Author: Frank Quisinsky
Date: 07:05:38 02/26/04
Go up one level in this thread
On February 26, 2004 at 09:38:04, robert flesher wrote: >Another 100 games almost finished and 2.10 is once again far ahead of 2.02. >Some questions come to mind when.... >Frank Quisinsky states he is 100% sure that 2.02 is stronger than 2.10! >Why would the author not name the engine 2.10 beta or alpha? If weaker? >What are the time controls 2.02 is considered stronger in? >Frank states 2.10 is tactically weaker!, Can we see the test results that >confirm this? >Once again i seriously doubt the claims made by Frank Quisinsky Hello Robert, I had start the test with Ruffian 2.0.2 and Ruffian 2.1.0 last week on two different systems. 01. Pentium IV Mobile 2.67 GHz, 128Mb for hashtables, 4-pieces tablebases with 4Mb Cache, Arena 1.0.4 GUI, Windows XP SP1a, ponder = off, Shredder with hyperthreading = on and the opening book by Andreas Runge. 40 moves in 20 minutes. Ruffian with the Leiden book as UCI engine! 02. Dual Xeon 2.8 GHz, 256Mb for hashtables, 5-pieces tablebases with 64Mb Cache, Arena 1.0.4 GUI, Windows XP SP1a, ponder = on, Shredder without hyperthreading and the opening book by Andreas Runge. 40 moves in 20 minutes. Ruffian with Leiden Book as UCI engine. You can find the results / games / log files in Arena Event Forum: http://f27.parsimony.net/forum67213/ It seems that Ruffian 2.0.2 played much stronger as version 2.1.0 and it seems that with ponder on Dual Xeon the results are not to good. Its only a test vs. one engine (Shredder 8.0). But I have 164 collect postions and I can see that in a lot of cases Ruffian solved the positions in the half of the time. Booth engine solved the most of the position (Ruffian 2.0.2 = 136, Ruffian 2.1.0 = 135). So the tactical playing level by Ruffian 2.0.2 is for my higher and this can be the reason that the results of Ruffian 2.0.2 in Eng-Eng matches stronger compare to version 2.1.0. I saw your results but I have not try Ruffian vs. Ruffian and not with soo fast time controls. Possible that Ruffian 2.1.0 is stronger with faster time controls. I have not the time to test all constellations but at the moment for me is interesting to see that Ruffian with ponder = on have not so good results. Possible that this is the reason for the bad SSDF results (so far I am not sure about it, have only 40 games vs. Junior got from SSDF). Best Frank
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.