So a one-sided rollout is done "on the fly" to derive the epc? If so, is this done on 0-ply with a static seed? How many trials? Can't see how the settings would matter much; I'm just curious.
Thanks. Chase On Sat, 2 Jun 2007 22:40:54 +0200, "Christian Anthon" <[EMAIL PROTECTED]> said: > These positions fall outside the normal database (15 checkers on 6 > points), so unless you have a larger database the epc is determined by > a one-sided rollout and small variations are to be expected. > > Christian. > > On 6/2/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > > Daniel Murphy noted this on bgonline.org. > > > > Two positions: > > > > Position ID: bbsFEQC1u00AAA > > Match ID: MAHgACAAEAAA > > > > Position ID: td1KAQC1u00AAA > > Match ID: MAEAAAAAAAAA > > > > The position of the player on roll is the same in each. The epc for that > > player is reported as 79.841 in the first position. but 79.759 in the > > second position. I experimented by making further changes to the > > position of the player not on roll and, sure enough, these changes > > caused the epc to change for the player on roll as well. > > > > Chase > > > > -- > > http://www.fastmail.fm - Access all of your messages and folders > > wherever you are > > > > > > > > _______________________________________________ > > Bug-gnubg mailing list > > [email protected] > > http://lists.gnu.org/mailman/listinfo/bug-gnubg > > -- http://www.fastmail.fm - One of many happy users: http://www.fastmail.fm/docs/quotes.html _______________________________________________ Bug-gnubg mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-gnubg
