Author: Majd Al-Ansari
Date: 13:46:45 01/01/06
Go up one level in this thread
I have to disagree with your comments. In my engine test Rybka can clearly gain at least 50 elo points with minor additional knowledge in the endgame. I am analysing the 400 + games now and will post why I think Rybka can be much stronger. Mind you that Rybka plays some amazingly strong endgames but can play at a level of a 1000 elo player in some positions. In the tournament I have done Rybka could have won or drawn at least 20 games outright if it only had some basic knowledge as to what endings are obviously drawn. I cannot count the many times that Rybka has thrown a won ending by simply going for an ending that gives it a higher evaluation (at times up to +8.0) while the other engine was showing an immediate drop to 0.00 On January 01, 2006 at 13:40:00, Jouni Uski wrote: >Hi, > >I have 10 hours spare CPU time and repeated engame match from Nunn/Larsson >positions with 60/10m level. Result: > >Fritz 9 ½½0½½½½½0½½½1½011001½10½½1½½01½1½0½½0½1½ 20.0/40 >Rybka 1.0 ½½1½½½½½1½½½0½100110½01½½0½½10½0½1½½1½0½ 20.0/40 > > >I feel, that Rybka 1.2 with added endgame knowledge and TB access will not add >many if any ELO points. Is it simple, that good search is enough to handle all >basic endings? Special trick/zugzwang/TB positions have not practical meaning. > >Jouni
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.