Author: Jaime Benito de Valle Ruiz
Date: 19:35:31 08/18/03
Go up one level in this thread
Very funny!
I can't believe that you sacrificed your knight & bishop only to prove how blind
a program can be in certain positions!
This is actually very similar to an old test position for computers that I tried
with Genius 5 long time ago... and I remember that no program at that time could
never find the solution. The program just "knows" that it's up two bishops, and
it does is avoid repeating the same position 3 times... until eventually reaches
the 50 moves, so it doesn't really have a clue about what's going on.
There are a lot of similar test positions based on the same trick to laugh at
them, if that makes you happy:
[D]3k4/8/p1p1p1p1/PpPpPpPp/1P1P1P1P/8/5Q2/2BK4 w - - 0 1
Although positions like these happen very rarely, it would be interesting to
find a quick & cheap way of detecting them during the evaluation without slowing
the search significantly. This kind of things cannot be left to be found through
deep search!!
It's a good test for chess engines, actually.
Keep posting things like this.
Regards,
Jaime
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.