Guides

Longer pieces on how to study, each built from something measurable rather than from the usual received wisdom about openings.

Most opening advice has the same blind spot. It is written from the engine's side of the board, where every move under consideration is already a good one, and so it never mentions the moves that will actually decide your games — the natural, plausible, popular ones that quietly lose half a pawn or a piece. Those are invisible to an engine precisely because they are beneath its notice.

Finding them takes two sources that disagree. A database of ordinary players says what people really do in a position; an engine says what it is worth. The overlap — played often and bad — is where the useful advice lives, and it is a different answer in every opening.

So everything here is built that way. Move frequencies come from the Lichess opening explorer at club ratings, 1600 to 2200, across blitz, rapid and classical. Evaluations come from Stockfish, screened at depth 20 and re-confirmed at depth 26 before anything is published, because shallow searches call all sorts of ordinary moves refutations. Where a number appears below, it was measured rather than remembered.