Author: Omid David Tabibi
Date: 04:13:21 10/14/03
Go up one level in this thread
On October 14, 2003 at 06:52:09, Ulrich Tuerke wrote: >On October 14, 2003 at 06:33:39, Omid David Tabibi wrote: > >>On October 14, 2003 at 05:58:27, Ulrich Tuerke wrote: >> >>>>... >>>> >>>>The version with all the extensions and checks in quiescence decisively >>>>outplayed Tiger in blitz time control (5 minute per game on my PIII/733MHz), >>>>while it got demolished in 1 hour per game time control... >>> >>>Doesn't this mean that the architecture of Tiger sucks ? >> >>No, it means that the architecture of Falcon using all the extensions sucks :) >>Sometimes at the 4th iteration it searches as deep as 40+ plies in some lines, >>which gives it a huge tactical strength in lower time controls. But the tactical >>advantage disappears in longer time controls, and what remains is a bad >>positional play with no decisive tactical strength. >> >>Although this results in a very good blitz performance on my slow hardware, I >>believe in faster hardware the tactical advantage will be insignificant even in >>blitz time controls. > > >Come on, Omid. > >You shouldn't sabotage my attempt to "pull Chris' leg" a bit. -:) :) > > >BTW i guess that Falcon is the successor of Genesis, is it a complete re-design >? Yes. >And it's really outplaying CT at Blitz ? Not the normal version. I was testing a version with many extensions and checks in quiescence which performed very well in solving tactical test suites. That version clearly outperformed some of the commercial engines (Tiger included) in fast time controls (on my slow hardware I have to mention), but the results were far from satisfying in longer time controls... > >Regards, >Uli >> >> >>>Looks like it plays strong on longer time controls and looses on blitz. >>> >>>Uli >>> >>>>...
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.