Author: Ross Boyd
Date: 11:55:33 10/25/05
Go up one level in this thread
On October 25, 2005 at 12:56:43, Dann Corbit wrote: >On October 25, 2005 at 12:28:07, Lar Mader wrote: > >>I realize that this is a difficult question... >> >>I'm curious about what makes Fruit 2.2 so strong. Also, it is impressive how >>quickly Fabien achieved this strength. Fritz 9 and Shredder 9 seem to be close >>to Fruit 2.2 in strength, and yet they have been in development for a much >>longer time with a lot more resources. Has anyone spent any time examining the >>2.1 source code, or have any other insights into what this program does that >>makes it so effective? > >Fabien does everything well. He is a magnificent and careful programmer. He >obviously puts a lot of effort into making things correct. He has a lot of >insight to pick out what is important and concentrate on that. > >His search is excellent and innovative. He does some things that nobody else >does. I'm not really sure how he is able to not hash the PV and still have a >stupendously fast search, but that is pretty amazing. > >His evaluation function is also very different from other evaluation functions. >He pays a lot of attention to control of the board. > >But it's kind of like asking "What makes Einstein so smart?" or "What makes >Michael Johnson so fast?" When you are talking about the best in the world at >anything, you are talking about a peculiar kind of genius or at least something >several standard deviations away from the norm. It takes a pretty smart person >to write any sort of a chess program. To make a world champion program in a >short time frame takes a sort of freakish excellence that you see only on very >rare occasions. > >IMO-YMMV. Good answer Dann. Fabien deserves much praise for his concise and methodical approach. Btw, what do you mean "I'm not really sure how he is able to not hash the PV and still have a stupendously fast search"? I'm guessing you mean he doesn't allow hash cutoffs at PV nodes. After a root search returns within the ab bounds he fills the hash with the PV but they seem to be used only as a suggested move. (His replacement scheme has some hard-to-comprehend ramifications.) Having said that I suspect there may be a minor 'bug' in his transposition table replacement logic - it shows up in Fine 70. Maybe its one of those 'bugs' that somehow help more than they hinder. Whatever, the guy's a grandmaster programmer if there is such a thing. Cheers, Ross If I were to describe his search in a few words it would be consistent and persistent. Its amazing how stable his PVs are.
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.