Stockfish on standard opening position 1 trillion nodes

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

whereagles
Posts: 565
Joined: Thu Nov 13, 2014 12:03 pm

Re: Stockfish on standard opening position 1 trillion nodes

Post by whereagles »

so, it steers away from the dreaded FRENCH into the sicilian at depth 43?
fastgm
Posts: 818
Joined: Mon Aug 19, 2013 6:57 pm

Re: Stockfish on standard opening position 1 trillion nodes

Post by fastgm »

Thank you for your effort, Kai, and sorry for the mistake i made.
I edited the eval values by hand.

And thanks for your interesting experiments!

Andreas
User avatar
Laskos
Posts: 10948
Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos

Re: Stockfish on standard opening position 1 trillion nodes

Post by Laskos »

fastgm wrote:Thank you for your effort, Kai, and sorry for the mistake i made.
I edited the eval values by hand.

And thanks for your interesting experiments!

Andreas
Another chart with your data: the absolute difference between 2 successive evals:



Image


The fit shows that it very slowly decreases. Stockfish 8 seems to have a fixation for 0.07 eval change between two successive evals.
User avatar
Laskos
Posts: 10948
Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos

Re: Stockfish on standard opening position 1 trillion nodes

Post by Laskos »

Laskos wrote:Stockfish 8 seems to have a fixation for 0.07 eval change between two successive evals.
And yet another chart (histogram) from your data showing this 0.07 fixation of Stockfish, the 0.07 difference between successive evals to depth=59 analysis makes for more than half of all differences:


Image