the idea of combining two chess engine is good. You can take advantage of features "best" of two different engines. The problem is what are the best features objective of an engine?! and above all, how many hundreds of tests are needed to judge a combination "good" ... you should ask the person who first posted to the application, I think it is in Rybka forum with this link:
http://rybkaforum.net/cgi-bin/rybkaforu ... pid=286301
Reality or joke
Moderators: hgm, Rebel, chrisw
-
- Posts: 6401
- Joined: Thu Mar 09, 2006 8:30 pm
- Location: Chicago, Illinois, USA
Re: Reality or joke
The first time I heard anything similar was from Ingo Althoefer, who posted a lot about it in r.g.c.c at the time, but the "third brain" was human.Xann wrote:I think it was the subject of Stefan Meyer-Kahlen's master's thesis (yes, in the 90's)! It was later added to Shredder GUI under the name "triple brain" (two engines + referee).cdani wrote:This idea of using two engines as one by selecting moves of one or another following i don't know which algorithm was in my long term list of ideas. Of course the algorithm can be optimized also by tests. I'm happy someone has done it.
Miguel