Singularity Chess | Hacker News Reader

Singularity Chess

abstractstrategygames.blogspot.com

231 pointsiamwil28 comments

Some further information on this chess variant from a 2006 blog post:

http://pmburgess.blogspot.com.au/2006/06/singularity-chess.h...

In fact, the text of the 2010 post appears to be almost entirely lifted (without attribution!) from this 2006 post you cite. The wood-inlay board is very nice but the design of the illustrations appears to be lifted from the earlier post as well.

This sheds some light on this comment about the pawns: '(Move "forward" indicated in white.)' The 2006 picture has white lines and this new one doesn't.

And he, in turn, attributes it all the way back to a calendar he saw on a wall in the 70's or 80's.

Clearly this idea has been kicking around for a while. I used to be very interested in chess variants. I'm a little bummed I didn't hear about it "back in the day".

The board looks neat but this concept seems only half baked, since it begins to confuse directionality once a turn is taken, and the new board greatly reduces the number of available moves.

There is just a single space for the pawns on the end to move to, after all, and it isn't well defined what moves the penultimate pawns could make should they move one square forward (can they attack end-pawns that have moved just one square?)

What might be a better variant would be to assign directionality to all pawns (starting as forward), and allow them to take left or right turns, perhaps diagonally. This greatly increases the number of game possibilities while introducing no new confusing scenarios.

A couple other ideas: don't hack the pawn moves, just change the promotion rules to allow for promotion on either side of the board. Or, don't hack the pawn moves and don't change the promotion rules. If a pawn gets stuck on its own side of the board, the pawn acts as a blockade but runs out of moves. This means the only way to promote is to capture an opponent's piece.

Yeah I was thinking the same thing. For this board to work I think the pawn design should come with a "face" so you can remember what direction it is pointing as it seems possible to turn a pawn around or make them move sideways.

For example, following the rules of the layout I _think_ the end unit pawns should technically be facing away from the board. That is to say, "A2" pawn's forward direction would actually be H7. That's pretty crazy to think about.

I'd like to get this board, I wouldn't be all that interested in a serious game but I think it'd make a fun time waster and argument creator.

One of the side projects I'm pondering is a multiplayer chess variations site, with an engine flexible enough to handle modes like this (and 4 player, etc).

A different variation would be promoted weekly, but you'd be able to play any of them whenever via email or real time.

Would that interest anyone here?

There is a great library of them here, except it's single player only which IMO takes a lot of the fun out of it:

http://www.pathguy.com/chess/ChessVar.htm

To kill some time many years back we wrote a 32-player "chess" game. We had some slightly modified rules, most notably anyone can move themselves whenever they want to and "killing" the king was a matter of checking him for a specified amount of time (say, 10-15 seconds). It was still black vs white.

The game turned out to be very fast paced, to the point where we had to add measures to slow it down. :)

As I recall, it was written with the "board" as a separate entity from the pieces. The board contained the game-rules, for instance whether it was turn-based or not. Now, this is what got me thinking... The platform we used, allow for live recompiling and upgrading of objects -- so we indeed did go in and change the rules even as a game was running (which was kind of fun, just to mess with peoples heads). But I do think perhaps that could be a nice twist to have as a "service" as well -- the ability for people to design their own chess-rules by giving them a framework (like, standard board, standard knight, pawn etc). Naturally saved online, for public scrutiny. :)

It would be be fun to see what people came up with. The target group for this may be somewhat limited. :-)

A friend has a version of what we made hosted on his personal site, but I'm afraid to link to it because it'll probably bring down his server. Throw me a mail if you want the link.

I started a project a while ago which was going to be a variant of chess. Basically where you started out normally, but you could upgrade pieces, and earn game credits to buy new pieces with better abilities, as well as make modifications to the board. I got the normal version of chess working, and got a few abilities working, but then got bored with it, I should have started with a 2D game. Lesson learned.

http://www.metalliccloudgames.com/?page_id=8

Reminds me of a card game that augments chess called Knightmare Chess, out of print now but not hard to find: http://www.sjgames.com/general/outofprint.html

Each player gets a deck of cards which change the rules of chess, for example, you can play a card that makes the board cylindrical, so the edges wrap, and you can move pieces from one side to the next, or another card that allows the king to move two spaces at a time, etc. I always thought it was a neat game.

Looks pretty cool but the pawn rules are a hack. Since "forward" is no longer a well defined direction, I propose allowing pawns to move to any adjacent square, and capture to any diagonal square.

There's only one chess variant that's more interesting than the original: http://en.wikipedia.org/wiki/Kriegspiel_(chess)

I also reject this claim and augment the mention of bughouse with crazyhouse, which is basically a two-player version of bughouse: http://en.wikipedia.org/wiki/Crazyhouse

Crazyhouse has a much higher branching factor than regular chess, and because dead pieces come back onto the board, the game cannot be attacked with endgame tables.

I also personally find suicide chess and atomic chess quite interesting. I can't make the claim that they are more interesting than the original, but they do require delightfully interesting changes in mentality. And endgame analysis in each of these variants is vastly different from regular chess with some beautiful patterns.

I reject this claim and submit Bughouse as a favored variant: http://en.wikipedia.org/wiki/Bughouse_chess

An interesting variant of that variant would be Fog of War Chess, where you can only see an opponents piece if one of yours is attacking it.

Fischer-random is excellent.

I wonder how this changes up the nature of openings. I don't know too much about chess openings, but usually it seems lik you want to control the center. But in this case, it seems like all roads lead to the center, so would it be wise to have pieces sitting there?

Any piece that crosses the Event Horizon (I'd say from B4 to G4) would get torn apart under its own gravity and just disappear into nothingness. Although from the piece's perspective, time would come to a standstill? Now how about that...

An interesting aspect of the pawn movement is that a pawn on a semicircular square can capture, but not be captured by, a pawn on the semicircular square directly rimwards. In normal chess, all threats between equivalent pieces are mutual. (Well, except for the en passent rule.)

I think I agree with the others saying pawn movement is hacky. I'd be inclined to say that a pawn promotes when it reaches either side of the board, but this isn't great because you need to keep track of where a pawn started to know which direction it's going.

It would be cool to have an engine that supports multi-rule chess AI. I've always wanted to get around to writing an AI for Laser Chess [1] for instance, but it seems to me that since it's basically a search strategy, if you generalized the rules you could easily write an AI that can play any set of rules.

[1] http://en.wikipedia.org/wiki/Laser_Chess

I made this with the goal of eventually incorporating other boards and variations. Currently there are 3 options (click on GAME): Standard, Fischer-random (960), and a 7x7 board with no queens. http://chessfoo.com

quite a challenge to implement the singularity mode ;)

Here's a CSS example of a board: http://socketgaming.com/singularity-chess-css/

Is it still an interesting game to play?

Don't get me wrong, it sounds interesting, but with such games you usually only know for sure once you've tried it.

I love the idea! I want to try playing :)

But "Singularity Chess" is an awkward name. The center of the board is not a singularity in any reasonable sense of the word. The space of allowable movement trajectories appears to be nonsingular.

Maybe a more correct name, albeit drier, would be quadratic chess (because the transformation looks like a quadratic form).

Why does black have only 7 pawns?

"... a rook's pawn still on its original square could capture the opposing rook's pawn on its original square— not sure whether that should be permitted or disallowed."

It was at this point that I stopped taking this chess variant seriously.

Reply on news.ycombinator.com