Building a repertoire

A game on lichess.org, 2014. White began the King's Gambit, 1. e4 e5 2. f4, and Black replied with the Falkbeer Countergambit, 2… d5. Instead of the main move 3. exd5, White took the e5 pawn with the f-pawn and left the king without cover on the e1–h4 diagonal. Black to move.

White chose a sharp opening but did not know its main trap. Two queen checks, and Black is a rook up. That is the main lesson of this chapter: the opening you play has to be known at least a few moves deep and, above all, you have to know its traps. Two openings known well are worth more than twenty known a little.

What a repertoire is

An opening repertoire is the set of openings you always play. As White you choose your first move and your answers to the main defences. As Black you choose an answer to 1. e4 and an answer to 1. d4, and against rarer moves you keep a general plan. With such a set every game starts from a familiar position, and you never have to invent the opening at the board.

A repertoire is easy to picture as a tree. The trunk is your first move or your answer. Branches grow from it: the opponent's moves. From each branch grow your replies, and so on. The deeper the tree, the more there is to remember. A beginner needs a tree five to eight moves deep: beyond that, understanding the ideas matters more than remembering moves.

Below is a board with an opening book. Play the opening you usually play on it, for both sides. After every move the book names the opening and says where the line it knows ends. The book holds 3,865 named variations, and every opening from the earlier chapters is in it.

Rare first moves need no theory of their own. Against 1. c4, 1. Nf3 or 1. g3 you can set up your pieces so that the position looks like your defence against 1. d4: a pawn on d5, a knight on f6, a pawn on e6. Against really rare moves such as 1. b3 or 1. a3, the rules from the chapter on opening principles are enough: take the centre, develop the pieces, get the king to safety.

Choosing to suit your style

An opening should suit your style. If you like attacking and sharp positions, open games as White and the Sicilian or the King's Indian as Black will suit you. If you prefer quiet play and long manoeuvring, choose the Queen's Gambit, the London System, the Caro-Kann or the Slav. No opening is good or bad for everyone: there are openings that suit you and openings that do not.

Your own games help you find your style. Look at which ones you won with pleasure and which ones you lost without a fight. If you more often win by attacking the king, look for openings where such attacks arise. If you get lost in sharp positions and play the endgame confidently, choose openings where play is calmer and the pieces come off sooner.

Here is an example of a simple repertoire for a player who is just starting. As White: 1. e4; against 1… e5 the Italian Game, against the Sicilian the Alapin, 2. c3, against the French and the Caro-Kann the Advance Variation, 3. e5. As Black: against 1. e4, 1… e5; against 1. d4, the Queen's Gambit Declined. You have seen all of these in the earlier chapters.

As White, the first question is 1. e4 or 1. d4. After 1. e4 the play is more often open: the pieces come into action quickly, and the games are full of tactics. After 1. d4 the centre is usually closed by pawns, the fight goes more slowly, and understanding pawn structures matters more. Beginners are usually advised to play 1. e4: open positions teach development and tactics faster, and that helps in any opening. If you prefer slow play, start with 1. d4 without a second thought.

Notice how White's answers are chosen. The Alapin and the Advance Variation give positions with pawns on d4 and e5 or d4 and e4 and similar plans. Once you have learnt them, you are ready for three defences at once. This trick, choosing openings with shared structures, saves the most time.

The tree of variations

Let us take one branch of a repertoire and look at it more closely. White plays 1. e4, Black replies 1… c5. Now White has a choice: the Open Sicilian with 2. Nf3 and 3. d4, the Alapin, 2. c3, the Closed Sicilian, 2. Nc3, the Smith-Morra Gambit, 2. d4. Try them on the board below and see what the book calls each branch.

Of each branch you need one, your own. The rest you only need to recognise: which variation it is and where it leads. In the book, the Alapin appears after 2. c3, and the Smith-Morra Gambit after 2. d4 cxd4 3. c3. If you play Black, on the contrary, you need answers to all the main branches, because the opponent chooses.

It is worth writing the tree down. A plain text file or a notebook will do: the opponent's move, your reply, one line about the idea. For example: "2. c3: I reply 2… Nf6 and attack e4; after 3. e5 Nd5 the knight stands actively". Such a note is shorter than a table from a book, and it is easy to go over before a tournament. Note separately the opponents' moves that caught you off guard: in a month those very lines will turn out the most useful, because such moves come up in your games again.

