Check out Glinski's Hexagonal Chess, our featured variant for May, 2024.

This page is written by the game's inventor, Patrick O'Neal.


By Patrick O'Neal

I have been convinced of the superiority of shuffle-chess since shortly after learning the game, as it avoids the reams of memorized and recorded opening-theory.  The three best forms of shuffle-chess, in my opinion, are standard Shuffle-chess (no castling), Shuffle-chess utilizing the King's leap for castling, and Fischer-random.

Each one of these has it's drawbacks:
1. Standard shuffle-chess, without castling, leaves out an exciting element.
2. Shuffle-chess, with the King's leap, has the drawback of possibly of restricting castling to only one direction if the King is on either the 'b' or 'g' file, and if a Rook is on the 'a' or 'h' file at the same time, its action is limited with no hope of improving via the King's leap.
3. Fischer-random has the drawbacks of limiting the number of possible starting positions to 1/3 of shuffle-chess because of the requirement of placing the King between the Rooks, and also the castling seems to be a little  'fudged,' by designating the 'c' and 'g' files as having a magnet to attract the King, and the 'd' and 'f' files a magnet to
attract the Rooks.

My idea, I think, eliminates these drawbacks, and allows for a superior game of shuffle-chess.

The game is played exactly under the same rules as standard chess, but with the difference that the initial position of the pieces on the back rank is randomly chosen, with the Bishops occupying opposite colored squares as in standard chess, and Black's position must be a mirror image of White's, also as in standard chess, to avoid any unfair advantage to one side.  This allows for 2,880 possible starting positions; three times as many as Fischer-random.

The only other difference is the castling rule, which is:

The King is allowed once during the game to castle, which is done by 'sliding' the King, either left or right as far as it can go along any vacant squares on the back rank to occupy the square adjacent to another piece on the back rank.  The piece that the King slides next to is then  moved to the adjacent square on the opposite side of the King; that is, it 'leaps' over to the square on the other side of the King on the back rank.  The piece involved in the castling can be any piece, not just a Rook. This is accomplished under the same rules as standard chess-- the King may not pass through check on its 'slide;' the King and the piece involved in the castling must not have previously moved; etc...  Under this castling, if a Bishop is the castling-piece in question, it presents no problem, since it will occupy the same color square after castling as it did before castling.  It is also possible to castle without moving the King, e.g., a piece already sitting next to the King on the back rank may 'leap' over to the other side of the King if that square is vacant.

This version of shuffle-chess eliminates the drawbacks that I have mentioned.  I believe it to be the best form of chess.  It is interesting to note that conceivably a typical standard game can result if the standard setup happens in the randomizing process.  In that position, a King-side (h-side) castle, utilizing the Rook as the castling piece, would create a typical standard-game castled position.