chess-pipeline
chess-pipeline copied to clipboard
Pulling games from the Lichess API into a PostgreSQL database for data analysis.
My games for 2023-04-30 took up all the server ram during `GetEvals` (which was running simultaneously with another user's `ExplodeClocks`), leading to a forced server reset. I had 18 bullet...
If someone played a category X of games, but we try running the email pipeline with Y category, the pipeline will still find games and not trigger the "you didn't...
Inspired by [this Reddit post](https://reddit.com/r/chess/comments/uwcn01/evaluating_the_value_of_every_piece_in_a_given/), there might be some way of evaluating relative piece worth by using an engine and manipulating the board directly. There's a few ideas in that...
This is far out, but hopefully doable someday: Maybe Lc0 evaluations could be added (as win probabilities? or maybe Q score directly?). This would require a GPU most likely, at...
Right now the positions/evaluations/eval depths get exploded separately in `GetEvals` which causes a bunch of issues (namely, what happens if there is a position but no evaluation for it?). It...
It's pretty well known that having the bishop pair is an advantage in most positions in chess. But is this actually the case? We could find out by marking out...
A lot of the data was run with an off-by-one error present in the positions/evals/clocks taken from the cloud or from a Lichess-side computer analysis. To fix: - truncate data...
A lot of the pipeline issues that crop up can't really be found by running tests without a proper staging environment. I think I can set up something under `tests/`...
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...
[Stockfish 12 is out](https://www.chess.com/news/view/stockfish-12-released-130-elo-points-stronger). Not only does the evaluation method need to change, we also need to make sure that it doesn't take too much longer with NNUE (via benchmark...