Skip to content
Aback Chess
Sponsors

Chess glossary

The words an analysis board uses, defined plainly: 34 terms on notation, engine analysis, move quality and the patterns that decide games.

Board and notation

The vocabulary of describing a position: squares, files, ranks and the two text formats every chess tool speaks.

Algebraic notation
Also calledSANstandard algebraic notation
The standard way of writing a chess move, using the piece letter and the destination square, such as Nf3 or Qxd5+.
Captures use x, check uses +, checkmate uses #, castling is O-O or O-O-O, and promotion uses =Q. When two identical pieces can reach the same square, the notation adds just enough of the origin square to disambiguate, as in Nbd2.
How to read standard algebraic notation
File
A vertical column of eight squares, lettered a to h from White's left to White's right.
The a-file is the queenside edge from White's point of view, the h-file the kingside edge. Square names are always defined from White's side, regardless of how the board is displayed.
Chess board coordinates explained
Rank
A horizontal row of eight squares, numbered 1 to 8 starting from White's side.
Rank 1 is White's back rank and rank 8 is Black's. A rook reaching the seventh rank is a common attacking motif because unmoved enemy pawns sit there.
Chess board coordinates explained
FEN
Also calledForsyth-Edwards Notation
A single line of text that fully describes a chess position: piece placement, side to move, castling rights, en passant target, and the move clocks.
FEN records one position, not a game. It is what you paste into an analysis board when you want to start from an arbitrary position rather than play into it.
What is FEN? Chess positions as textFEN fields explained, with examples
PGN
Also calledPortable Game Notation
A plain text format for recording a complete chess game, made up of a header block and a movetext block.
Headers carry players, event, date and result; movetext lists the moves in algebraic notation, with comments in braces and variations in parentheses.
What is PGN? How chess games are recorded
Ply
A single half-move: one move by one player.
White's move and Black's reply together are two plies, or one full move. Engine depth is measured in plies, so a depth of 16 means eight moves by each side.
Chess engine depth explained: what depth actually means

Engine analysis

How a chess engine reports what it has found, and how to read those reports without over-trusting them.

Evaluation
The engine's assessment of a position, expressed in pawns from White's point of view, so positive values favour White.
An evaluation is specific to the depth and settings that produced it. Differences of a few hundredths of a pawn are within the engine's noise and should be treated as level.
Chess evaluation explained: how to read +1.2 and M3
Centipawn
One hundredth of a pawn, the unit engines use to report evaluations.
Scores display in pawns, so 100 centipawns appears as 1.00. The unit allows the engine to express small positional differences that are not worth a whole pawn.
What is a centipawn? The unit behind chess evaluations
Centipawn loss
The difference between the evaluation after the best move and after the move actually played, measured from the mover's point of view.
This is the quantity move classifications are usually built on. A loss of roughly a hundred centipawns means the move gave away about a pawn.
What is a centipawn? The unit behind chess evaluationsMove classifications explained: inaccuracy, mistake, blunder
Principal variation
Also calledPV
The sequence of moves the engine expects to be played with best play from both sides, and the line that justifies its evaluation.
A principal variation is evidence, not a prediction. Games rarely follow it beyond a move or two, and the line is recalculated whenever a player deviates.
What is a principal variation? Reading the engine's best line
MultiPV
Also calledmultiple principal variations
An engine setting that makes the search report several best lines instead of one.
Each line carries its own score. Raising MultiPV divides the engine's effort across more lines, so each one is slightly less deeply analysed for the same total work.
What is a principal variation? Reading the engine's best line
Depth
How many plies ahead the engine searched before reporting a result.
Higher depth generally means a more reliable assessment and always means more waiting. A score is only comparable with another score produced at a comparable depth.
Chess engine depth explained: what depth actually means
Nodes
The number of positions the engine examined during a search.
Nodes per second is a measure of how fast the engine is running. Depth and nodes together describe a search; depth alone does not capture how much work was done.
MultiPV lines
The separate ranked options the engine returns when MultiPV is greater than one, numbered from the best line upwards.
Seeing two lines with nearly equal scores usually means the position is flexible and more than one move is reasonable, which is often more useful than the single best move.

Move quality

The labels a review applies to moves, and what each one is claiming.

