Author: blass uri
Date: 03:26:22 05/31/99
Go up one level in this thread
On May 31, 1999 at 06:04:36, blass uri wrote: > >On May 31, 1999 at 05:26:24, Bernhard Bauer wrote: > > > >>The law "a null move program cannot do this simply doesn't exist". >> >>The question is: How do you implement the null move technique. >>And if you do it wrong, it remains a bug. > >Fritz5(16 bit) can also solve it and it is a null mover. > >I believe that the reason that fritz5.32 cannot solve it is simply that >Fritz5.32 is weaker than Fritz5(16 bit) > >Uri I can explain the fact that fritz5.32 cannot solve it in a change in the evaluation function that cause Fritz to believe that there is no threat when older version believed that there is a threat. Saying that Fritz5.32 cannot solve something because it is a null mover is not the same as saying that all the null movers cannot solve it. I believe that Fritz5.32 is weaker not only because of this problem but because I looked at games of Fritz5.32. It lost one game against Fritz4 because of stupid queen sacrifice(it believed that it is winning(mate evaluation) when it sacrificed the queen and after it saw that it is losing). I saw also games against Rebel10 in 1 hour per move when Fritz did positional errors. Fritz5(16 bit) cannot do these mistakes. I believe that fritz5.32 is a buggy program and the bugs influence more significantly at slower time control. Uri
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.