Author: Uri Blass
Date: 02:32:12 07/05/02
Go up one level in this thread
On July 05, 2002 at 04:54:15, Sune Fischer wrote: >On July 05, 2002 at 03:26:10, Uri Blass wrote: > >>On July 04, 2002 at 10:27:05, Sune Fischer wrote: >> >>>On July 04, 2002 at 07:52:10, Geo Disher wrote: >>> >>> >>>> >>>>New position >>>>r3rb2/p1p2p1k/5P2/1p6/6Q1/q2p4/6NP/5R1K w - - 0 1 >>>> >>>>Analysis by Fritz 6: >>>.... >>>>1.Qf5+ Kg8 2.Qg5+ Kh7 3.Qf5+ >>>> = (0.00) Depth: 20/52 02:03:57 10943059kN >>>>1.Qf5+ Kg8 2.Qg5+ Kh7 3.Qf5+ >>>> = (0.00) Depth: 21/53 05:45:03 30486830kN >>>> >>>>Sure this is a mate in 14? >>> >>>The draw score is expected because Black has huge >>>material advantage, so white is quite happy with the >>>draw (perpetual check i suspect). >>> >>>White needs to see the mate to know it has something >>>better than a draw, and Fritz is not the best matesolver >>>around, probably too much pruning of the quiet moves. >>> >>>-S. >> >>No >>There is no forced mate. > >No? What do you mean, JM says its a mate in 8, that is always forced or it's >just not a mate in 8 :) JM said mate in 8 about a different position >Fritz reaches "depth" 21 and doesn't see it, that doesn't have to be a bug >because the 21 plies are not fullwidth. > >>Humans can often see mate when there is no mate because they have imagination. >>computers have no imagination so they usually cannot see mates when there is no >>mate. >>The only exception is when the program has a bug but usually the bugs of >>computers are not the same as the bugs of humans. >> >>Uri > >I don't understand your point. >Both humans and computers try and do the best move they can, and if you are down >in material, usually a repetition is good, of course mate is even better but you >have to see it through to the end, computers will never gamble to find a mate. > >-S. The point is that the poster believed that there is a mate without proving it. Humans can sacrifice because they believe that there is no way to prevent mate because they do not calculate everything. 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.