Blunder
A serious error, usually costing a pawn or much more, or turning a good position into a losing one.
A blunder is defined by a threshold in centipawn loss, so different tools draw the boundary slightly differently. A label describes a specific analysis rather than an objective property of the move.
Move classifications explained: inaccuracy, mistake, blunder
Mistake
A clear error, typically around a pawn or more in centipawn loss.
Between a mistake and a blunder the distinction is one of scale. A missed win is a special case: the position was winning and no longer is.
Move classifications explained: inaccuracy, mistake, blunder
Inaccuracy
A minor concession, usually a few tenths of a pawn, that may change the character of a position without changing who is better.
Inaccuracies are the most common grade in a normal game and the least worth dwelling on individually. Their value is in aggregate: a consistent set of them may point at a habit.
Move classifications explained: inaccuracy, mistake, blunder
Missed win
A move that lets a winning advantage go, judged from what the position was before the move rather than after it.
Labelling these separately is more useful than calling them blunders, because the lesson is different: you had the win and did not convert it.
Move classifications explained: inaccuracy, mistake, blunder
Accuracy
A percentage summarising how close a player's moves were to the best move, across the positions that were analysed.
Accuracy is derived from engine evaluations, so it depends on the depth and settings used, and it covers only the moves the engine actually assessed.
Move classifications explained: inaccuracy, mistake, blunderHow to analyze a chess game (a workflow that works)
Book move
A move that still follows known opening theory rather than being an independently calculated choice.
Book moves are usually excluded from quality judgements, since both players are following prepared lines rather than solving the position.

Patterns and concepts

The recurring shapes that decide games, and the principles that guide play when no tactic is available.

Fork
A move that attacks two or more enemy pieces at once, so the opponent can only save one.
Knights and pawns are the most effective forkers. The most common decisive fork in practice also gives check, because the check forces the king to move before the other threat is answered.
Chess tactics: forks, pins and skewers explained
Pin
A position where a piece cannot move without exposing something more valuable behind it.
An absolute pin is against the king and is legal to enforce; a relative pin against a queen or rook can be broken at a price. Pinned pieces can be attacked repeatedly because they cannot run.
Chess tactics: forks, pins and skewers explained
Skewer
An attack on a valuable piece with a less valuable one, which forces the valuable piece to move and exposes whatever stood behind it.
The mirror image of a pin: in a pin the less valuable piece is in front, in a skewer it is the more valuable one that must give way.
Chess tactics: forks, pins and skewers explained
Discovered attack
A tactic where moving one piece unleashes an attack from another piece behind it.
A discovered check is the sharpest form, since the check must be answered while the moving piece is free to create a second threat.
Chess tactics: forks, pins and skewers explained
Tempo
One unit of time, effectively a single move. Gaining a tempo means developing while forcing the opponent to respond.
Tempo is why chasing an early queen is effective: each attack on her forces a move that does not develop anything.
Zwischenzug
Also calledintermediate movein-between move
A move played in between a capture and its expected recapture, usually a check or a bigger threat.
An in-between check can win a piece, because the opponent must answer the check instead of completing the recapture. It is one of the most common reasons a recapture sequence goes wrong.
Opposition
A king and pawn endgame situation where two kings face each other with one square between them, and the side that must move loses control of the key squares.
The player who does not have to move holds the opposition. In pawn endings that single tempo frequently decides whether a position is won or drawn.
Passed pawn
A pawn with no enemy pawns on its own file or on the adjacent files that could stop it from promoting.
Passed pawns are worth more than ordinary pawns because nothing but pieces can stop them, and their value grows as pieces are traded off.
Isolated pawn
A pawn with no friendly pawns on either adjacent file, so it cannot be defended by another pawn.
An isolated pawn grants open files for the owner's pieces and a clear target for the opponent. Whether it is a strength or a weakness usually depends on how many pieces remain.
Fianchetto
Developing a bishop to the long diagonal by first advancing the knight's pawn, as in g3 followed by Bg2.
A fianchettoed bishop exerts pressure along a long diagonal for the whole game and is a staple of openings such as the King's Indian and the Catalan.
Castling
The only move in chess that moves two pieces at once: the king moves two squares towards a rook and the rook jumps to the square the king crossed.
Castling is permitted only if neither piece has moved, the squares between them are empty, the king is not in check, and the king does not pass through or land on an attacked square. Rights are lost permanently once the king or rook moves.
Chess opening principles: what to do in the first ten moves
En passant
A capture in which a pawn takes an enemy pawn that has just advanced two squares past the square it could have been captured on.
The capture is only legal on the very next move. In FEN the passed-over square is recorded so a program can offer the move, and in PGN it is sometimes written with an e.p. suffix.
FEN fields explained, with examples
Zugzwang
A situation in which a player would prefer to pass, because every legal move makes their position worse.
Zugzwang is most common in king and pawn endgames, where having the move rather than the position is the whole difference between winning and drawing.
Fifty-move rule
A rule allowing either player to claim a draw once fifty moves have passed with no capture and no pawn move.
The FEN halfmove clock counts exactly this. The counter resets to zero after any capture or pawn move.