Principles first, combinations follow: where chess tactics come from

A combination can look like it came from nowhere. A queen is offered, a rook takes a defended pawn, and three moves later the game is over. Play the game back a few moves, though, and you usually find the winner doing ordinary work before the sacrifice: developing pieces, castling, getting rooks to files that were about to open. The combination came at the end of that work.

Irving Chernev built Logical Chess: Move by Move around this idea. Strong players rarely go looking for combinations; they make the moves that give a combination something to work with. Two classic games below show it happening, and the later sections cover how puzzle training fits beside it. For the study habit itself, see the Chernev method.

Count the pieces before the sacrifice

The Opera Game is famous for its finish, a rook sacrifice on d7 followed by a queen sacrifice on b8. Look at the position before 13.Rxd7 instead. White’s queen, both bishops and the d1 rook are already at work, and the h1 rook reaches d1 one move later. Black has spent its moves trading the light-squared bishop, defending with the queen and pushing queenside pawns. The bishop on f8 and the rook on h8 have never moved, and the queen on e7 stands in the bishop’s way.

White’s combination uses up exactly that lead. The knight on d7 is pinned to the king by the bishop on b5, and the knight on f6 is pinned to the queen by the bishop on g5. Each time White gives up material, a defender disappears that Black has no spare piece to replace, because the spare pieces are still at home. Morphy’s sacrifices worked because he had more pieces in the fight and the open d-file let them reach the king.

Ask the same question in your own games before you capture or attack: how many pieces does each side have near the action? When the count favours you by two or three, look hard for forcing moves. When it is even, a sacrifice is more likely to lose material than to break through.

Rooks to the open file, king still in the centre

Steinitz against von Bardeleben, Hastings 1895, is remembered for a rook that checks the black king from e7, f7, g7 and h7. The moves that made it possible came eight moves earlier, and none of them gives check.

White castled on move 9, and Black’s king never left e8. 14.Re1 puts a rook on an open file pointing straight at Black’s king and pins the knight on e7. Black would like to castle, but 14…O-O allows 15.Rxe7, and nothing can recapture the knight. 15.Qe2 adds the queen to the same file, and 16.Rac1 brings the other rook to the half-open c-file. Then 17.d5 gives up a pawn to clear d4 for the knight, which travels to e6 and later to g5.

Each of these moves follows a plain rule: castle early, put rooks on open files, and keep the opponent’s king where it is. By move 22 Black’s king is back on e8 with the knight still pinned in front of it, and 22.Rxe7+ starts the finish. You can follow the whole game move by move in the library.

A short list, used in every game

Chernev returns to a small set of ideas across his 33 games, often in nearly the same words. In our own words, they are:

• develop quickly, usually knights before bishops, and move each piece once in the opening;
• fight for the centre with pawns and pieces;
• castle early and leave the pawns in front of your king alone;
• put rooks on open files, or on files that are about to open;
• don’t grab pawns while your pieces are still at home;
• don’t start an attack before your development is finished.

The list is short on purpose. A principle helps only if you apply it at the board without stopping to remember it, and that comes from using the same few ideas in game after game. Chernev also shows, about once per game, a strong player breaking one of these rules for a concrete reason. So a rule tells you where to look first, and the concrete position decides whether it applies. Our opening principles lesson works through the first three in detail.

Training the eyes: the same puzzles in cycles

A good position creates chances, and you still have to see them. Spotting a pin or an overloaded defender quickly is a separate skill, and repetition trains it well.

In Rapid Chess Improvement (2002), Michael de la Maza described solving the same set of puzzles in seven passes, each one faster than the last. Axel Smith and Hans Tikkanen developed the idea in The Woodpecker Method (2018). You solve a fixed set of exercises, then work through the same set again in cycles, aiming to finish each cycle in about half the time of the one before. Repeating puzzles you already know is intended. In the later cycles you stop calculating each one from scratch and start recognising the pattern straight away.

The Woodpecker and Seven Circles training here builds that fixed set from the motifs you miss most often, so the cycles practise the patterns you need most. Faster recognition doesn’t decide whether you should play for the pattern in a real game; that judgement still comes from the position. The lesson on chess tactics covers how to check a combination’s replies before you commit to it.

How the two halves fit together in a week

A week of training mixes both. Lessons and narrated master games show the principles at work move by move, with the reason for each move spoken aloud. Puzzle sets and cycles train the patterns those positions produce. Your own games, against your trainer or on Lichess, connect the two: they show which principles you keep under a clock, and which combinations you find or miss.

When your trainer reads one of your games with you, the moments that decided it get the same treatment as a master game. You hear what your move neglected and what it allowed, how a strong player would think in that position, and which habit would have found the better move.

Your games are the test

A puzzle tells you there is something to find, and a game never does, so your games show which habits hold up on their own. As your trainer reads them, your theme map fills in from the moves you actually played, judged by the engine: whether you got your pieces out, castled in time, kept your king’s pawns in place, put a rook on the open file, or left a piece loose. Those results sit next to your puzzle ratings, so your next week can aim at the habits your games reveal as well as the patterns your puzzles show.