There Might Exist Simple Rules For Accurate Position Evaluation

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

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Toadofsky
Posts: 27
Joined: Sat Dec 03, 2016 1:20 pm
Contact:

Re: There Might Exist Simple Rules For Accurate Position Evaluation

Post by Toadofsky » Fri Oct 25, 2019 10:46 am

towforce wrote:
Sat Oct 19, 2019 9:48 am
A quick thought experiment which, admittedly, entails a combination of optimism and plucking numbers out of the air.

Suppose that, using methods discussed in this thread, a set of 100,000 linear expressions could evaluate a chess position quite accurately. There are, on average, 30 legal moves in a chess position. This would mean that choosing a move would entail calculating 30 x 100,000 = 3 million linear expressions. A computer could do this as quickly as human perception, making move selection, from a human perspective, instantaneous.
I would like to see some web site run an experiment where humans submit evaluations for selected positions, and Leela is rewarded for her ability to predict those evaluations (and players can see what her evaluation is). Whether this would produce the same (engine) result as https://github.com/2014mchidamb/AdversarialChess , I do not know.

Post Reply