The Chess Variant Pages Menu

Butterfly Chess


This large board (10x10) form of chess includes Butterflies (in place of knights), Advancers (in place of queens), and Flying Dragons. I came up with the game after going through long lists of piece types that had already been invented, i.e. as found on wikipedia, besides under Piececlopedia on

Butterflies are fairly close in value and in their powers to knights, yet they have more novelty to them as a piece type, which appealed to me in trying to invent a 10x10 variant. The piece type was also appealing since, on such a board, bishops are definitely a bit stronger than knights on average (after some value estimating calculations of mine, butterflies also seem to me to be a bit weaker than bishops on a 10x10 board, too, but they are still interesting pieces). After not liking any number of often seemingly complex, overly powerful or overly weak chess variant piece types during my search, I found that Flying Dragons seemed to suit my needs too, as is explained in the Notes section further below. Advancers were the sort of roughly equivalent substitute for queens that I was also looking for in a piece type, in terms of their value, and in having more novelty. Hopefully the result of my labours is a rather refreshing and enjoyable variant to play.

A Game Courier preset for play is available.



In Butterfly Chess there are 7 piece types. Pawns, Bishops, Rooks and Kings are used, as in chess, and there are 3 other piece types:

the Butterfly - moves as a knight, except it can move one space diagonally backwards instead of being able to make the knight's two rear-most moves, and it can move like a camel sideways in a forward direction instead of being able to make the knight's two sideways moves that are also forward. Ralph Betza alluded to such a piece type's way of moving in an article on his Mushroom piece type (Butterfly powers are diagrammed further below, too): Mushroom

the Flying Dragon - moves like a bishop or camel, and is also known as a caliph; Charles Gilman discusses this piece type in the following link (Flying Dragon powers are diagrammed further below, too): Caliph

the Advancer - The Advancer's method of movement and capture are different from each other: The Advancer moves as a chess Queen - that is, the Advancer can move any number of squares in any direction, orthogonally or diagonally, as long as all the squares it passes over are unoccupied by other pieces. The Advancer can never move to a square occupied by a piece of either side. To capture, the Advancer must move directly towards an opposing piece, and end its move on the square adjacent to that piece. Again, the Advancer can cross any number of unoccupied squares in any direction to make its capture. Note that if the Advancer moves directly towards a piece and lands adjacent to it, capturing that piece is not optional - the piece must be removed. Ben Good discusses this piece type: Advancer

Here's a diagram showing all the legal moves (i.e. leaps) that a Butterfly can make on an empty board - note that forward movements are not symmetrical with rear ones, and that the pattern as a whole looks like a butterfly (as Betza put it):

Here's a diagram showing all the legal moves that a Flying Dragon can make on an empty board - note that the flags show where it can move to when acting like a bishop, and the stones show where it can move to when leaping like a camel (notice also that it is a colour-bound piece):


Pawns can promote to any piece type in the setup, except for a king. They can initially advance 2 or 3 steps, rather than just 1 step, if unobstructed. Capturing en passant, is allowed, even if the pawn took a triple step. The capturing pawn would land on the square that is diagonally in front of it on the same file as the enemy pawn that just moved. Here's a link to Omega Chess, which has very similar rules governing legal pawn moves; note the examples of en passant captures: Omega Chess

Castling is allowed, with similar conditions as in chess, though in Butterfly Chess the king would move three squares to the left, or three squares to the right, as part of making such a move, and otherwise the rook involved would end up adjacent to the king, as in chess.

Stalemate is a draw, as in chess.


I'd tentatively estimate the piece values as follows:

Pawn (P) = 1; Butterfly (Y) = 3; Bishop (B) = 3.5;

Rook (R) = 5.5; Flying Dragon (D) = 6; Advancer (A) = 9.5.

In chess a King (K) is thought to have a fighting value of 4 (noting that it cannot be traded); on a larger 10x10 board I'd put the fighting value of a K as 2.5.

