Author: Paul Jacobean Sacral
Date: 02:29:19 12/01/05
Go up one level in this thread
On December 01, 2005 at 04:00:30, Mig Greengard wrote: >(The original composition was given as mate in 21, but the >comps find a faster way.) That seems to be acknowledged here in recent postings, BUT: If standard engines designed for games (only) are used, it is not 100% sure that they perfectly check for all defensive ideas. In other words, there is small risk that the best defense is missed due to pruning. You would either have to run a mate solving engine like chest with this position (I have no idea if it would take minutes or month for such a long mate), or you would have to enter a complete solution variant and check it by backwards analysis* with a good hash learing engine. *) = going move by move through the variant from that last position towards the starting position, in analysis mode. My experience is though, that even good "backwards" engines like Shredder 9 or Hiarcs 9 will have trouble over such a long mating variant, sometimes. >It is quite good in endgames for an engine despite apparently not accessing EGTB >in the search. Does its cousin Fruit 2.2 do this? Not yet 2.2, but the current version is 2.2.1 (as a free update) which has been enhanced with nalimov tablebase access. The fruit website http://www.fruitchess.com/ offers an unrestricted 14 days trial version. For ratings to compare Fruit, Toga and other new engines with the "established" ones like Fritz, Shredder etc., there are many rating lists based on thousands of games. I guess Computer versus computer is not be the main thing you are interested in, but nevertheless these ratings are informative. For example, your good impressions of Toga are acknowledged: Toga II 1.1 ranks between Fritz 9 and Shredder 9, currently at CEGT: http://kd.lab.nig.ac.jp/chess/cegt/rating-table-shifted.shtml Yours truly Paul J. Sacral
This page took 0.01 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.