This month, I thought I would follow in the footsteps of my esteemed co-contributor Bruno Faidutti and talk about the invention of two of my own games.
I've been fond of Chess for pretty much my entire life, though I've always been too much of a dilettante to study it seriously. I've long enjoyed a number of Chess variants, starting with the completely lunatic Bughouse Chess in high school and continuing through the wide world of CVs ever since.
My study of the variety of chess variants led me, of course, to dabble in creating my own. When I first started studying chess variants seriously at around the time of the initial publication of Richard Garfield's Magic: The Gathering. I found myself toying with the idea of a "collectable chess game"—a Chess variant in which players would bring their own arrays of pieces to the game. But every time I tried to picture what a game of "Chess: The Gathering" would look like, I kept running into the image that everyone would just want to buy a "deck" of queens.
I can still remember the exact sentence that finally bubbled out of my lower brain: "And why is that a bad thing?" From that question almost immediately came Hecatomb.
Start with a standard chessboard. Place the kings in their normal starting positions. Fill the rest of the board with queens, 31 on each side. That's it—all of the other rules of Chess are in effect.
Hecatomb plays quickly, but with a fair amount of tactical depth. After some preliminary capturing, almost every move is a check, which pleases my inner patzer and keeps the tension level high. Complex crystals of queens develop as the board thins—emergent complexity at its finest. If you want to try it out, there's a free Hecatomb module available for Zillions of Games. Hecatomb is actually part of a family of games, which I collectively call "Zanzibar Chess" (after John Brunner's seminal novel about overpopulation, Stand on Zanzibar). One can just as easily play Hecatomb with rooks instead of queens, or with dragon chariots (a Shogi piece that moves like a bishop but can also move one square orthogonally), or pawns.
The Devolution of Chess
Some time later, I was reading David Pritchard's wonderful The Encyclopedia of Chess Variants when I realized that one of the things I liked about Hecatomb is its simplicity. Not its shallowness; as I said, there's a fair amount of depth to it. I mean, rather, how concise the rules are—25 words describe it completely, if you already know Chess.
Western Chess (a.k.a. orthodox Chess, the rules of which are codified by FIDE, the International Chess Federation) is an elegant game. But it could be more elegant than it is. It is full of what I call "fiddly little rules," rules that cover unique situations or allow pieces to move or act in unusual ways in certain circumstances.
So I set out to strip chess of its "fiddly" bits.
My guiding conceit was, whenever possible, to draw rules from existing forms of Chess—orthodox chess, the East Asian chesses (Xiang Qi [Chinese chess], Shogi [Japanese chess]), and so forth—and pretend that I was "recreating" the primeval game of chess from which all others emerged: "Ur chess."
First to go was stalemates. In orthodox Chess, if a player has no legal move and his king is not in check, the game is a draw. This is, I feel, somewhat awkward and is not shared by the East Asian chesses.
New rule: A player who cannot make a legal move loses. (No stalemate.)
Since this has the effect of reducing draws, I thought I should continue in that vein:
New rule: A player who is reduced to only a king loses.
(This rule, "robado," was present in some historical versions of Chess. I suspect it was present in the very first chesslike game.)
To avert draws through repetition, I borrowed a rule from many modern abstract board games. The "ko" rule in Go prevents the immediate repetition of moves. I opted for the stronger form.
New rule: No player may repeat a previous board position.
These three rules together constituted what I called "Renormalized Chess." But I never played it; it struck me as a half-measure—not enough of a different game to constitute a "real" chess variant. I'd be delighted to see those rules adopted by FIDE for chess, since I think that a reduction in draws would improve the game. But I'm not exactly sitting by the phone waiting for them to call.
Anyway, I started looking at the other special cases in the FIDE rules. I came up with two major fiddly bits, castling and pawn movement. So, for Ur chess:
New rule: No castling.
New rule: No pawn doublestep (that is, a pawn always moves one step forward, never two).
New rule: No en passant.
Now, pawn doublestep and castling both serve useful purposes in orthodox chess. Pawn doublestep cuts in half the number of moves before the opposing pawns come into contact, and allows the pieces to develop more rapidly. Castling brings a rook into the center of the board, where it can hook up with the other rook. So, drawing upon the Xiang Qi array and upon Christian Freeling's Grand Chess I did the following:
New rule: Pawns start on the third and sixth rows, not the second and seventh.
This gives an empty row between the pawns and the main pieces that the rooks can move into, supporting each other; the other pieces can develop more easily, too. This also makes the board look a lot more like the East Asian chesses, so I thought I'd borrow another rule from those games:
New rule: Pawns can promote when they move in the opponent's territory.
White pawns can promote when they move onto the sixth or seventh row, and must promote on reaching the eighth row; black pawns symmetrically. This is pretty much the promotion rule from Shogi.
Another "inelegant" rule is the unique diagonal capture move of the pawn. No other piece in any major form of chess moves differently when capturing than when moving. So:
New rule: Pawns capture by moving one space forward.
Again, this is directly from the East Asian chesses.
Finally, after a discussion with Michael Keller on a different topic, I realized that I'd always found it problematic that not every game of orthodox Chess could be played with a single chess set—i.e., a player could promote a pawn to a second queen or third rook. Shogi solves this problem by having each piece's promoted form on the flip side of the piece, but I went with a different approach:
New rule: Pawns promote only to eliminated pieces.
This helps explain why a pawn wouldn't necessarily want to promote as soon as possible.
So there it is.
I first played this with Michael in 1997 and was surprised at how much fun it was. The pawn rows begin in almost immediate contact and development of the major and minor pieces proceeds very quickly. Fast promotion of the pawns means that the game remains full of major and minor pieces for a long time, which means that endgames are much more furious than in orthodox Chess.
As with Hecatomb, a great deal of the emphasis is on offense. In Ur chess, the pawn lines get ripped to shreds very quickly; the long pawn chains of orthodox Chess are almost impossible to build in Ur, and thus the board is much more open. In Xiang Qi and Shogi, the king has bodyguard pieces to help fend off attackers, but nothing comparable exists in Ur chess. Castling in orthodox Chess allows the player a chance to build a shelter for the king, but in Ur chess not only is there no castling, the pawns are moved forward, leaving a huge space behind them, which is hard to defend.
Is Ur chess a better game than Chess? Almost certainly not. But it has a charm of its own, and I think it's a faster and cleaner alternative, a variant which keeps the strongest points of Chess while sanding down just a few of the sharp edges.
Thanks go to the Chess Diagram Construction Utility for the illustrations used in this essay.
And thanks again to Michael Keller for all his help and encouragement with both Hecatomb and Ur Chess. Word up.
- Kevin Maroney