Built a tool to find out which openings and endgames are costing you wins

I kept noticing patterns in my opening repertoire but couldn't put my finger on it. So I built something that measures two checkpoints on top of the standard metrics: the engine eval at the position the opening actually gave me, and the eval at the endgame I entered.

For example I found in my own games: my Vienna Opening looks healthy overall, but I don't play well when I face Anderssen defense.

Everything shows 95% confidence intervals with raw counts, because most of these samples are small and I did not want a 100% built on one game to look like a strength.

It's free, no signup, and runs entirely in your browser. You can expand the openings to see specific variations and then the corresponding games.

Happy to explain the methodology, and genuinely want to hear how the numbers look to stronger or beginner players.

Try Plyscope here

I'm quite new to chess, but this looks like a great tool, and really easy to use. Hopefully I can use it to work on my weak areas and build me ELO.

Really nice work!

Thanks a lot, let me whether you learned something interesting from your play and what features you are missing if there are any

Plyscope has been updated to v0.3.

The main changes are that the reported elo changes are now real and not hypothetical. Lost games are now analyzed more in detail: three new columns split every game you did not win based on where the game was lost.

Report tables are much tighter. Every cell is now a single line for clarity. Columns are banded under the question they answer, every name fits on one line, and each cell names its own exact counts and intervals on hover. Percentages still carry the number of games behind them.

Let me know which features you are missing or would want to see!

https://plyscope.com/