Check out Chess with Different Armies, our featured variant for July, 2024.

This page is written by the game's inventor, Charles Gilman.

Nearlydouble Chess

In Honeycomb Chess each rank is a triangle of side 5, comprising 15 cells. I fill each end rank with a King and twice FIDE numbers of other symmetric pieces, and the next ranks in with Pawns. I wondered how I could deploy a set of nearly twice FIDE size in 2 dimensions. One possibility was to lay the 15 files out flat with the King file in the middle, a Queen file either side, and the rest in pairs beyond, but that seemed a bit unoriginal as well as weakening Bishops compared to Rooks.

Then I considered a square board. 11x11 would not be far short of twice 8x8 - in fact, just a single cell larger than the Honeycomb 3d board or its flattened version. That gave Bishops and Rooks the same maximum range of 10 steps and an overall piece density of 51.24%, very near the 50% of both FIDE and Honeycomb.

Next came fitting all but half the Kings of two sets into sensible camps. As the back rank had room for only eleven of the fifteen symmetric pieces aside I gave the Knights (actually Catapults, see below) the second rank to themselves. This pushed Pawns forward, suiting the increase in total ranks, so I filled the fourth rank with Pawns and put the remainder on the third. This gave a front rank entirely protected by at least two pieces.

Since this page I have similarly enlarged other established games - see Nearlydouble Wildebeest, Long Yang, Strong Yang, Shogi-set Nearlydouble Variants, Nearlytriple Raumschach, Nearlydouble Hex variants, Bachelor Nearlydouble, Half Nearlydouble, Fragnurasian Qi, and Small Game Nearlydoubles. All these have odd numbers of files, with the King on the centre file other pieces placed symmetrically around it. In retrospect symmetry, of the kind found on this page, was one of the goals of Nearlydouble.


Face-to-face version:

Corner-camp version:
These arrays serve equally well for Nearlydouble versions of other variants using the FIDE array (and no River, 11 being prime) such as Bishogi, my Modest Variants, and (without the Nearlydouble enhancement to the Knight) Carnival of the Animals.


As the enlargement weakens Knights compared to linepieces I enhance them to CATAPULTS - a name explained in Man and Beast 08: Diverse Directions - by adding a 3:0 leap as unblockable as the 2:1 one. This is in preference to other possible enhancements. I quickly rejected the Gnu (Knight+Camel) as it lacks the original Knight's colourswitching character. The Gazelle (Knight+Zebra) took longer to reject but eventually I judged the Zebra's leap, more than half as long again as the Knight's, too long on a board less than half again the FIDE one's dimensions. For nearly twice the total size of board means nearly root-2 times the dimensions, making a choice of adding a move just shorter than the Camel's eminently better.

Other pieces are, as regards their usual moves, the FIDE ones.

Since posting this variant I realised that for both arrays all pieces of a kind can be replaced with corresponding ones from strengthened-piece 8x8 variants. Three come to mind, two based on later variants of my own and one based on an earlier variant by Ralph Betza:
* Nearlydouble Overkill Chess substitutes Aces for Queens, Marshals for Rooks, Queens for Bishops, and Cardinals for Catapults;
* Nearlydouble Tripunch Chess substitutes Combines for Queens, Reapers for Rooks, Harvesters for Bishops, and Nightriders for Catapults;
* Nearlydouble Quadripunch Chess substitutes Werewolves (Wolf+Fox) for Queens, Wolves for Rooks, Foxes for Bishops, and Double Rhinos for Catapults.


Pawns have a Eurofighter double-step move (see Man and Beast 02: Shield Bearers) from their own third or fourth rank - but not both for the same Pawn as such a move from the third rank takes the Pawn to the fifth rank. This move can comprise two noncapturing steps orthogonally forward, two captures diagonally forward, or one of each in either order. In the corner version the move is available from any Pawn starting square. An enemy Pawn can capture En Passant a Pawn making any double-step move, although if the second step was capturing the captured piece is not recovered.

Castling involves moving either the King to the inner Bishop square and the inner Rook to the Queen square, or the King to the outer, and the outer Rook to the inner, Bishop square. Usual restrictions apply. This extends to Marshals and Reapers.

Promotion in variants using FIDE pieces and Catapults:
* A player with fewer than four Catapults can either promote a Pawn reaching the enemy front Pawn rank to a Catapult which cannot later be promoted further, or leave it a Pawn that can.
* A player with fewer than four Bishops can either promote a Pawn reaching the enemy rear Pawn rank to a Bishop which cannot later be promoted further, or leave it a Pawn that can.
* A player with fewer than four Rooks can either promote a Pawn reaching the enemy Knight rank to a Rook which cannot later be promoted further, or leave it a Pawn that can.
* A player can promote a Pawn reaching the enemy King rank to any piece except King, with the restriction that only if they have at least four of every type can they promote to a type of which they have four or more.

Promotion in variants substituting other pieces substitutes them in the same way.

For a Nearlydouble Carnival of the Animals, the board suits longer leapers than in normal-equipment Carnival of the Animals. Therefore I would suggest interpreting the higher-scoring of the two dice - or one but not both when they score equal, as one higher than it actually is. This would mean that rolling 1/1 gives a Knight, 2/1 a Camel, 2/2 a Zebra, 3/1 a Giraffe, 3/2 a Charolais, 3/3 an Antelope, 4/1 a Zemel, 4/2 a Satyr, 4/3 a Gimel, 4/4 a Rector, 5/1 a Flamingo, 5/2 a Crane, 5/3 a Chamois, 5/4 a Zherolais, 5/5 a Parson, 6/1 a Namel, 6/2 a Stork, 6/3 a Samel, 6/4 an Ox, 6/5 a Famel, and 6/6 a Curate.

Check, Checkmate, and Stalemate are standard.


It is notable that Overkill substitution does not enhance the basic Knight move. This is because the Knight is not an Overkill piece and its compounds, being with long-range radial pieces, should be strong enough on an 11x11 board. Applying Nearlydouble to 10x10 variants using the Knight as well as its compounds, such as Grand Chess and Échecs De L'Escalier, is a different matter. The resulting board - 15 files by 13 ranks, or vice versa - weakens Knights but a Catapult is not that suitable for compounding with radial linepieces. For such a size I do therefore suggest using the Gazelle that I reject for an 11x11 board - and its compounds Rook+Gazelle=Razorbill, Bishop+Gazelle=Basilica, Queen+Gazelle=Sheherezade.

A possible intermediate stage between FIDE and Nearlydouble Chess which would require neither Pawn initial moves nor, I contend, Knights, to be enhanced would be DOUBLEBACK CHESS, an idea arising from the double numbers of capturable symmetric pieces in certain hex variants of mine. This would combine Nearlydouble's 15 symmetric pieces aside and 11 files with fewer Pawns aside and the 8 FIDE ranks. I would suggest the array:

the Knights placed to preserve the FIDE feature that a single noncapturing move cannot open up a Rook or Queen file.

This 'user submitted' page is a collaboration between the posting user and the Chess Variant Pages. Registered contributors to the Chess Variant Pages have the ability to post their own works, subject to review and editing by the Chess Variant Pages Editorial Staff.

By Charles Gilman.
Web page created: 2007-03-08. Web page last updated: 2016-03-09