Computer Chess Club Archives


Search

Terms

Messages

Subject: Rather large EPD test suite is completed and can be found at my ftp site

Author: Dann Corbit

Date: 16:29:41 01/08/99


Here are the results after running crafty with time set to 1/100 second.  It
somehow manages to solve 1/3 of them.

White(1): epdscor epd.01
total: 7716
total acd: 23779   mean total acd: 3.08
total acn: 5383410   mean total acn: 697.69
total acs: 7717   mean total acs: 1.00
total mean node frequency: 697.60 Hz
solve: 2503   solve percent: 32.44
solve acd: 7748   mean solve acd: 3.10
solve acn: 1671453   mean solve acn: 667.78
solve acs: 2503   mean solve acs: 1.00
solve mean node frequency: 667.78 Hz
unsol: 5213   unsol percent: 67.56
unsol acd: 16031   mean unsol acd: 3.08
unsol acn: 3711957   mean unsol acn: 712.06
unsol acs: 5214   mean unsol acs: 1.00
unsol mean node frequency: 711.92 Hz

   ftp://38.168.214.175/pub/bench3.epd
or
   ftp://38.168.214.175/pub/bench3.zip

for the compressed version.

I actually simply intended the .01 second run just as a syntax test.  I am quite
astonished that 1/3 should be solved in this way.  I would be grateful for any
corrections, omissions, etc. that others can provide.



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.