Snowie chooses the best move out of the list of possible moves. If all moves are equal (all with 0%) the first move in the list is as good as all others.
So Snowie plays a 'random' move. Nothing bad or good about that one. Just the best move. As did GNU Backgammon a long time ago. N. On Thu, Nov 18, 2010 at 10:09 AM, Peter Nikolov <[email protected]>wrote: > Back in time, I wrote this remark to Snowie forum group and also Cc: the > same message directly to them. > > For almost a year, they didn’t reply. No comment… > > > > Why Snowie never resign? > > Discussion started by *Peter Nikolov on 11 Dec 2009 18:21* > > > It’s a very strange fact that Snowie never resign… even in an absolutely > lost positions! Probably the creators of the program are not sincere and > noble enough openly & loudly to confess that sometimes Snowie also can lose > a game- just as anyone else does… Bad rolls, you know… lol > Maybe the so called “perfect” algorithm is not good to calculate when the > chance of win is equal to 0% (zero percent)!? > > In such case Snowie is starting doing some really very strange moves. > Instead to resign, or at least to play still properly and lose with a > dignity- leaving less possible checkers on the board and avoiding the full > disaster to be gammoned. But in these circumstances Snowie behaves exactly > like a small irritated child who can’t reasonably accept his defeat. Is it > not time to grow-up? Version 5 perhaps? > > Shame, Snowie, shame… > > > > > > > > PS. Meanwhile someone responded to me that if in such circumstances you > double, then Snowie will resign the game 100% for sure. True. But just > imagine you can’t do it under current result in the match! > > > > Viewpoint, anyone? :P > > > > _______________________________________________ > Bug-gnubg mailing list > [email protected] > http://lists.gnu.org/mailman/listinfo/bug-gnubg > > -- "He is no fool who gives up what he cannot keep to gain what he cannot lose." -Jim Elliot
_______________________________________________ Bug-gnubg mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-gnubg
