Author: Rolf Tueschen
Date: 10:19:54 07/24/02
Go up one level in this thread
On July 24, 2002 at 12:13:11, Vincent Lejeune wrote: >After 14h40, not many changes : >Nouvelle partie >r1r1q1k1/6p1/p2b1p1p/1p1PpP2/PPp5/2P4P/R1B2QP1/R5K1 w - - 0 1 > >Analysis by Hiarcs 8: > >1. +- (1.65): 1.Db6 De7 2.axb5 Tab8 3.Dxa6 e4 4.Fxe4 De5 5.Ff3 De3+ 6.Rf1 Dd3+ >7.Rf2 Fe5 8.De6+ Rh7 9.Df7 Dxf5 10.g4 Dg5 11.Fe4+ Rh8 >2. ± (0.96): 1.axb5 axb5 2.Db6 Txa2 3.Txa2 Fc7 4.De6+ Rf8 5.Rf1 e4 6.Fxe4 Fe5 >7.Ta3 Ta8 8.Dxe8+ Txe8 9.Ff3 Fg3 10.Ta1 Te3 Thanks Vincent. Don't take me wrong when I say that your results speak a clear language. Of course we don't know what the program would play after 1099000456*12 hours! Perhaps then the two evals will get slowly _narrower_ and _narrower_ together as the widely accepted theory of traditional CC declared, but perhaps in 100 years we all will be too old to be capable to read what is on the display. BTW is HIARCS 8 an especially "stupid" program compared with CRAFTY18.15 or FRITZ 7? Or also this, is it perhaps not enough tweaked for this traumatic position for Kasparov? Did you use enough Hash? Was the position a priori implemented in the learning files? If not, please try to fussle a bit, so that your HIARCS finally gets in the neighbourhood of the top programs, speaking of DB2, CRAFTY10-18.15, FRITZ2-7! Sorry, DB2 is no longer available, but we have enough class with CRAFTY and FRITZ. Next year the two will take part in the Paris-Dakkar rally. This will be certainly a new world record so far. So, till then and a la prochaine fois. :) Rolf Tueschen
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.