Author: Dann Corbit
Date: 10:59:08 03/17/04
Go up one level in this thread
On March 17, 2004 at 08:00:29, Vincent Diepeveen wrote: >On March 17, 2004 at 04:07:12, Bas Hamstra wrote: > >This endgame is dead won. A few years ago i indepth analysed it. Black has >*nothing*. But the evaluation problem is indeed this position and a bit further. >If you do a deep search from here you will see it is a win for white. > > > >I'm sure the GM in question missed this line unless world top, but would he have >seen it he would have played Rxe6 only faster :) > >Probably he/she saw: > >rxe6 qxe6 re1 qd6 bf4 f5 qh3 qxf4 qxh7+ kf6 bxd5 nd8 ne5 and now black can do >nothing but the mate is still a number of checks to go. So if diep nullmoves >here it will find the checks still in qsearch and find the mate. Crafty needs >another 10 ply here i guess :) > >I'm sure crafty & co are missing the above line. > >Diep sees this line at around 8-10 ply already using some extensions (singular), >but it currently has problems seeing directly the endgame is won after the line >with qe6-f5 x f5. > >That's why it doesn't find Rxe6 instantly. > >Corbit should claim less and implement checks in crafty. I did not use crafty for any analysis. I used Shredder 7.04, Ruffian 2.02 and DeepSjeng 1.6, all world class chess engines with very good capabilities. Did you actually look at the analysis or just brush over it without trying to read the trajectories? To me, it is very convincing.
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.