or, 10k on just One Game a Day
Stop me if you've heard this one, but here's an idea for a large variant that doesn't require much new in the way of pieces or rules … just a bit of a twist in the way we look at movement potential.
Consider a very boring game of 10x10 Chess, with a couple extra pawns, a couple extra middle rows, and extra knights between the bishops and rooks. Yawn, ho hum, et cetera; not even a nightrider or an amazon to play with. Imagine next that you go to all the trouble of specifying movement rules this way: A rook can move (v: +/-x, h: 0) or (v: 0, h: +/-y). A bishop moves (v: +/-z, h: +/-z). A pawn moves (v:+1, h:0), or (v: +1, h: +/-1) to capture, and a king moves (v: +1/0/-1, h:+1/0/-1), and a knight either (v: +/-2, h: +/-1) or (v: +/-1, h: +/-2) … again, ho hum, right?
Well, now instead of a single 10x10 board, take 100 of them and lay those out in a 10x10 square. We specify the locations of the pieces with a handy binomial, "a1-A1" meaning "the square in the lower left-hand corner of the board in the lower left-hand corner" (where there's a white rook in the picture) and "d7-H4" meaning … well, you get what it means. There's a magenta knight there in the picture. Anyway, then we let the movement-and-capture rules apply to either half of the binomial, but only one half at a time.
So the whole thing looks like this:
There you have the white pieces at the bottom, in position to start a new game. Above them, a green king, blue bishop, yellow rook, cyan pawn and the aforementioned knight disport themselves, showing off the squares they can attack. Not, perhaps, the most profound piece of theory in the world. But it provides an imposingly large game and a challenge to the players' panoramic vision of the board, without adding unduly to the number of pieces or piece-types. Plus you might be able to get a game in and have the rest of the weekend free!
Long-range pieces ("riders") should be limited to x, y, z < 10, to make their new powers uniform with the others'.
Pawn promotion should take place upon reaching any square in the 10th rank of boards, not the 10th rank of squares nor (good heavens!) the very last rank of squares.
Rules for Castling could no doubt be worked out and put to use. Aside from the sheer thrill of it, though, I have no idea why you'd want to; the extra protection to be gained thereby would be practically nil.
Capturing en passant is a fairly recent refinement to the laws of FIDE Chess, meant to address a very delicate problem of balance. When Taiga or something like it has been played for a century or two, and the odd million games exhaustively analyzed, if someone cares to propose a similarly precious revision of its code I shall wish them well. But for the moment I think we can cripple along without it.