Author: Uri Blass
Date: 00:35:12 02/21/01
Go up one level in this thread
On February 20, 2001 at 23:19:30, Robert Hyatt wrote: >On February 20, 2001 at 17:17:01, Uri Blass wrote: > >>[D]8/8/3p2r1/3P3k/2R1PQ2/1P4pP/6P1/r2q2NK w - - 0 1 >> >>Here Gandlaf played 64.Qf5+(the right move) but after Kh6 65.Qf4+ Kh5 it >>blundered with 66.Qe3. >> >>It was better for gandalf simply to trust previous search. >>the position before 64.Qf5+ is the same as the position before 66.Qe3 so Gandalf >>could do better by trusting previous search. >> >>I guess that Gandalf had less time for all the game when it played 66.Qf5+ and >>had a fail low without having time to solve it. >> >>I am interested to know the score of Gandalf when it played 64.Qf5 and the score >>of it when it played 66.Qe3 that is the losing mistake. >> >>Uri > > >It must be broken. I tried this and Crafty sees that Qe3 leads to a -3 score >in 2 seconds on my PIII/750 notebook. Qf5 has a score of 0.00 as it did from >the original position. Gandlaf4.32g on PIII450 wants to play Qf5+ with a positive score of more than +2 and never considers Qe3 in the first time. If I give it the moves Qf5+ Kh6 Qf4+ Kh5 then it suddenly wants to play Qe3 and has an horrible branching factor at depth 9. The score of Qe3 at depth 8 is positive and the score at depth 9 is mate against itself. It needs only 3 seconds to get depth 8 but needs more than 8 minutes to get a score for Qe3 at depth 9. The score at depth 9 is mate against itself and it can finds Qf5+ some seconds after finding that Qe3 is bad. Uri
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.