Regarding these values, note I'd estimate a chess knight worth 3 Pawns on a 10x10 board, while I'd put a camel at half that value, since while it has 8 moves maximum, like a knight, it's colour-bound. Then, just as in chess a Q=R+B+P, I found my estimate of a Flying Dragon's value on a 10x10 board to be B+camel+P, or 6, as given above. Note also that on a large (10x10) board, the value of long range pieces (such as Bs & Rs) increases.

My value for the Advancer is based on the assumption that it is almost worth a chess queen (see the link given earlier in the Pieces section regarding this); I arbitrarily decided the difference between the two pieces ought to be less than a pawn, and estimated an Advancer as worth a Flying Dragon plus a Bishop on a 10x10 board.

Betza's mentioned that such a piece as the Butterfly could ('in the general case') be worth a bit more than a mushroom (which in turn he said is close in value to a chess knight; see the link given earlier in the Pieces section), and that prompted me to suppose the Butterfly as having about the value of a Bishop on a 10x10 board. However, my later calculations were perhaps not quite in line with this: just as a chess Q=R+B+P, I now reckon a Butterfly's value on a 10x10 board to be 3/4*(knight-P)+1/4*(K-P)+P = 23/8, or approximately 3, as given above.

I should note that I 'merged' the knight & camel components of the Butterfly's move in this calculation, i.e. into "3/4*(knight-P)", since the camel's colour-boundness is eliminated in a Butterfly, making a camel's component have the same per destination square value as a knight's in my estimation. That's plus adding in the K component (1/4 of the Butterfly's move), in order to obtain the final result.

Regarding my estimate for a Butterfly Chess K's fighting value, recalling that a chess K has a fighting value of 4 (even though it cannot be exchanged), this value in my view might be rather oddly expressed (for lack of a known formula) as chess K = 32 x (max. # cells chess K moves to [eight]) divided by (# of cells on a chess board [sixty-four]) = 4; similarly, a Butterfly Chess K has a fighting value of 32 x (max. # cells Butterfly Chess K moves to [eight]) divided by (# of cells on Butterfly Chess board [one hundred]) = 2.5 approx.

Note that White can threaten a mate in one move, right at move one, with 1.Dg1-f4 (eyeing the square g7), but Black has at least three defences (1...Pe9-e7, 1...Pf9-f8 and 1...Ph9-h8; perhaps 1...Pe9-e6 is also perfectly adequate, or Black can even consider moving his D on g10 right off the bat, too). Instead 1.Dg1-h4 also threatens the same mate in one, besides 2.Dh4-i7, but Black can (thankfully) defend at least with 1...Ph9-h8, when it seems White has gained little by committing his D at move one; perhaps 1...Dg10-h7 is playable, too, besides 1...Pe9-e6, with the idea if 2.Dh4-i7 Yi10-j8 3.Di7-j10 Ph9-h8, intending ...Yj8-h9 winning the trapped D, when Black may have compensation for his ultimate slight material deficit (depending if one estimates the values for this variant's pieces more or less as I do).

Regardless, I like the setup I've chosen, since bishops can be deployed in similar ways that they are in chess, and it's possible for either flying dragon to soon be developed in a way that it continues to protect the pawn positioned directly in front of its adjacent bishop. I also liked that rooks are almost fully worth flying dragons on a 10x10 board (in my estimation), which adds one more type of near equivalent 1 for 1 piece exchange that a player could possibly make during a game (besides B for Y).

By way of comparison, the original version of 10x8 Capablanca Chess had a similar 'defect', in that White could threaten a mate in one right at the first turn. This was alarming enough that a second ('final') version of Capablanca Chess was made by altering that game's original setup, yet many players are still happy to play with the original setup. Here's a discussion of Capablance Chess on Capablanca Chess

This 'user submitted' page is a collaboration between the posting user and the Chess Variant Pages. Registered contributors to the Chess Variant Pages have the ability to post their own works, subject to review and editing by the Chess Variant Pages Editorial Staff.

If you are the author of this page, you may edit index information or edit the contents of this page. You may also, if your page uses graphics, upload files.

By Kevin Pacey.
Web page created: 2016-12-04. Web page last updated: 2016-12-04