? > > > We haven't measured that but I think "move history" is an important feature > since Go is very much about answering the opponent's last move locally > (that's also why in Go we have the term "tenuki" for not answering the last > move).
That's pretty much how I looked at it as well. For getting a high prediction rate it is indeed a very useful feature, but it is unclear to me how important that really is. Some increased tenuki power may also have its merits. Perhaps I'll just run some experiments with Steenvreter to see what happens. I wonder how much worse a 6d human predictor would do without move history ;-) Erik _______________________________________________ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go