Skip to content
ChessAITools
Chessbook logo

Chessbook

Opening repertoire builder that recommends moves by win-rate at your rating, not just engine evals. Formerly Chess Madra.

Freemium · from $5 webiosandroid
Visit Chessbook →

No affiliate relationship with this tool. The link below goes directly to the product site.

Our take

Chessbook's core insight — that the best move against a 1500 is the one that scores best against 1500s, not the one Stockfish prefers at depth 40 — changed how I build repertoires for faster time controls. The coverage dashboard is quietly brilliant: it shows exactly which percentage of your likely games your prep actually covers. It won't teach you plans the way a Chessable course does, so I use both: Chessbook for the tree, a course for the understanding. At roughly $5/month for Pro, it's an easy recommendation.

Chessbook (formerly Chess Madra) is an opening repertoire builder by indie developers Marcus Buffett and Ollie Campbell. Instead of feeding you a grandmaster’s engine-checked theory, it mines the Lichess game database for what players at your rating actually face, and scores candidate moves by practical win-rate combined with engine evaluation.

You build the repertoire move by move, the app shows what fraction of your expected games it covers, and a spaced-repetition trainer keeps it sharp. Free covers about 400 moves; Pro (around $5/month, cheaper annually) removes the cap.

Pros

  • Recommends repertoire moves by actual win-rate at your rating from the Lichess database — far more practical than engine-first theory
  • Built-in spaced repetition keeps the whole repertoire drilled, with coverage stats showing the gaps
  • Indie developers who ship fast and respond to feedback; pricing is among the fairest in the category

Cons

  • The free tier caps you at around 400 saved moves, which a single serious repertoire blows through quickly
  • It's data-driven rather than explanatory — you get numbers, not a coach telling you why a move works
  • Occasional app stability complaints and evaluation quirks in less common lines

Related tools