Author: Dan Kiski
Date: 06:34:15 12/15/98
Go up one level in this thread
On December 15, 1998 at 08:34:08, Micheal Cummings wrote: > >On December 15, 1998 at 08:30:10, Dan Kiski wrote: > >>On December 15, 1998 at 06:19:06, Arda Renkver wrote: >> >>>I have read many interesting postings in here about those tests you guys >>>conducting. My question is that if you are just testing the engine strength for >>>different settings shouldn't the opening books be turn off in any case.Otherwise >>>you get no feeling as to how the "setting" is able to handle a position without >>>finding itself in a comfortable situation in which it arrives by using opening >>>theory from memory.In this case you only have match results without knowing >>>whether or not the result due to opening advantage or disadvantage.Maybe playing >>>with a set of predetermined openings with white and black would clear the cloud >>>a bit.I believe Mr.Kaufman (IM) used to suggest that a long time ago in print. >> >>This subject has been touched on many many times, I myself as previously stated >>use the Nunn test positions from Chessbase. > >When I asked about this test I pretty much got an answer that the Nunn tests are >not worth much these days, because some programmers can set up their engine to >find this positions easily and thus when people test the engine on these >positions they can give misleading results. So I do not bother using that test, >I do not trust it, I rather have two personalities battle it out, or have CM6K >battle it out with another program. While naturally any programmer can set up to play any move whatsoever, the Nunn positions let the engine play after move 10, naturally both sides should be allowed white and black. You can easily tell if the engine is actively seking the best move or playing a programmed response. Also with most programmes you can modify it's opening book to only know these opening positions that eliminates any possible way the engine would have a pre-programmed move. Again the advantage to this type of set-up is that it prevents opening book wins, and compares only the engine has they both get white and black from the same starting point. That isn't to say as Harald Faber suggested that all openings couldn't be worked to an equal point in time, they could if someone is prepared to do all that work, but any opening point can be used made up by anyone as long as both sides are given white and black from the same position.
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.