3 Dimensional Cubic Chess 8x8x8 Board

Nah that's weird, only 1 king regardless of the number of dimensions lol

Another interesting question, how many 3d queens needed to control every vacant cube on the board. With 2d we know it's 5 ("The mathematics of chessboard problems" is an interesting read)..

I think the history of this was more just people would play over boards just spread across the table. So 8 boards with all those extra places to move. The link you provided in chess variants was kinda of cool and it warms my heart that people are into that stuff. Of course, 3d chess is actually quite easy to play online tho : https://store.steampowered.com/app/3463010/ChessFinity_Demo/ Where you can play against your friends for free in a medley of different sizes. In my experience and I do play a lot of 3d chess, an 8x8x8 board is just too big , too time consuming and rather confusing without having to use tooling for marking grids and move preview modes being turned on. A comfortable size is around an 8x8x3 with parties staring at the 1 and 3 levels Anyway. Enjoy

What even is this game!?!?!?!?!

I would have to disagree that uneven dimensions are better than even 4x4x4 for example. Part of what makes chess such a great game is the square symmetry of the board. Positions without pawns are exactly equivalent regardless of reflection or rotation. I'm not saying this always isn't the case either with uneven dimensions, but it makes it harder to visualize basic patterns and piece coordination if the distance to the edges of the board are uneven. At least with the 4x4x4 board you can have 1 total set of 3D pieces that take up the same proportion of the board spaces, but pawns would get a little funky.

A 3D queen is basically a scaled up 2d queen in the simplest form. It can move any number of cubes in any straight line, including all the new diagonal paths and 3D diagonals. But, there are soo many more cubes. Any patterns that each queen fails to cover in 2d is only multiplied enormously in 3d. Board space expands much faster than piece power rises as you add dimensions. Max control of 27/64 squares in 2d but only around 90 out of 512 in 3d. The "lines of control" spread out in more directions, but there are far more gaps between said lines on a 3d board! In fact think of it this way, a 3d king can move to as many places as a 2d queen!

Depends. Should the King have the adjacent diagonals? If not then its only the same as a queen with only 1 propagation. So ; 8 + 5 + 5 = 18 If it has the AD's then its 8 + 9 + 9 = 26
Its worth bearing in mind that not all boards are symmetrical and pieces don't need to have infinite movement.

Yes, all pieces, but especially the queen and king should have all 3D type moves included. Even just for the reason of how "king and queen moves" are defined. The king can move to any location that it is next to, regardless of direction, which obviously includes "triagonal" jumps through the corner of a cube. The moves a king has completely surround it in an enclosed space. The only difference with the queen is that the distance is unlimited, but the same directions as the king apply. If you start omitting some of them then it's not even behaving like a king or queen anymore, regardless of dimensions. In fact in any dimension (label it n) the number of moves a king has from a central position on the board will always be (3^n - 1). There may be a way to calculate this with the queen too I'm not sure. Whether triagonal moves should be applied to bishops knights and pawns is up for debate, but the king and queen, just from the definition of how they move, will always include any and all directions.

The reason other pieces are more debatable is because including 3d diagonal moves with them can make the range of locations they cover, bizarre. For example a bishop that moves in a flat diagonal move (in any of the 3 planes) where only 2 of the coordinates change, AND also moves in the 3d lines where all 3 coordinates change, could essentially fork adjacent pieces in the same plane. Same thing with a knight, if you allow the 1-cell perpendicular move to the initial 2-cell move to be diagonal, the L-becomes tilted and literally these 2 pieces could be forked from a knight on a different plane:

A white knight on one of the E3s exactly 1 level above or below this one, forks both the queen and king, which seems awkward. But once again, you have to take into account the actual definition of a knight move. Which isn't "L-shape". It's "can move to any of the closest squares that are not on the same rank, file, or diagonal. That is what mathematically defines a knight move. Otherwise, there would be nothing excluding this type of move:

It's an L too! But such moves are longer distance-wise. A 2d knight move will always be of length [radical 5]. Now here's the problem, a 3D knight move could theoretically include a 2-1-1 movement rather than just more available 2-1 directions to go, but such moves would be length [radical 6], so such moves now fail to meet the definition of a knight move. Also, a knight would then have Two sets of 24 moves available for a total of 48, which makes it much more powerful relative to other pieces. But going back to the bishop, here's the other quirk, if we deny a bishop 3d diagonal movement, that means a 3D queen could see a bishop that the bishop wouldn't see back! So a bishop, queen, king, and possibly even pawns need to have the same consistency in terms of move availability.

We shouldn't be discussing how the pieces, that we know from 2D, would move in 3D.
Instead we should acknowledge that we need to come up with new piece definitions, and we need more of them than we now have in 2D.

