Do You Have to Memorise Chess Openings? What to Do Instead

The popular image of a chess player is someone who has memorised thousands of variations. It is true of professional chess and misleading for everyone else: below 2000, opening memory is one of the lowest-yielding uses of study time there is.

The number that settles the argument

Look at your last twenty games and count the move on which your opponent played something outside main theory.

The answer is almost always between move four and move eight.

That means the theory of move fifteen — the hardest to memorise and the fastest to forget — almost never appears. And that the hours spent there defend no games at all.

What happens when you memorise without understanding

You reach move nine perfectly, your opponent plays something different, and the script runs out. The problem is not that move: it is that there is no criterion for the tenth, because you never learned what the position was after.

The classic symptom: games where the opening "went well" and by move sixteen you have no idea what to do. The opening did not go well: it went memorised.

What you actually study about an opening

In this order:

1. The idea. One sentence. "The Italian aims for fast development and pressure on f7." "The London builds a solid structure and attacks on the kingside." If you cannot say it in one sentence, you have not understood it yet.

2. The pawn structure it produces. This determines everything else: where the open files are, which squares are weak, which side you play on.

3. Where the pieces go. In every opening, each piece has two or three natural squares. Knowing them settles most moves without calculating anything.

4. Which is the key break. Almost every opening has a pawn advance that frees the position — c5, d5, e5, f5 depending which. Recognising when to play it is half the opening.

5. The first six to eight moves of the main line. Yes, some memory. A little.

6. What people actually play, which is almost never the main line. Ten blitz games tell you faster than any book.

The test of whether you understood

Take the typical position in your opening after move ten and answer without looking anything up:

• Which is my worst piece and where does it want to go?
• Which break am I preparing?
• Which wing am I playing on?
• What is my opponent trying to do?

If all four answers come, you have the opening. If they do not, you know moves.

When memorising is right

There are three concrete cases, and they are few:

Traps and short refutations. If your opening has a line where your opponent can lose quickly, or where you can, those six moves need to be exact. There are not many and they are learned once.

Forced lines. Some variations are forced eight or ten moves deep: there no criterion is possible, they have to be known. Also few.

Above 2000. There, yes — opponents prepare, and memory starts to weigh.

What to do when you are taken out of book

This is the most frequent situation and almost nobody trains it. When your opponent plays something odd on move five:

Do not look for the immediate punishment. An odd move almost never loses by force. Hunting for the refutation is how you lose a comfortable position.

Ask what it weakened. Every move that does not develop weakens something: a square, a pawn, the king, a tempo. That is what to aim at exploiting, without hurrying.

Go back to principles. Development, centre, castling. Against an odd move, playing well is enough.

And count tempi. If your opponent spent two moves on something that does not develop, you are two moves ahead. Spend them on development, not on a premature attack.

How to build a repertoire you do not have to memorise

A well-built repertoire remembers itself, and the trick is not in the memory but in the choosing: openings that resemble each other. If with White you play systems where the bishop goes to g2 and with Black you do too, you are not memorising two things: you are learning one idea and using it twice.

Three criteria for choosing, in this order:

Structures should repeat. That is what makes the experience of one game useful in the next. A hundred games with the same structure teach more than three hundred with different ones.

They should not depend on your opponent cooperating. An opening that wins quickly when the other side errs and has no plan when they play well is a trap, not a repertoire.

You should like them. It sounds soft and it is not: you are going to play these positions hundreds of times, and a position that bores you gets played badly because it gets thought about little.

With that, a complete repertoire is four or five lines — one against 1.e4, one against 1.d4, a system with White and two answers to the odd stuff — and none of them needs more than ten written moves.

How many moves you actually need to know

The question has a fairly concrete answer, and a much shorter one than people fear:

Up to 1400: zero memorised lines. The three principles — centre, development, castling — cover everything, and games are decided long before theory matters.

From 1400 to 1800: six to eight moves per line, plus knowing the idea of each: where every piece goes, what structure it aims for, where the play is. Eight moves is what it takes to reach a familiar position.

From 1800 to 2100: ten or twelve moves in the main lines, and rather more detail in the sharp variations you have chosen. This is where move order starts to matter.

Above that: there you genuinely do need memory work, and there a database earns its keep.

Put in hours: for almost everyone reading this, the complete repertoire is four or five hours of work, not four months. What takes time is not learning it, it is playing it until you understand it.

What to do after each game with your opening

This is the method that replaces memorising, and it is half an hour a week:

Note the move where you left known territory. Not the move where you went wrong: the move where you no longer knew what was coming. It is almost always much earlier than you think.

Look at what gets played there. A free database is enough. Do not look for "the best move": look for which move is played most and why it fits the opening's plan.

Write it in your own file. One document per opening, with the lines that have actually turned up against you rather than the ones a book considers important. A repertoire built this way has exactly the lines your opponents play.

Re-read it now and then, five minutes. Not studying it: reading it. That is enough.

After three months of this, the repertoire is built and it is yours, in the literal sense that you wrote it out of your own games. That gets remembered in a way that a file downloaded from the internet never does.

In short

Studying openings is useful. Memorising them, hardly ever. The difference is that the first gives you criteria for every position that arises and the second gives you a script that runs out on move nine.

And most importantly: an opening you understand teaches you chess. One you memorised only teaches you that opening, until you forget it.

Related questions

What do I do if my opponent plays something I do not know on move three? The three principles: occupy or contest the centre, develop a minor piece, and castle. That gets you out of practically anything odd, and odd things are almost always objectively worse than normal ones.

Is it better to always play the same opening? Yes, a long way up. Repetition is what turns an opening into knowledge, and playing five different things guarantees understanding none. Variety gets added once the first one is understood.

Are opening traps useful? They win games against people who do not know them and they teach nothing. The problem is not that they are dishonest — they are not — but that when your opponent answers well they leave you in a position worse than the normal one, with no plan.

When do I start using a database? When you can name the exact move where your knowledge runs out. Before that, the database gives you a thousand lines for a problem you do not yet have.

Related questions

How much theory does a 1600 player know? Far less than people think: six or eight moves per line, and above all the idea. What separates a 1600 from a 1300 is almost never opening theory, but what they do with the position the opening left them.

Can I learn an opening just by watching games? Yes, and it is one of the best ways. Ten master games with the same opening give you the whole idea — where the pieces go, where the play is, what the typical endgame looks like — which is exactly what a file of variations does not give.

What do I do if I am out of book on move five? Go back to the principles and breathe: centre, development, castling. Leaving the book early almost always means your opponent played something objectively worse, so the position is good even if it is unfamiliar.

Is preparing against a specific opponent useful? At club level, almost never, and it consumes time that returns more elsewhere. It starts being worth it when you know your opponents by heart, which means closed tournaments and team leagues.