Author: karen Dall Lynn
Date: 05:08:02 01/23/02
Go up one level in this thread
On January 22, 2002 at 23:23:29, Brian Katz wrote: >What is the purpose of the most recent Fritz 7 7,0,0,6 upgrade? >Changes to the interface? >Bug fixes? >Stronger engine? >I am at the moment running a 5 minute engine tournament with Fritz 7 >(7,0,0,6)with 32MB Hash tables vs Fritz 6 16 with MB Hash tables and Deep Fritzn >with 32 MB Hash tables. Small round robin. Four games between each engine. I >know this is not very many games to judge by, but it is all I have time to run >for now. >Any opinions on the performance of the New 7,0,0,6 engine? >Brian I have been running Blitz games overnight for the last week, with the following preliminary results over 429 games: ALDEBARAN, Blitz:4'+2" 0 1 Fritz 7 134.0/224 2 Shredder 6 66.0/183 3 Chess Tiger 14.0 61.0/111 4 Deep Fritz 24.0/51 5 Gambit Tiger 2.0 1.0/3 6 Junior 7 1.0/2 287 games - Engine version 7.0.0.2 ALDEBARAN, Blitz:4'+2" 0 1 Fritz 7 82.5/151 2 Chess Tiger 14.0 68.0/149 3 Shredder 6 0.5/2 4 Junior 7 0.5/1 5 Deep Fritz 0.5/1 152 games - last 50 games with Engine version 7.0.0.6 Tabulation for the last 50 games: 1. Fritz 7 wins 21 2. Tiger 14.0 wins 12 17 draws All games run in a Pentim 4 1.5Ghz 768Mb Rambus Tablebases: full 3,4,5 endgames from Crafty Website (7.03Gb big) Opening books: Tiger.ctg for Tiger, Fritz7.ctg for Fritz In the last 100 games, Fritz used a book to which Tiger book, Junior 7 Book, and an "hiperbook 2001" database were imported; I sensed a slight improvement over Fritz ability to raise from the openings with some advantage in the middlegame start. Some games (about 40) were performed with no endgame database, no book; just the forcing power of each engine. Fritz won, I do not have tabulations. I have been placed Fritz 7 agains Hiarchs, and minor engines too, but assistematicaly; Fritz won more often than not. I also found that as playing against top ranked machines over ICC and FICS, the introduction of some personal knowledge about openings, forcing some moves that are unexpected for the machines routine analysis, increases larguely the chances for Fritz winning. But usually Fritz crunchs the opponent chances in a very though way. I could not stand all the time watching evaluation profiles in front of my computer, that would be quite boring, but some games I saw showed that Fritz really has the ability to invert a bad initial middlegame position with little increases of position quality for each move. It is also worth notice that sometimes I found an amazing discrepancy between the engines evaluation - Tiger 14.0 reporting +- (0.47) while Fritz reported -+(-1.07) for instance; but along the game the evaluations came to match and in general Fritz was right. I used to be a skeptical and to believe that any of the six major engines would do a decent job as playing. But now I favor Fritz. My conclusion is that the only serious opponent to Fritz 7 is so far Tiger 14.0, however as the number of games increases it becomes evident that Fritz 7 supersedes every known cyborg brain. These results are of course revisable, IMHO Karen.
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.