chess-pipeline
chess-pipeline copied to clipboard
guidopetri
"Complicatedness" of a position
This goes a little hand in hand with #14. GMs also often comment that positions are complicated. What does this actually mean? Recapturing a piece when you're a piece down isn't complicated, but when the position is very closed and there might be only one pawn push that is winning the game while everything else looks just as innocent but is a draw instead might be considered "complicated".
Much like the other issue, I think we could start this with heuristics, but then later develop a better explanation/formula for this metric. I have some commentary from the Lichess discord server saved somewhere that I'll try to post here that revolves around this topic.
Specifically, the relevance for this is: Is Leela/Lc0 just really good at coming up with complicated positions? What does a complicated position look like for a human, and why is it complicated? (this probably changes with player skill) How can we practice complicated positions?
I'm also interested in this because it might explain some games where stockfish can see obvious wins but humans can't.
https://chess.stackexchange.com/questions/9815/what-does-it-mean-when-a-position-is-complicated
Found out that ChessX has a complexity measure. It's undocumented, but I found this quote from (I think?) the chief maintainer on chess.com:
The value given with "complexity" tries to give a hint on the complexity of a position with the help of the engine as suggested by Matej Guid and Ivan Bratko. As complexity metric for an individual move, the sum of the absolute differences between the evaluation of the best and the second best move is calculated. It is invoked at every time that a change in evaluation occurs when the search depth is increased.
Therefore, the number of lines has to be set to '2' to start the calculation.
Guid/Bratko use this metric to evaluate the performance/quality of a game with human players, therfor they need to introduce a metric.
I found some very interesting information on the following site (on web archive since it is down):
https://web.archive.org/web/20090219174044/http://www.truechess.com/web/champs.html
with accompanying FAQ:
https://web.archive.org/web/20090220234552/http://truechess.com/web/champsfaq.html
and more details:
https://web.archive.org/web/20090220234628/http://truechess.com/web/projectdetails.html
Additionally - the Guid/Bratko reference is to a paper: https://www.researchgate.net/publication/220174548_Computer_Analysis_of_World_Chess_Champions