| Author |
Message |
Ronald W Brierley (Ronaldwb)
New member
Username: Ronaldwb
Post Number: 61
Registered: 2-2003
| | Posted on Sunday, April 01, 2007 - 4:20 am: | |
My TorusChess10x10 has been accepted as a new Zillions game.
TorusChess10x10 is chess on a 10 by 10 board which has its opposite sides connected.
All cells of the board are identically connected to their surrounding cells. There are thus no sides to the board and it has no centre. The concepts of controlling the centre of the board or promoting pawns or consolidating by castling or checkmating using a side of the board do not apply to the torus board. Pawns move in orthogonal steps unless capturing when they must take a diagonal step.
The strengths of pieces on a torus board differ from those on a standard board. To assist the Zillions AI to play an acceptable game the piece strengths have been tweaked.
Each player starts with a standard set of pieces with as many pawns as is needed to make the starting position interesting. One variant uses a restricted number of pieces.
A position has been saved to illustrate the tweaked values of the pieces. A variant exists which is an almost empty board. Although it is impossible to force checkmate with a rook and king against a king, it is possible with two rooks and a king against a king. What is the weakest piece set which allows checkmate to be forced against a sole king and how is it accomplished? |
M Winther (Kalroten)
New member
Username: Kalroten
Post Number: 14
Registered: 1-2007
| | Posted on Thursday, April 05, 2007 - 9:04 am: | |
You could always invent a new form of rook that is the same as a regular rook except when it controls the squares on the last rank or file. In these cases only, it could also control the rank or file on the opposite side. In this way it could give mate to the enemy king together with a king. /Mats |
Ronald W Brierley (Ronaldwb)
New member
Username: Ronaldwb
Post Number: 62
Registered: 2-2003
| | Posted on Saturday, April 14, 2007 - 3:43 am: | |
Hello Mats,
I would like to suggest that you are solving the wrong problem.
As I stated, a torus board is so connected that there are no special squares. It is like being on a sphere where there also are no special places, except that a torus board is worse since there is also a hole in its space.
All this implies that there is no centre of the board nor are there any sides to the board. If you use the boards and pieces I have provided all will soon become clear if a little surprising.
The boards I provide are displayed statically. This I found preferable to attempting to provide any movable display. Uwe Schaechterle, Corpsman, has done some remarkable programming for a smaller board with fewer pieces. On his screen it is possible to move the board up the screen so that pieces at the top of the screen immediately reappear at the bottom. Such sophistication is more than I could currently achieve in a Zillions program.
I also like to keep to standard chess pieces. Only the pawns have any alteration to their standard modes of movement. After accepting the loss of forward and backward directions on the board it is natural to allow pawns to make any orthogonal step when not capturing or any diagonal step when capturing.
The piece you suggest introducing would best perhaps be called a SuperDuperRook since it would have to attack not only, as you suggest along a file, which you name the last, but also along a neighbouring file and so using the implied symmetry of a torus board it would also need to attack along the other neighbouring file. That is it would have to attack along three adjacent files. This goes way beyond being a reasonable piece. If such a piece moved into position to check an opposing king it would in fact produce checkmate without the intervention of the friendly king.
Try the boards and pieces as they are.
Ronald W |
|