Hacker News Reader

The rating system for problems seems like it needs to add in some kind of variance factor, or at least a provisional period. I'm altering problems' ratings by 20+ points with my default 1200 rating despite having less than 10 problems completed. My actual strength is hundreds of points above 1200, and so if I play on this site without making an account a bunch of times, I'm going to artificially lower every problem's rating...

View on HN

Yes, we're currently using Glicko-2 but we're not accounting properly for the Rating Deviation that incorporates previous games played (tactics attempted). This will be fixed in a future update. Thanks for the report!

I also think it would be really nice if the position switched to an "analysis mode" after completion. Sometimes, especially if one gets a problem wrong, it's nice to examine the position more deeply. Bonus points if some open-source chess engine is integrated (I presume this can be done client-side somehow?).

Reply on news.ycombinator.com