For example the 3D counterparts of the Queen:
- a piece that can move like a Queen in x, y direction, but only one square othogonally in z
- a piece that can move like a Queen in x, z direction, but only one square othogonally in y
- a piece that can move like a Queen in y, z direction, but only one square othogonally in x
- a piece that can move like a Queen in x, y, z direction (much stronger than previous pieces)

Nah way too confusing. Just simply scaling up the "any number of squares in any direction" to 3D works fine. We can have more of the standard pieces, but no bizarre new pieces.

You can have whatever you want but I doubt that FIDE or any league will be setting any rules other than what we want to. 
I have no idea what Poco is trying to illustrate with those images. 
On the queen and the king. They are not the same - when coding a pieces moves,its quite simple to just do a bishop and a rook and combine their moves - when doing a king its simpler to just do a proximity of 1. The result is the same but it still doesn't really answer if the king should have the extra 8 grids in 3d as they do make a big difference. A counter this would ironically be to make the pawns more powerful but thats a different discussion.

A kings movement is 1 square/cube in any direction, which includes all the cubes in a 3x3x3 section it is sitting in. Omitting certain 3d king moves would be equivalent to removing the kings ability to move diagonally in current chess. It's ridiculous. I don't understand why so many people think that 3D chess should just be slopping together different perpendicular planes of 2d moves but not include actual moves that utilize all 3 dimensions.

A queens 3d movement is combining a rook and bishops moves. The 3d diagonals are the bishop moves. Or at least included with their existing 2d moves. Are you suggesting that a queen totally undefended should be able to check a king 1 cube away through a corner vertex of the cube, but the king couldn't capture it?? Because that's why denying those moves to the king but allowing them for the queen would entail. As well as the absurdity of a queen being able to attack a bishop along a diagonal that somehow the bishop (by definition a diagonal moving piece) couldn't attack back.

A pawn having access to all the 3d diagonal moves on top of the 2d ones (in terms of capturing) would be 100% consistent with normal chess. An upward-shifted king move, with the proportions being exactly equal. 3/9 squares the king could go to from its position, 9/27 cubes the 3d pawn could go out of the respective king moves.

I would hope that people who propose rulesets for 3D chess variants, stick to the same names for pieces that do the same thing.
To give an example of this, how many times, and under how many names, did somebody invent the Bishop-Knight compound for 2D?

Therefore, at this moment, when it comes to 3D, a nice next step would be to come to a set of definitions that yourselves and future inventors can use.

People can disagree with the ideas that I read here, there are more possible approaches to extending 2D chess pieces to 3D boards.
One could for example state that in 3D, any classical piece can do it's own movement rotated into the additional dimension by 90 degrees. That approach would rule out the space diagonals.
But then one would want to come up with new pieces that can actually utilize these space diagonals.

The rook movement is easy to imagine in 3D space, but bishop get complicated.

But again why should space diagonals be excluded from the normal pieces to begin with? I understand if you want to add a bishop that can only move in space diagonals and is thus restricted to 1/4 of the board (would need 4 of them), but why do people want the king to not be able to move to a cube it touches, such as through corners? Space diagonals are a fundamental part of a 3d board, so why should those weirdly be ignored? There are 3 possibilities for a bishop. One that only utilizes the flat diagonals within any of the 3 planes it is in. One that only utilizes space diagonals and not any 2d ones. And one that uses both. As weird as it seems, I think the typical bishop should use both. Think of how a king and bishop on top of a king at the edge of the board stalemate it. If both the king and bishop (and even pawn) are free to utilize all 2d and 3d moves available, the same stalemate exists on an edge-plane of the 3d chess cube. Here's an even better way to think about it, if a queen and king couldn't move in space diagonals, than a basic support-mate with a queen right on top of a king couldn't exist! The king could always move in a 2d diagonal that puts it 1 space diagonal away from the attacking queen, so even if a queen is right on top of the king, it couldn't checkmate it even in the corner! Kings could be 1 space away from each other which is awkward.

The rook is one of the weaker pieces in 3d chess given how each extra dimension only adds 7 more moves to it where almost every other piece it gets multiplied. But it's important to note that the power ratios of each piece (like a rook vs bishop) doesn't mean the weaker piece can be trapped easier in 3D. Sure, a bishop with both 2D and 3D movements has a whopping 60 positions it can move to from the center of a 3d board, while the 3d rook only has a measly 21. But the rook can still easily avoid this bishop due to all those extra cubes in 3d. A 3d bishop simply controls more directions than the rook, but at the same time there is a much larger volume of open spaces between those lines of control on a 3d board. In 2d chess a bishop and pawn can trap a rook on the edge near a corner. In 3d chess they can't trap a rook anywhere despite being nearly 3x as "powerful" as the rook.