Hacker News new | ask | show | jobs
Show HN: FireChess – Find the chess mistakes you keep repeating (firechess.com)
1 points by leonfresh 108 days ago
Hey HN, I'm a solo dev and I built FireChess because I kept losing to the same openings without realizing it.

Lichess and Chess.com both have great post-game analysis, but they review one game at a time. I wanted something that looks across hundreds of games at once and says: "You've played this position 14 times and lose 70% of the time — here's what to play instead."

What it does:

- Scans your Lichess or Chess.com games (up to 5,000)

- Finds repeating opening leaks — positions where you consistently pick the wrong move

- Cross-references the Lichess opening explorer to tell you how popular/sound each line is

- Detects missed tactics and endgame mistakes across all your games

- Runs Stockfish 18 (WASM) entirely in your browser — no server-side engine needed

- Includes a drill mode so you can practice the correct moves until they stick

- The free tier gives you 300 games at depth 12. Pro ($5/mo) unlocks 5,000 games, higher depth, and full tactics/endgame scanning.

Tech stack: Next.js, Stockfish 18 WASM, Lichess Explorer API, Stripe. All engine analysis runs client-side in your browser — game data is only stored server-side if you save a report.

I have a short trailer here: https://www.youtube.com/watch?v=m7oUz7t8uZA

Would love feedback on the analysis quality or anything else. Happy to answer questions about the Stockfish WASM integration or the pattern-detection approach.

1 comments

This is a fantastic idea and I would love to try it. Am I missing something? I don't have a google account and I’m on chess.com not lichess - how do I make an account?
Cool, I think you can scan your existing chess.com games by clicking the chess.com toggle? and enter your chess.com username. You don't need an account to try the tool
Aha! That worked. Brilliant idea, really useful - I’ve often thought ‘ i wonder what my better move is in this oft-repeated opening’ and this tells me. Thankyou!
The radar measure was surprising - it told me my play falls of a cliff when i’m in a worse position- I would have guessed i’m quite good at battling on, so that was an eye opener. To be clear, I am sure the analysis is better than my gut feeling.
Amazing, thanks for the great feedback :) I'll add it to the reviews!

Yeah I find the estimate of the radar chart to be quite accurate. I scanned a range of players from beginners, to pros like magnus and hikaru and myself who is in between (2200 elo) and it gets the rating estimates correct. So I think if we work on all the fields it's suggesting we can definitely climb!