Game board (6x6x4)

In deciding on the 6x6x4 board as the main one for this game, there were various canditates which are found among other 3D chess variants. The most obvious ones are 5x5x5 and 8x8x3. 8x8x8 is too large, the game would take too long and nobody will play this game. Conversely, 8x8x3 feels like "2D Chess with layers" at times, not having enough vertical z axis movement. 5x5x5 is what Ferdinand Maack (mathematician and designer of Raumschach, the first 3D chess) settled on. 5x5 boards (or anything smaller) feels too cramped for my tastes though. Raumschach also has an odd number of squares, which also seems off. So, I began looking at other ideas, based on stacked 6x6 boards and settled on a playspace of 6x6x4 (144 cells).

At first I thought about 6x6x6 (216 cells) or 6x6x5 (180 cells). 6x6x6 feels best aesthetically, since it is nice to have even symmetry in all axes. However, we now have a play space that is still very large (more cells than Heian dai shogi!). The taller a board space becomes, the more difficult it is to store, manufacture and play in. I also think the more boards you stack, the more intimidating and demanding the game becomes. This has implications for how many people are willing to play it at all. It is also why the most popular commercial 3D variants tend to be 3 or four levels. One only needs to look at the various 8x8x3 and 4x4x4 commercial variants that have been produced, and the Star Trek Tri-D chess (4x4x3, plus 4 small "attack boards").

6x6x4 seems like a decent compromise between the boards that feel too small and the larger ones. All axes being equal is not really necessary. Several classic 2d variants like Courier chess and Xiangqi don't even have equal axes in 2d and they are eminently playable. 6x6x4 is also simpler and less expensive to make than bigger stacks. One could even buy one of the standard 8x8x3 boards and modify it, placing one more chessboard below or above the bottom board, and blacking out the outer squares to get 6x6 boards. It's also nice that all axes are made up of even numbers. This was my thought process in choosing the 6x6x4 board.

Pieces and movement

Various writers on 3D chess, like Ferdinand Maack and V. R. Parton (see his Chessical Cubism or Chess in Space), have said that when looking at a 3d cubic space, there are three basic movements a piece can take three-dimensionally. Imagine lattice of cubes, pick a cube with a chess piece in the middle of it, now there are three ways to move into another cube in the lattice:

Moving into one of the faces of the cube

Moving into one of the edges of the cube

Moving into one of the corners of the cube

These three form a natural basis for 3D movement design since they correspond exactly to adjacency relations in a lattice of cubes.

Before we begin to compose our army of cubic chessmen based on these three main movement types, let me make two observations that I have noticed in my cursory study of 3d chess variants.

In a cubic 3D lattice, obstruction is possible but rarely causes major blockages or problems. Leapers cease to be necessary or very interesting. Furthermore, leaping type movements are one of the hardest for newcomers to visualize, and there are numerous possibilities for a "3d knight". Yet, since making a "3d knight" doesn't seem useful or necessary given the wider range of movement in a cubic lattice, there is no point is adding them, especially if they will be confusing to use for players.

The orthodox pawn collapses under 3D conditions. The pawn’s restrictions only make sense in a flat, frontal battlefield. It was designed with this idea in mind. I noticed that some designers discuss 3D pawn movement with the idea of mimicking in some way the classic way pawns are used. Ralph Betza agonizes over how to make 3d pawn chains and so forth. But what if we abandoned this completely? 3d chess should not be seen as mimicking a 2d battle. It should be seen as a three dimensional conflict full stop. The closest real life comparisons would be outer space or submarine warfare. It is indeed suitable that the first true 3D chess we have the rules for was called "Space chess". Thus, in Astro chess, there are no "pawn chains" just like there are no shield walls in a space battle, a dogfight or a submarine battle. This means pawns can have different kinds of movement. Or we could do away with the idea of having a bunch of "pawns" (a bunch of weak pieces) altogether, and just make an army with different kinds of pieces of more equal strength. Many possibilities present themselves here and I will present one configuration below.