Also, you might find different predictors of winning in lower rated games, so analyzing games between masters only would provide more insight into the game itself, whereas games between all levels might show an advantage to unsound but aggressive attacks, sort of bluffing their opponents.
Two questions about game stat correlations
Sort:
Oct 14, 2021
0
#2
Some additional parameters might be to take games of similar opening type—say Ruy Lopez—due to variance of strategy for different types of games. (For example, Modern openings sit back and defend the center from the flank).
Also, you might find different predictors of winning in lower rated games, so analyzing games between masters only would provide more insight into the game itself, whereas games between all levels might show an advantage to unsound but aggressive attacks, sort of bluffing their opponents.
Also, you might find different predictors of winning in lower rated games, so analyzing games between masters only would provide more insight into the game itself, whereas games between all levels might show an advantage to unsound but aggressive attacks, sort of bluffing their opponents.
Hi everyone!
Long time reader, first time poster. I wasn't sure which forum to post these questions in, but I would like to analyze some games, albeit a large set of them.
I've been wondering about two possible correlations in chess:
1.) Is there a correlation between winning games and the first player to pass a pawn over the halfway mark (4/5th rank)?
2.) Is there a correlation between winning games and the player who moves his pieces the most distance? This one seems more obvious as losing positions get "boxed in," but if we, say, measured the "distance travelled" by pieces and pawns (A Qh8 from a1 would be a distance of 7, a rook from a1 to a5 would be 4, etc.) before the game tips to a lost game, could we measure piece mobility as a predictor of outcomes when considered over a large set of games?
I'd like to know the answers to these, but I'd really like to know how someone would actually find out - by pulling up a chess database and running some python algo. I'd like to know how that sort of chess research is performed by the pros, and I'd like to know if these correlations exist at certain ELO ratings more than others.
After running this algo, I'd like to develop principles for play based on them. It might be prioritizing creating real estate on your side of the board as you encroach onto your opponent's, or knowing that fianchetto'd bishops threatening a bunch of squares is generally better than positioning them through a lot of movement.
Even if there are no strategic or positional principles to develop from the data, I'd still like to know the correlations.
Lastly, if there are famous players known for prioritizing (or not prioritizing) crossing the halfway mark of the board or "mobilizing" pieces.
Thanks in advance,
Homer