I posted a couple of variants here -- Scirocco and Typhoon -- about a
decade ago. Both are heavy on short-range pieces, and if you'll excuse my
immodesty, I still like the way they look. They might play well, too, but
I've never played them except against Zillions, and I don't know of
anyone else who's played them at all.

Scirocco is a 10 x 10 game, which seems about the right size to me for a
basic chess variant. Typhoon is 12 x 12, with the same number of pieces
(in the same relative positions) as Chu Shogi, which is apparently more
than most people are willing to take on, although Zillions playing a game
this size against itself makes wonderful wallpaper. I've never seen a
draw in either game.

With Scirocco I included an 8 x 8 variant, Shatranjirocco, with the same
number of pieces (of about the same strength) as in Shatranj. This game
*is* drawish, probably almost as much so as Shatranj.

The points I'd like to make are:

. I think Mr. Joyce is right in saying that 8 x 8 games are, as a rule,
too drawish. You have to introduce something like drops in order to make a
game that size complex enough to hold human interest for more than a few
centuries.

. And I agree with Mr. Joyce again about FIDE chess. Part of the problem
here is that the stronger pieces are too much stronger than the weak
pieces; too often it's hard to arrange exchanges of approximately equal
value (except of pieces of the same type), so from many positions it only
makes sense to play defensively. If the pieces available to exchange are
closer in value, it becomes a lot more plausible to exchange material for
a positional advantage.

. For all the interest weaker pieces have in the opening and the midgame,
they are tiresome to try to checkmate with, at least if you're using
traditional, unconfined Kings and no drops. For that reason I like the
Shogi-style every-piece-promotes-in-the-end-zone convention: it all but
guarantees that even large variants come to an end in a reasonable amount
of time. (I don't care for the promtion-on-capture rule of some of the
giant Shogis, because it puts too much power on the board too early and
obscures the role of the weaker pieces.)

. There remain a number of directions in which to push the invention of
short-range pieces. The great Shogis concentrate mainly on movement
patterns: with a few interesting exceptions, they are combinations of
leaping and short ranging movements through empty spaces, in either case
capturing by displacement. However, there are other things pieces can do:
consider, for example, shorter-range versions of Ultima or Rococo pieces.
I tried to explore a little of this space in my not-quite-finished 16 x 16
variant Jupiter, but you could certainly use many of the same ideas on a
smaller board.