Author: Harald Faber
Date: 04:51:13 12/16/03
Go up one level in this thread
On December 15, 2003 at 14:43:55, Andre van Ark wrote: >On December 15, 2003 at 13:48:27, Harald Faber wrote: > >>On December 15, 2003 at 12:56:25, Andre van Ark wrote: >> >>>On December 15, 2003 at 05:06:29, Harald Faber wrote: >>> >>>>On December 15, 2003 at 02:29:13, Kurt Utzinger wrote: >>>> >>>>>On December 15, 2003 at 01:45:44, Harald Faber wrote: >>>>> >>>>>>On December 14, 2003 at 18:59:53, William Rex wrote: >>>>>> >>>>>>>On December 14, 2003 at 16:02:26, Kurt Utzinger wrote: >>>>>>> >>>>>>>>******************************************************************** >>>>>>>>Athlon 1.3/256 MB, using Fritz 8 GUI >>>>>>>>time control: 120'/60 plus 60'/30 plus 30' >>>>>>>>ponder=off, no books, no EGTB, no learning >>>>>>>>games starting with 5 pre-set positions >>>>>>>>1.e4 / 1.d4 / 1.c4 / 1.g3 / 1.f4 >>>>>> >>>>>> >>>>>>They use a completely different setup! >>>>>> >>>>>>To point the differences out: >>>>>> >>>>>>1. SSDF uses *2* PCs >>>>>>2. SSDF plays ponder=on >>>>>>3. SSDF plays 120'/40+60'/20+60'/20+.... >>>>>>4. SSDF plays WITH books >>>>>>5. SSDF plays WITH TBs >>>>>>6. SSDF plays WITH learning >>>>>>7. SSDF plays without such stupid (IMO) setup positions. >>>>> >>>>> Hi Harald >>>>> You are definitely wrong as far as Junior 8 is >>>>> concerned. All other engines had the same handicap >>>>> in NBE 2003 but none of them did so badly. The only >>>>> explanation is that the "naked" Junior 8 has weaknesses >>>>> other engines do not have. As to Shredder 7.04: you can >>>>> run this engine under conditions whatsoever, it will >>>>> still be top. This can't be said to Junior 8. And in >>>>> our opinion the missing opening book was the main reason >>>>> why Junior 8 got such a bad result. When we play matches >>>>> with books [with/without EGTB, with/no learning] Junior 8 >>>>> always gets very good results in our tests. >>>>> Kurt >>>> >>>>Indeed Shredder is the best allround program and Junior depends on special >>>>conditions. But the conditions under which Junior plays well are different than >>>>yours and several other tests run. One important point is the book. Hundreds of >>>>posts and opinions have been written how senseful the usage is. However, Junior >>>>8 is sold WITH book. Judge yourself and decide if you want to castrate Junior by >>>>disabling this book. >>>>Furthermore permanent brain is another point Junior absolutely needs. And >>>>finally, 40/120 plus pb is much slower than 60/120 without pb. >>>> >>>>Sure, one can say: Bad program or bad engine design when it needs own book, >>>>permanent brain and all the 5man TBs. But this is a COMMON setup, even with the >>>>5man TBs. >>>>I cannot remember an offical tournament (Thueringen, Aufsess, Leiden, >>>>Maastricht, Graz, New York, etc.) where these conditions have not been >>>>fulfilled. So I do not see any reason why complain. It is "only" the engine >>>>testers/users who use a worse setup and get worse results. Is this a wonder? >>>>Certainly not. >>> >>>Hi Harald, >>> >>>Most of us are engine-testers and not book testers ;-) >>> >>>André >> >>Well, usually the book is contained in the engine package. Do you want to see M. >>Schumacher run 60km vs. Raikkonen, Coulthard, Montoya etc. to see which of them >>is the better F1 racer and leave out their cars? >> >>BTW if you are "engine tester", why don't you test without any book at all? > >Hi Harald, > >Because the engines make the most ridiculous moves without a book. So who decides after which move it makes sense to test them after 1/3/5/10 moves? And which are those moves? >Every engine need a book what suits its style best. But I also like to see how >an engine peforms in different circumstances. OK, but is an engine bad or disappointing when it does not perform best in YOUR positions? Generally speaking certainly not, only for you maybe. >"Schumacher run 60km vs. Raikkonen, Coulthard, Montoya etc. to see which of them >is the better F1 racer and leave out their cars?" > >No I think it would be fair if all of them ride in the same car. Would it? Consider Schumacher driving MUCH better with more traction, different seat or whatsoever. Consider Uhlmann (specialist for the French defence) play Sicilian vs. another GM who by chance is a strong Sicilian player. Do you think their equal ELO guarantees a drawish overall result? >Well, in a closed position Junior is a blind horse and Shredder and Hiarcs >perform much better then. That is why nunn-matches are interesting. I don't say such matches aren't interesting. It is just that the results of such tests/matches might not reflect whyt you get when you let them play with their strongest weapons. If <enter program name of your choice> plays the worst result in Nunn test but is in the top5 of SSDF I see no reason to doubt SSDF or question how this can be. There is still an easy answer: The <enter program name of your choice> simply does not play the Nunn positions well. Nothing more, nothing less. IMO no reason to complain or to be disappointed. >Not only the elopoints count, but also the ability to deal in different >positions. > >André
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.