Comments by DavidCannon

I feel flattered by all the 'excellent' grades people gave me, but at that time, there were not so many games posted on the CV pages so people didn't have much to compare it with. I would probably rate it as 'below average' or 'poor' myself now, which is why I've never programmed it and have never made any diagrams. But this was my first serious attempt to design a game, and I've left this page standing to remind me of where I came from.

I did use the idea of pieces merging with other pieces for many of my other games (Diamond Chess 306, Dürer's Chess, Matrix Chess). You could saw that I see this game, not so much as a game worth playing, but as more of a source of raw material for other games.

Another thing you might notice: most of my games allow pieces to merge with other pieces; this one does not. That is deliberate: as this board was designed with the Bishop in mind, I want to let the Bishop shine without merging with other pieces.

I wouldn't worry too much about 'stealing' others' pieces. Anything completely original wouldn't be a chess variant - it would be something else entirely! Of course, you can't just copy the game in toto and present it as your own, but there's nothing wrong with borrowing ideas, as long as you acknowledge the source. I did this with a number of my games, such as Diamond Chess 306. The tiling pattern and many of the pieces are taken from Parachess, an earlier game by Tony Paletta. I did invent a number of new pieces, but many of them are logical extensions of Paletta's ideas. I briefly acknowledged my debt to him (and to Fergus Duniho, for some other ideas) on the web page, and in more detail on the information file that I included in the Zillions package.

One weakness, as I see it, with a lot of 3D games is that the King is too mobile, making checkmate problematic. We'd all do well to explore ways to address this.

There are a number of ways to increase the number of pieces without getting too complicated. One way is simply to have more of the familiar range of pieces - e.g., four Rooks instead of two, etc. Another is to make new pieces with new moves, or to borrow pieces from other well-known variants (such as the Elephant and the Cannon from Korean Chess). A third way is to have compound pieces (a number of my own games explore this possibility).

I appreciate your offer to 'host' both Hafsteinn's version and mine (which is still on the drawing board). I'll have to crystallize my ideas a bit more and get back to you about that as soon as possible.

If you want to experiment with your ideas, we've got quite a few Zillions experts around here. (I know a little Zillions programming, but I'm no expert and it takes me weeks to produce a functional game, but there are others here who could do a better job much more quickly). If you have a Zillions implementation and let the computer play against itself, you can see what goes on and analyse it - which I've found very helpful.

I think I spoke a little too hastily last night. I've seen a few too many games that seemed just slapped together, but now that I've taken a look around some of your work, I can see that's not what you're about. It's probably a good thing that you've submitted your 'first draft' raw for others to look at, and I'm sorry I shot from the hip last night.

1. 3D chess creates extra paths on which pieces may move. The most obvious of these is the so-called 'trigonal' path (like a diagonal path, but not colour-bound); various variants have introduced a UNICORN or a MACE as a line-piece to move on this path. A variant that misses this is really lacking something important, I feel.

2. One weakness with the starting position is that a couple of simple moves by the Rook or the Queen can check the opposing King. As white moves first, this gives quite a head-start to white, which is not fair.

3. The piece density is too low. FIDE chess has 32 pieces for 64 cells (50%); Shogi has 40 pieces for 81 cells (almost the same, although Shogi's pieces are somewhat weaker). Changgi (Korean Chess) has a lower density of 32 pieces for 81 points (40%). Now, there's nothing sacred about these percentages, but they have stood the test of time. Having designed quite a few games and playtested them on Zillions of Games, I've found that Changgi's 40% density is close to the lower limit at which one may play a satisfactory game. With too low a density, the players just chase each other around the board forever. (My own Diamond Chess 306, whose Zillions file you can download from this site, has a 38% density, but in two of the variants each piece is really a three-piece compound that can be unpackaged). The popular 5x5x5 variants have a 16 percent density, which I find too low. Your own 32 pieces per 320 cells is only a 10 percent density. You've got to be joking.

Now, some suggestions for you, Hafsteinn: 1. Either borrow a Unicorn/Mace-like piece from other variants to ride the Trigonal path, or modify one of the existing pieces to utilize it. 2. Give us something most of the existing 3D variants haven't got. Leapers, for example. Most existing variants simply extrapolate the Knight's move (one orthogonal step plus one diagonal step) to the 3D board. How about bringing in some new leapers that cover the orthogonal plus trigonal, and diagonal plus trigonal, steps? 3. Increase your piece density! Either reduce the size of the board, or increase the number of pieces. (I know the solution is problematic - I'm working on the idea myself, and it's tying my brain in knots - but we've got to try).

You may ask why this fussy distinction. There are two reasons. The first is that I could either allow every piece to assimilate every other piece, or I would have to draw the line somewhere. The first option would create such a huge number of pieces that even I wouldn't be able to remember them all. The second option, which I chose, means that some arbitrary distinctions have to be made. I settled on confining assimilation to within 'families' - and in the case of families whose membership partially overlaps, a certain degree of subjectivity creeps in.

The second reason for this is programming. There were a number of things I would have liked to do, but the limitations of Zillions (or, rather, of my knowledge of it) forced compromises. The case you mentioned is one, although I think I would now be able to work around it if I was so inclined. I may may tweak the program that way when I get around to it.

'As far as I know, that remains intact are the Knight, the
King and the pawns.' Not so fast, mate.  I believe the Rook has remained fundamentally unchanged throughout the entire history of chess.

Anyway, this game, along with the associated Zillions page, is now ready to go live:-)