Transpositions

The same position can be reached by different roads. In the chapter on closed games we called such coincidences a transposition. The book checks the position after every move against its database, so it names the opening even when the moves came in a different order. On the board below the game began with the English Opening, 1. c4, and after 3. d4 it became the Queen's Gambit Declined.

Transpositions help and hinder. They help because you can sidestep an unpleasant variation. A player who dislikes the Nimzo-Indian, after 1. d4 Nf6 2. c4 e6, plays 3. Nf3 instead of 3. Nc3: without a knight on c3 there is no pin …Bb4. If Black plays 3… Bb4+ all the same, the book names a different opening, the Bogo-Indian Defence. They hinder because the opponent can steer you into a position from someone else's repertoire. So it pays to know which positions your openings can turn into.

Traps in your openings

The game at the start of the chapter showed that not knowing the main trap costs the game. For every opening in your repertoire, write down two or three traps: the ones you could fall into and the ones opponents often fall into. We have already looked at many: the Siberian Trap in the Smith-Morra Gambit, the Elephant Trap in the Queen's Gambit Declined, the trap with Nd6# in the Caro-Kann, the Lasker Trap in the Albin Countergambit and the Budapest Gambit trap.

Here are two examples from games played in 2014. The first is the Philidor Defence, 1. e4 e5 2. Nf3 d6. After 3. d4 Bg4 White took the pawn with the knight, 4. Nxe5, and forgot that the g4 bishop looks along the diagonal at the queen on d1.

The second is the London System. The black queen has come out to d6, right onto the f4 bishop's diagonal.

Both mistakes are simple, and both are typical of their openings. In the London System the f4 bishop looks at d6 and c7, and it is dangerous for Black to put the queen there. In the Philidor the g4 bishop and the d1 queen stand on one diagonal, and before moving the knight off f3 White has to check whether the queen will be left hanging. Knowing such details is what marks a player who knows their opening.

How to study openings

It is best to study an opening in this order. First the ideas: which pawn structures arise, where the pieces go, which pawn breaks the two sides make. Then the traps. Then the main line, five to eight moves deep. And only then the side lines. A player who understands the ideas will find a good move even when the opponent leaves theory on move three.

Model games help a great deal. Find three to five games by strong players in your opening and go through them move by move. You will see how the opening turns into the middlegame and what plans lie behind the first moves. Games stick in the memory better than bare variations: they have a story.

The moves of the tree need going over. A board with the opening book is handy for this: start from the initial position and play your line for both sides without looking at your notes. Where you go wrong is your weak spot. Before a tournament ten minutes are enough to go through the whole tree this way. Without repetition half the moves will be forgotten in a couple of weeks, and that is normal: just go back to your notes.

Most important of all is to play and review your own games. After every game open it on the analysis page: the engine shows where you left the best move. If you went wrong in the opening, find that spot in your tree and add the right answer. That way the repertoire grows out of your practice, and you learn exactly the variations your opponents play.

Check your repertoire

When the tree is ready, check it with four questions. Do you have an answer to each of the opponent's main moves: as White to the main defences, as Black to 1. e4 and 1. d4? Do you know, for each opening, where the pieces go and which pawn break is the main one? Do you know two or three traps in each? And do you enjoy playing the positions you get? If the answer to any question is "no", start with that one.

A repertoire changes over time. As you get stronger, you will want to try more complex openings. Change one branch at a time: first a new defence against 1. e4, then a few months later, if you feel like it, something else. That way the old knowledge stays, and the new is added to it.

Typical mistakes

The first mistake is changing openings after every defeat. You lost with the French, switched to the Sicilian, lost there too, took up the Caro-Kann. In the end no opening has been studied. A defeat in the opening is a reason to add a branch. There is no need to throw the tree away.

The second mistake is memorising moves without ideas. A player remembers fifteen moves of the Najdorf, but on move ten the opponent played something else, and the knowledge ran out. If you know why each move is played, you will find a plan in an unfamiliar position too. The third mistake is choosing an opening because it is fashionable or because the world champion plays it. The champion knows thousands of variations in it, while you need an opening whose positions you understand.

What's next

The opening ends when the pieces are developed and the kings are safe. Then the middlegame begins, where you have to evaluate the position and make a plan. How an engine evaluates a position, and what of that is useful to a human, is the subject of the next chapter.