Author: Thorsten Czub
Date: 14:09:54 04/13/98
Go up one level in this thread
>Maybe I did not make my point clear. What I am saying is That Fritz 5 >may be the best program on the faster chips out. Its fair the way you >tested, and fritz was not the best program you have seen. I tested fritz >on a P II 300 and it was the best program. So Fritz may not be the best >program for the slower chips but given the faster chip it is the best. I >also tested Nimzo 98 cm5000 Rebel(all versions) Mchess pro 7.1 >genius(3,4,5) ect. ect. Fritz 5 tested the best out of all the programs >I have tested by far. All the programs were tested on the P II 300. I understand completely what you say. But I am sure that CSTal would have played the same way against Fritz5. CSTal has no problems to play vs. Fast-searchers. The main problem is not the opponent but the question if chris has again created some nice bugs ... :-) >You tested Fritz 5 on a slower computer and you tested the other >programs on the same computer. You find the Fritz does not play well. >You see bad positional moves and just plain bad chess from fritz. I >don't see this in my testing at all. So I can only conclude from this >that Fritz is best on the faster chips. You are making a mistake. You test Fritz5 exactly against the programs it was tuned to play good. Try out NONAMES or programs this engine was not designed to play against and you will see Fritz5 lose. My suspicion is that this fritz version plays good against the top programs (rebel, hiarcs, genius) but weaker against e.g. Diep, Eugen, Zarkov, CStal, and and and. Or take the Paris games where Fritz played (also on a 300 Mhz machine) against ISIchess and many different NON-TOPS. The games were lousy. Don't tell me this is an exception. If you do only test exactly what this engine was designed to be tested against, you make a mistake. ChessBase - as Matthias said himself, prepared this engine to play against different opponents. Genius etc. They selected the engine with the best results. And sent it to sweden. Ergo will this engine produce the best results. If you change the test-environment, e.g. changing the opponent, you change the outcome of the whole thing. >And the other programs are best >on the slower chips. Now thats with all things being equal. Fritz(fast >chip) vs program X(fast chip)=Fritz is the strongest program out. >Fritz(slower chip) vs Program X(slower chip)=Fritz sucks and plays ugly >and bad chess. It is possable for both to be true and would explain what >you and I have seen. I understand your point. You made your points very fair and clear and I subscribe to them in all parts. I do have all programs and have made benchmarks of all of them on different machines and I know that different programs behave different on different pc's's depending on first-level-cache-size and instruction-set of the CPU etc.
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.