The Chess Variant Pages


There are several different ways in which a chess game can be notated. The best known are the algebraic notation (in long and short flavor), and the (English) descriptive notation. In the latest rules of FIDE, the (short) algebraic notation has become the official standard, and the only one that one is allowed to use in official chess matches. These new rules caused some upset with people that used the descriptive notation for their entire chess life. The debate reminded me of a chess variant, that is based upon this notation: U-chess.

Here, I describe what is needed to know about the English descriptive notation in order to play U-chess. Some fine points of this notation that are necessary to deal with cases to avoid ambiguties are omitted here (mainly to hide my ignorance in those matters.)

The long name of the variant is Unambigiguous Three-Symbol Chess. In the Encyclopedia of Chess Variants, Pritchard gives Mannis Charosh after Irving Chernev as inventor, with 1953 as year of invention, but then continues C. E. Swanson recalls playing a similar game under the name Telegraph C. and thinks the origins may go back as far as WWI (citing Nost-algia, issue 223. The game is still frequently played, and has also been used as a theme for fairy chess problems.


A move is only legal if it can be written down in English Descriptive Notation with at most three symbols. The `-' symbol does not count, but the `x' symbol counts. The player that takes the king of the opponent wins the game.

A few details and examples follow with the explanation of the main points of this chess notation.

Names of squares

First, I describe the names of the squares from the view of white. The board is seen as to have two halves: the `kingside' and the `queenside'. For white, the kingside is the right half of the board (as the king starts the game at the right side of the board - on e1 in algebraic notation.) The queenside is the left side of the board (again for white.)

Columns are names with one or two letters, depending on what kind of piece starts the game on the first row. Thus, from left to right, the columns are named: QR, QN, QB, Q, K, KB, KN, KR. To find the name of a square, after these letter, a number 1 to 8 is added - the first row at whites side gets the number 1, the next the number 2, etcetera.

The following diagram shows the names of the squares, as seen by white. (White is sitting at the bottom of the board.)

Note that the N stands for kNight - the K for knight would not be possible as this letter is already used for the king. In some cases, the knight is instead noted by Kt, or by S.

Black does not use the same names for squares as white, but instead counts the rows from his own side of the board. The names of the columns are the same. So, for black, the left side of the board is the kingside, and the right side of the board is the queenside. The names of columns for black are from left to right: KR, KN, KB, K, Q, QB, QN, QR.

So, each square has actually two names. To recapitulate, two examples:

However, squares can also have shorter names: from the rows KR, KN, KB, QB, QN, or QR, the first K or Q symbol can be dropped when there is no misunderstanding. So, for instance, the moves e2-e4, e7-e5, B f1-b5 are written down as P-K4, P-K4, B-N5. (Note that in the last case, the Q is dropped, as there is no bishop that can move to KN5, i.e., g5 from the starting setup.) So, each of these moves is legal in U-chess, as they use three symbols each (the hyphen - is not counted as a symbol.)

Notation for the pieces

The pieces are described by one letter symbols: K for King, Q for Queen, R for Rook, N for kNight, B for Bishop, and P for Pawn. The remark about the knight above also applies here: some (usually older) texts use instead Kt for the knight - we cannot use K for knight as the K is already used for the king. In a few cases, one also uses the symbol S for the knight (S from Springer.)

In cases where two of the same pieces could move to the same square or make a similar capture move, such possible ambiguity is removed by adding a Q or K symbol, describing whether this piece came from the queen- or kingside of the board. However, removing such ambiguities always means that at least four symbols are necessary, making such moves not legal in U-chess. (So, if white has a rook on d1 and a rook on h1 with no pieces between them, none can move to e1, as such a move would either be denoted as QR-K1 or KR-K1.)

Describing a move

Each move description first starts with giving the letter that notates the piece. Thus, if we describe a move made with a queen, the notation always starts with a Q; if we describe a move made with a pawn, the notation always starts with a P, etc.

After this, either a dash -, or a cross x follows. A move that does not make a capture has a dash (-), and a capturing move has a cross (x).

In case of a usual move, the square follows, as discussed above; in case of a capture, the name of the piece that is captured follows.

So, a move where a queen captures a pawn is described as QxP. Note that if the queen can capture different pawns, then such capturing moves are illegal in U-chess, as each of them needs more than three symbols (recall that the x is counted as a symbol.)

Castling is described as 0-0, or 0-0-0, and is always legal.

In standard descriptive notation, ambiguities can also be removed by check, mate, or en passant symbols, but these are ignored for the purposes of U-chess.

As checks are ignored, moves possible by pinned pieces are considered to cause ambiguities, so, if a player has two pieces of the same type that could move to the same square, and one of them is pinned, then still moving the other to that square is not legal.

A pawn that promotes needs a symbol to describe the type of piece moved to, but if only one pawn promotes, one can drop the squarename. This means that pawn promotion is legal, if and only if only one pawn can promote, and the move is a non-capturing move, or the move is a unique move of the type PxK (which ends the game.)

A stalemate example

A very short game leading in four moves to a stalemate is the following:
  1. P4R R4R
  2. P4D P4D
  3. C2R C2R
  4. C2D C2D
This is a stalemate, as any move legal in `normal chess' cannot written down in three symbols without ambiquities.

Note, the above example may be confusing for our English readers. Here is my (David Howe's) translation of the above example using letters for the pieces in English:

1. PK4 PK4
2. PQ4 PQ4
3. NK2 NK2
4. NQ2 NQ2

WWW page created by Hans Bodlaender. Information on the English descriptive notation put on the WWW by Joe Brooks is thankfully acknowledged. Thanks to Fabio Forzoni for the stalemate example. Thanks for Mark J. Tilford for pointing out the confusing example.
WWW page created: December 12, 1997. Last modified: February 20, 1998.