Author: Laurence Chen
Date: 06:41:32 11/25/00
Go up one level in this thread
On November 25, 2000 at 05:57:58, Stefan Fredriksson wrote: >On November 25, 2000 at 05:53:31, Steve Maughan wrote: > >>I came across this position when analysing with CM 8000. >> >>[d] rnb3kr/ppp3bp/3pq3/3N3Q/5R2/8/PPP3PP/R5K1 w - - 0 18 >> >>CM 8000’s search behaviour is ‘different’ and could be a bug. At first >>ChessMaster thinks Nxc7+ is the best move with a score of –1.19. Then at ply 7 >>this increase to +0.65. Everything is OK so far. Then at the same ply the PV >>changes to Raf1 with a score of +0.00. This implies that CM would prefer a move >>of +0.00 over a move of +0.65 (the fact that Raf1 is probably a better move is >>an aside). Now there could be a sensible explanantion e.g. CM could be fibbing >>when it comes to its PV. >> >>Does anyone have an idea what is going on? >> >>Regards, >> >>Steve Maughan > >Well, I dont see what is odd here. The King changes its mind about the >evaluation as it goes deeper and sees more. > >/Stefan Agreed, there's nothing odd about CM changing its mind. I believe that the so called "bug" is not really a "bug", as CM searches deeper it changed its mind with the previous evaluation, and that the move it intended to play 18. Nxc7 is a mistake because of 18. ... Qe3+. The bug seems to be that it did not display this evaluation as negative in the window, and instead it display the next best move Raf1 with a score of 0.00. Hence, there's nothing odd with the choice made by CM. Laurence
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.