How to improve at chess

The single highest-return habit is reviewing the games you actually play. Books and videos teach ideas in the abstract; your own games show the exact mistakes standing between you and the next rating band. Start every study session by looking at your most recent loss and asking where the game really turned — not where it ended.

Reviewing your losses with an engine is free. The hard part is being honest about what went wrong — not blaming time pressure or luck, but finding the move or plan where you diverged from what was correct.

Improvement is mostly subtraction. Most players lose the same way over and over: the same hung piece, the same bad opening, the same rushed move in a winning position. Find your recurring errors across many games and fix those first — one fixed pattern is worth more than a dozen one-off lessons.

A single-game review tells you what went wrong once. Looking across your whole game history reveals what goes wrong repeatedly, which is a completely different and more actionable picture.

Below around 1800, most games are decided by tactics — one player sees a fork or a hanging piece the other misses. Learn the core motifs (fork, pin, skewer, discovered attack, back-rank) and drill them until you spot them without calculating.

The goal of tactics training is not calculation speed — it is pattern recognition. Once you see "this looks like a fork" automatically, the calculation to confirm becomes fast and reliable.

You do not need many openings; you need a few you understand. Pick one reply to 1.e4, one to 1.d4, and one opening as White, and learn the plans behind them — where the pieces go and what you are trying to do — rather than memorising long lines.

The benefit of a small repertoire is that you reach familiar structures more often, which means you spend less clock time and make better strategic decisions out of habit. Depth beats breadth at every rating below master.

Endgames are the cheapest points in chess: a small amount of knowledge converts drawn or lost positions into wins for the rest of your career. Know how to checkmate with a queen, with a rook, and with a king and pawn. Know the basics of rook endings and the idea of the opposition in king-and-pawn endings.

The payoff is asymmetric. A few hours of endgame study will save you a point per month for years. The same hours spent on openings usually yield far less.

A one-move safety check — "what does my opponent threaten, is anything hanging?" — before every move removes most blunders. Managing your clock so you have time at the critical moments removes most of the rest. Habits, not knowledge, are what separate players at the same level.

These two habits cost nothing to learn and compound immediately. A player who runs a pre-move check on every move and reaches the endgame with time on the clock will beat a more knowledgeable player who does neither.

Vague goals ("get better") do not guide practice; concrete ones do. Pick the next band — 1000, 1200, 1500 — learn what actually separates it from where you are, and check your progress against your own game data every few weeks.

Progress in chess is not linear: you will plateau, then improve in a jump as a pattern clicks. Measuring over weeks rather than days keeps you from abandoning habits that are working.