Part two

Why the word is study, not play

This site is called how to study chess for a reason. When researchers went looking for what actually separates strong players from weak ones, they compared everything: how many tournaments people played, how much coaching they’d had, how many books they owned, how many hours they’d put in and where those hours went. One activity kept coming out on top, in sample after sample: serious, solitary study. Time spent alone with games, thinking hard, working things out. Not playing. Not watching. Study, in the oldest sense of the word.

The scale of it is humbling. By the time grandmasters reached their title, they had typically logged around five thousand hours of that kind of work in their first decade, roughly five times what intermediate players had managed in the same span. Nobody is asking you to do that. The point is the direction of the arrow. The activity most amateurs do least is the one the evidence ranks first.

And what does “serious study” actually mean in chess? Strip away a century of method books and it comes down to spending real time with two kinds of games.

Your own games, first

The tradition here runs straight through the greatest chess school there has ever been. In the Soviet training system that produced a half century of world champions, every student arrived with their own games annotated by hand, including at least one loss, and defended their notes in front of the room. The annotations were published so that others could hunt for errors in them. The founder of that school considered self-analysis so central that he called chess itself the art of analysis. Every serious coaching tradition since has kept that spine: the player who studies their own games improves, and the player who won’t, doesn’t.

The reason is simple once you see it. Your games are the only place your actual chess exists. Not the chess you understand when someone explains it, which is always better than the chess you play. Your real openings, your real time trouble, your real habit of grabbing pawns on move twenty. A stranger’s course addresses a generic student. Your last hundred games describe exactly one player, and they do it in merciless detail.

The masters’ games, second

The second half of study is time spent with players far stronger than you. Every generation of chess players has learned this way. Morphy’s opera-box miniature from 1858, the Immortal Game, thirteen-year-old Fischer beating Donald Byrne in the Game of the Century: these get replayed a century and more later not out of nostalgia but because a well-annotated master game is the densest chess lesson that exists. You watch a whole plan unfold, move by move, with the reasoning attached.

Learning science has a name for this. A worked example: a complete solution you study before attempting problems of your own, which is among the most reliably effective ways to learn anything difficult. An annotated master game is exactly that, a worked example of good chess. And there’s a way to sharpen it that coaches have prescribed for generations: cover the moves and guess. Pick one side, decide what you would play, then compare against what the master played. Being wrong is the productive part. Committing to your own answer before seeing the real one is what makes the real one stick, and the research on learning says guessing first beats reading alone even when your guesses miss.

Here’s the strange thing. In that big survey of thirty thousand improvers, people listed everything they do: playing, puzzles, videos, books, courses. Studying master games barely registers. The method every school of chess teaching agrees on is the one almost nobody actually practises. Which means it’s also the cheapest edge available to you, because your rivals aren’t doing it either.

Where do puzzles, endgames and openings fit? They’re support work. All real, all useful, and all of them sharpen tools that the two kinds of study teach you to use. A tactic you spot in a puzzle only wins you games once you also spot it across the board with a clock running, and that transfer gets built in exactly one place: whole games, yours and the masters’.

Study means two things in chess: read your own games honestly, and spend time with games from players far better than you. Everything else is support work.

Part three

The method: five steps, then you go round again

What follows is the loop that the research and the coaching traditions agree on. None of it needs an account here. All of it works with a board, a notebook and patience. Read it as a whole first; the steps feed each other.

1. Play games long enough to think in

Rapid or slower. 15+10 is a good floor. Three games a week is plenty, and there’s no upper limit. Blitz is fun, and it isn’t training.

Everything downstream depends on the quality of the games you feed it. A slow game is one where you had time to look, to calculate, to feel the tension of a real decision. Those games contain information about how you think. A sixty-second game mostly contains information about your mouse. Play fewer games, slower, and treat each one as material for the next step rather than as a lottery ticket for rating points.

One habit to carry into every one of these games: before each move, spend one beat asking what your opponent’s last move threatens, and whether the move you’re about to make can be safely met. That single discipline, checking instead of hoping, is worth more than any opening file you will ever download. What you end up with: games worth reading.

2. Read your losses before the engine does

Soon after a game ends, give it ten minutes with the engine off. Write down what you felt and what you calculated at the hard moments. Then, and only then, check.

This is the step almost everyone skips, and it’s the one that separates players who improve from players who just play. The engine can tell you a move was wrong in half a second. What it cannot do is tell you what you were thinking when you played it, and the thinking is the thing you’re trying to fix. Were you calculating a line and miscounted? Did you see the threat and forget it two moves later? Did you not look at all? Three different diseases. The engine prescribes the same pill for all of them, which is why engine-first analysis teaches so little.

Do it soon after the game while the thoughts are still warm, because a week later you won’t be able to reconstruct what you actually saw, only what you now see. Write in your own words. “Felt scared of his kingside stuff, traded queens to make it stop” is a better annotation than a page of engine lines, because it’s true and it’s yours. Losses matter most. Wins flatter you; losses inform you. What you end up with: one named weakness, in your own handwriting.

3. Spend time with the masters

One master game a week, done properly: thirty to forty-five minutes, not five. Under about 1000, read a well-annotated game and follow the reasoning. Above that, cover the moves and guess for one side, then compare.

