Author: Dann Corbit
Date: 12:06:29 02/26/01
Go up one level in this thread
On February 25, 2001 at 13:31:23, pete stein wrote: >[reposted/tree was extended] > >If Eco.epd contains the computer analyses, how come there are 7857 analyses in >Eco.epd whereas there are 2614 openings in Eco.pgn; it implies the same openings >were analyzed several times with different programs, right ? The EPD positions are all of the unique board positions from the openings. The analysis is 12 minutes of PII 300 MHz time or more. Some of the positions were analysed at time control up to 24 hours. >Whether you can trust the analyses or not depends of which engines and which >time settings were used to analyse the openings ? You cannot trust any of the analysis beyond this: There are no obvious blunders within the sphere analyzed. You can get a much, much stronger analysis by stringing the analysis together by minimaxing it. >It would be even more interesting to have deep analyses of only the terminal >position of each opening. Fritz would be able to set a value for each of the >terminal positions and on the basis of these values a number of openings could >be selected to build an opening book database. That is a very good idea. Fritz cannot be used because Fritz is incapable of writing out EPD records.
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.