Super cool! Downloading and running against existing engines doesn't seem that hard to me? Just running against stockfish a bunch of times would narrow your error bars a lot. I've seen chess engine battles on twitch. I'm sure someone in that community has what you need. Sorry if that is obvious and unhelpful. I'd love to try out your engine if you are looking for testers. I can compile stuff (unless its Windows).
M.
I am curious if anyone knows how to find an accurate rating for a chess engine I have spent the past few months developing.
I've come across resources stating I need to pair it up against other engines and play thousands of games. This makes complete sense, but unfortunately, I don't have the time (or computing power) to do this. I've tried using Arena and CuteChess, but I'm wondering if there's something better that can be used for rating purposes.
The other thing missing is that I need a few engines with already established, accurate ratings to compare against. Preferably, I start with low-rated engines (1800-2000) to establish a baseline, then move up into the 2000s until I find the limit. My program can consistently beat engines with estimated ratings of ~2000, but I'm unsure how accurate some of the claimed ratings are without sources of thousands of games to back the claims up.
Does something exist to make this easier to do, or do I just need to manually install some engines and let them play, which is very time-consuming?