The Chess Variant Pages

Check out Cylindrical Chess, our featured variant for March, 2023.


By Éric Angelini


Steno-chess (SC) purposes to play with a series of simple chess instructions which are assembled into a character string.
Example of such a string : ghLxLR3xo#

Here is the complete SC instruction set (all the traditional chess rules are kept):

[One will keep in mind the difference between o and 0]

A string is read from left to right. Odd instructions are for White, even instructions for Black. Now, let us look again at the example on top of page. The string ghLxLR3xo# should be read like this (in absence of other indications, the string of instructions starts with the first move) :

  1. White has to move a piece in order to occupy a square on file g (either capturing or not)
  2. Black has to move a piece in order to occupy a square on file h (either capturing or not)
  3. White has to move a Bishop (either capturing or not, either giving check or not)
  4. Black is forced to make a capture somewhere on the board (x)
  5. White has to move a Bishop (either capturing or not, either giving check or not)
  6. Black has to move a Rook (either capturing or not, either giving check or not)
  7. White has to move a piece in order to occupy a square on rank 3 (either capturing or not)
  8. Black is forced to make a capture somewhere on the board (x)
  9. White has to castle on the kingside (o)
  10. Black has to checkmate his opponent (#).

Steno-Chess is not a new standard for describing chess positions; its aim is to play with strings of instructions which may seem, at first glance, weird or contradictory. As one will quickly notice, the “standard ASCII” SAN notation (or even the Chess Informant’s one) can lead to several SC strings. For example, the short game:

1.g4 h5 2.Bg2 hxg 3.Bxb7 Rxh2 4.Nh3 BxB 5.0-0 Rh1#
... could produce, among others, SC strings like ghLxxR3xo1, 45ghLxNLoh or ghLxLR3xo# (the latter being the one which opens the page). The aim of Steno-Chess, in reality, is to play with the assumed blur of its own set of instructions. How about, for instance, trying to replay the following game? But take care: the seemingly haziness attached to some moves disappears when you look further in the string. Every move is forced, in fact – and this is the beauty (?) of the exercise: dg6LxgxK+x5Nx4+NxLgx+N1+x2xg+5ebxR6+dQNQg4~#

Further remark. Nicolas Graner (a friend of mine) warned me, when we had our first discussion about SC: “With your system, it is impossible to describe the move < 1.a3 > because the notation < P > doesn’t fit (too ambiguous – there are eight possible pawns), neither < a > (three movements are possible, bringing a piece on the a file), nor < 3 > (twelve movements – between Knights and Pawns – fulfil the requirement of occupying a square on the 3-rank)” I answered him that SC are not to describe a game or a standard sequence of moves (as we have seen), but to build labyrinths through which only one path leads to the exit. Nicolas’s remark, however, was still challenging me... So I tried to find SC strings (the shortest ones) in order to fix all the twenty white moves which can open a real game.

Here is the (almost) full display. A single line should be read like this: “If you obey the string at the right of the arrow, you cannot play another first move than the ASCII one at the left”.

1.a3 ——> 3e2x
1.a4 ——> ab3x
1.b3 ——> be33
1.b4 ——> b~5
1.c3 ——> ?
1.c4 ——> cb5x
1.d3 ——> 3gK+
1.d4 ——> dex
1.e3 ——> 3f+
1.e4 ——> 4f+
1.f3 ——> 3~~#
1.f4 ——> 4Nf+
1.g3 ——> gd23
1.g4 ——> g~5
1.h3 ——> 3d2x
1.h4 ——> hg5x
1.Na3 ——> a~1
1.Nc3 ——> c~1
1.Nf3 ——> f~1
1.Nh3 ——> h~1

Here is the same exercice for the first twenty black moves:

1...a6 ——> NaNx
1...a5 ——> bax
1...b6 ——> ab5x
1...b5 ——> abx
1...c6 ——> ?
1...c5 ——> bcx
1...d6 ——> c6+
1...d5 ——> cdx
1...e6 ——> ~6~#
1...e5 ——> dex
1...f6 ——> ?
1...f5 ——> efx
1...g6 ——> hg5x
1...g5 ——> fgx
1...h6 ——> NhNx
1...h5 ——> ghx
1...Na6 ——> eNx
1...Nc6 ——> ~c~8
1...Nf6 ——> ~f~8
1...Nh6 ——> dNx

We will leave to the reader the pleasure of finding the 4-symbol strings above which were replaced by < ? >. Answers will be posted soon here... We would be delighted if some composers should adopt the Steno-Chess notation. This would enrich the beautiful domain of no-diagram chess (link in French).

Thanks to Gilles Esposito-Farèse for his remarks and for suggesting the % symbol. Many thanks too to Mario Richter who suggested a lot of improvements in the English notation of promoted pieces – and for allowing the King to castle when it has to obey the < K > instruction! Mario checked and corrected a lot of the above strings – thanks again! And thanks too to Olivier Pucher for the = symbol. French original version of this page, click here. Main page of this site (in French), there.