Author: David Dahlem
Date: 09:02:46 11/02/05
Go up one level in this thread
On November 02, 2005 at 10:34:14, Uri Blass wrote: >On November 02, 2005 at 10:11:39, Tord Romstad wrote: > >>On November 02, 2005 at 09:31:13, Jouni Uski wrote: >> >>>Yes Uri seems to be right in this. In most of my tactical test suites engines >>>like Shredder, Fritz and even SlowChess are much = twice as fast! But still >>>Fruit has won all matches so far. So Fruit's secret is NOT the search, but?! >> >>What makes you arrive at this conclusion? It definitely doesn't follow from >>your observations about tactical test suites. You can have a very good >>search and do only moderately well on tactical test suites, and you can have >>a mediocre search and score exceptionally well in the same test positions. >> >>Tactical test suites does not measure playing strength, the quality of the >>search, nor even the tactical abilities of the program (my own program >>usually scores OK in tactical test suites, but is a relatively poor >>tactician in real games). Running such test suites can be fun, but the >>results don't tell us anything of importance. >> >>Tord > >I think that the question about being better in tactics is which test suite you >use. > >Maybe the arasan test suite is better because it include positions from computer >games and not from human games and other test suites include too many sacrifices >and tactics is not only about sacrifices. > >Uri Has the arasan test suite been debugged? Regards Dave
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.