Choose games with words attached, not bare move lists. The classics exist for a reason: the attacking games of Morphy, the endgames of Capablanca, the collections written specifically to explain every single move to a club player. One game studied slowly beats ten games skimmed, the same way one book read beats ten blurbs.

If you want somewhere to begin, begin old. The nineteenth-century attacking games are direct enough that every idea is visible, which is why teachers have handed students the same handful of miniatures for a hundred and fifty years. Work forward from there toward games in the openings you actually play, so the patterns you collect start appearing in your own positions the following week.

When you guess along, something changes in how the game goes in. You stop being an audience and start being a participant. Each position becomes a small exam, and each master move becomes feedback on your answer. Some weeks you’ll match the master six times in a row and feel like a genius. Other weeks you’ll fail every guess in the middlegame and discover, precisely, where your understanding runs out. Both are wins. What you end up with: patterns you’ve made your own, and a growing sense of what good moves feel like before you can prove they’re good.

4. Drill your one biggest weakness

Take the single biggest leak from step two, not the top five. Train it in short daily reps, ideally with positions from your own games, hard enough that you get about four in five right. Openings: an hour a week at most, ideas over move orders.

A weakness doesn’t go away because you understood it once. It goes away through repetitions, spaced over days, until the right response becomes the automatic one. You can know everything about back-rank mates and still lose to one on Tuesday, because knowing and doing are stored in different places. Ten focused minutes a day on one named problem beats a two-hour Sunday binge across five of them, and it fits inside a normal life, which matters more than any other property a training plan can have.

Resist the urge to fix everything at once. The player who spends a month genuinely repairing their habit of ignoring threats gains more than the player who spends the same month lightly touching openings, endgames, tactics and time management. Narrow and deep, then move on. What you end up with: a drill streak on the leak, and games where the old mistake starts feeling wrong before you make it.

5. Check the number moved, then go again

After your next twenty-five slow games, measure the same thing again. If the leak’s number moved, the flag rate down, that opening’s score up, you’ve graduated. Pick the next leak and go round again.

Progress you can’t see in a number has a way of not being real. Rating alone is a noisy judge, it swings with sleep and luck and who you happened to get paired against. But the underlying numbers, how often you lose on time, how you score with your black openings, how many of your losses trace to the same mistake, those move honestly when the work is working. Measure before, train, measure after. When a number you targeted actually moves, you get the one thing that keeps adult improvers going: proof.

You graduate when the number moves, not when the checklist is finished.

Part four

What doesn’t count as study

None of this is bad. It’s just not training, and it’s where most chess hours quietly go.

Watching videos. It feels like study, but nothing gets tested, so very little sticks. Enjoy them as entertainment and count them as entertainment. Endless blitz. Your hundredth fast game of the week practises the mistakes you already make, at speed. Racing puzzles. Speed-solving positions you already know how to solve is a warm-up, not a workout; training happens at the edge of what you can do, slowly. Memorising openings. Your games probably leave known theory by move eight. Learn the ideas behind your openings and keep the memorised part short.

The test for whether an hour counted as study is uncomfortable but reliable: did you, at any point, have to commit to an answer you weren’t sure of, and then find out whether you were right? Reading your own loss passes. Guessing the master’s move passes. A video autoplaying while you nod along does not, and neither does your ninth blitz game of the evening. The hour has to come from somewhere. Spend most of it where the test passes.

Part five

The honest numbers, before you commit

About a third. That’s roughly how much of the difference between players comes down to hours of practice, across the major studies of chess skill. The rest is what you train and how. Aim beats volume, which is exactly why the method above starts with diagnosis rather than a bigger schedule.

Eight times. In one famous study, the slowest learner needed about eight times as much practice as the fastest to reach the same level. People are different, and the honest response to that fact is to measure your own numbers rather than trust anyone’s timetable. Anyone promising you a specific rating by a specific date is guessing with your money.

Plus 110 a year. The median gain for adults who study seriously and consistently. A quarter gain less than 25. Improvement is real, and it’s slower than the ads say. We’d rather tell you that on the first page than have you discover it in month three and quit. The players who make it are the ones who set up a loop they can sustain, and who can see their own progress in numbers smaller and faster-moving than rating.

Part six

How this site walks you through it

Everything above works with paper and patience, and we mean that. What we built is the version where the slow parts do themselves.

For step one, there’s always an opponent: a coaching bot you can set to any strength, with every move graded as you play. For step two, the free report reads your last hundred games in about a minute and does the honest accounting most of us avoid: which openings leak points, where the clock hurts you, what happens right after you lose two in a row, which mistakes keep coming back. Every claim comes with the number of games behind it, and when there isn’t enough evidence to say something, it says that instead. Here’s exactly what the report contains.

For step three, there’s a masters section with great games to play through and a guess-the-move trainer, so the cover-the-page method works without covering anything. For step four, your own blunders come back as spaced drills, and when you want to know why a move was wrong, a coach explains it in plain language, with every explanation checked against the engine before you see a word, so it cannot invent a line that doesn’t work. For step five, every report is saved, so this month sits next to last month and the difference is right there in front of you.