Additional types of halflings are also possible.
For example, there is the Floor Halfling, that moves up to half as far as the corresponding normal piece, rounded down. You could not make a piece that had Floor-Halfling powers and no other powers, because it would be unable to move to the edge of the board (except when moving from one edge square to another). A Relative Floor Halfling could never make a capture! However, a Floor Halfling power could be combined with a more normal power to make a reasonably useful piece.
In addition, there is the Anti-Halfling, as per http://www.chessvariants.com/d.betza/chessvar/pieces/anti-r un.html In essence, an anti-halfling can move to all the squares a Rook could reach but which a Halfling cannot reach. It is a whimsical power, of course.
By special decree of the Chess Variants Rules Supreme Court, Castling is legal, and the Pawns, if they have not yet moved, may make a two square advance, as in Chess. The plaintiffs argued that these moves were special cases, not to be counted as distance moves for the purposes of halving, and the majority of the Court concurred.
It is amazing how a simple and obvious chess variant often turns out to have subtle ramifications when you examine it closely.
For example, in Halfling Chess, after 1. e4 e5, the normal and natural 2. Nf3 is subject to the theoretical objection that one is developing too powerful a piece too soon, which will make it subject to later harassment from weaker pieces -- and therefore the theoretically correct move must be 2. Bf1-c4. In fact, since the only weaker piece in this game is the Halfling Bishop (which is worth approximately 1.5 Pawns), the objection is not so acute.
Because Halfling pieces have trouble reaching the edge of the board, the shortest possible game of Halfling Chess that ends in mate is with 1...Nc6 2...Ne5 and 3...Nd6 or ...Nf6; and the shortest mate not given by a Knight is 1 d4 e5 2 Qd2 Ke7 3 Qf4 Ke6 4 Qe5.
After 1 e4 e5 2 Bc4, the B at c4 does not attack f7; in order to do so, it would need to go to d5 first, but then ...c7-c6 would chase it away. Therefore, a reasonable opening might be 1 e4 e5 2 Nf4 Nc6 3 Bc4, threatening Bc4-d5 -- and in fact it is a disruptive threat which makes me wonder if 1...Nc6 isn't simply a bad move!
Instead, one might try 1 e4 e5 2 Nf3 Nf6, and after 3 Ne5 Ne4 4 Qe2?! Nf6 5 Nc6??, Black is not in check because the Q can't go that far and can simply eat the N. So, 1 e4 e5 2 Nf3 Nf6 3 Ne5 Ne4 4 d4 d5 5 Bd3 Nf6 6 O-O Bd6 7 Nf3 Be6 8 Re1 Bg4 9 Nf3-d2, neither player is getting any advantage. The open file is unimportant, and advancing with Knights first simply loses time when the Bishops drive the Knights back.
Given the pattern of the way the Halfling Bishop moves, a finchetto development seems attractive; it may turn out to be tactically useful to be able to move to the edge of the board. 1 e4 e5 2 g3 d5?! 3 ed5 Qd5 4 Bg2 attacks the Q, but the Q does not attack the B; or 1 e4 e5 2 b3 Nf6?! 3 Bb2 Ne4 4 Be5 d6 5 d3 might be good for W.
The pieces move in the same directions, but with different lengths of movement; and the game is completely different.
The statement in the previous page, which said they were worth 0.6 to 0.66 as much as a normal piece, was incorrect, and was caused by incorrect figures in the short rook page. (http://www.chessvariants.com/d.betza/chessvar/pieces/short rook.html)
A Halfling is worth half as much as a normal piece.
Once you understand the value of the zFF (crooked Bishop), then it becomes clear that in my example analysis on the page that introduced the Crooked Bishop, I got bad results because I was using it incorrectly. Save it for the endgame! Once the board opens up, the Crooked Bishop will spread fear and terror everywhere among the enemy ranks.
A Halfling Crooked Bishop is therefore worth half as much as a normal Rook; in other words, it's a piece that's worth about 2.5 or 2.25 Pawns.
The Halfling Rose should not be used on an 8x8 board, of course; but the Rose and the Halfling Rose will be popular pieces when Hans holds his contest to design a chess variant on a board of 169 squares.