Author: Dusan Dobes
Date: 00:30:55 02/12/99
Go up one level in this thread
On February 11, 1999 at 17:21:26, Dann Corbit wrote: >On February 10, 1999 at 07:48:22, Dusan Dobes wrote: > >>On February 10, 1999 at 05:18:48, Jouni Uski wrote: >> >>>Hi Dusan! After testing Phalanx in couple of days I must say that it's >>>tactically unbelievable fast. In ECMF5 testsuite it scores almost same result >>>as Fritz5 and Junior5!! And it beats Nimzo, Rebel, Mchess, Genius by far in >>>same suite. >> >>Oh, thanks :-). Good Phalanx's results in tactical test suites are often >>achieved via huge king safety bonuses. And material sacrifices for king >>safety are always correct in test suites (that's the way how the positions >>are selected), but not in practical games. Maybe pawngrabber style would >>be stronger but I prefer high risk of incorrect sacrifice and 'king hunting'. >Can Phalanx be used for EPD test analysis (EPD in / EPD out)? >If so, it might be a very excellent addition to the C.A.P. project. It can read a file of EPD positions from it's standard input and analyze each position for specified time. This is intended to run test suites, so it expects a solution move (long notation) on the input line too. If it's not there, phalanx just says "solution incorrect" after the analysis. This feature still needs some programming as i ignore castling and en passant flags (originally i used the test suites from the gnuchess distribution and there are no castling/EP flags at all). Dusan
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.