output statistics

[ghost]

Maybe write to the console for each player encountered in the pgn file, how many inaccuracies, mistakes, blunders etc he or she made:

30000 games processed kasparov, 1 blunders, 3 mistakes, 4 inaccuracies deepblue, 8 blunders, 1 mistakes, 3 inaccuracies

That way one can use it as an evaluation tool of the strength of e.g. a chess program: make a change, let it play 30k games, count how many bl/mis/inac on average the program made. Repeat for each new version. You can then graph or so in e.g. a spreadsheet program the progress.

[ddugovic]

If anyone feels motivated to do so, please consider the winning chances and from that (not from raw centipawn loss) somehow judge move accuracy.

[fsmosca]

@flok99

This is interesting when someone let chess artist analyzes many games at once.

Here is what I am thinking so far. Just create an option to get summary. Example

chess-artist -infile a.pgn -outfile eval_a.csv -job evalsummary

This way other files analyzed by chess artist in the past can be processed.

I need to look in detail again how chess artist annotates position like,

15... c6 $0 {+0.13} ({Better is} 15...b5 16. Bd5 Rb8 17. Qe2 Bd7 {-0.10})

By reading that info in the annotated game we need a rule of when a move is blunder, mistake etc.

I am aware how Lichess implements this categories the past 1 or 2 years, I am not sure now if they revised this.

I am not sure if chess artist has bad move categories already such as blunder, mistake and inaccuracies.

Defining those categories is also not easy at all. In my other program chess game analyzer, I have different system of categorizing than what Lichess is doing.

Thanks for the input. I have not much time at the moment to work on chess artist.

[fsmosca]

@ddugovic

Thanks for the info, I will look at it in detail when I got the time.

[ghost]

I've implemented it myself, now waiting for fsmosca to give his agree because it uses some code of chess-artist.

[ghost]

@fsmosca https://github.com/flok99/my-eval