Chess engines for circular chess boards

Discussion of chess software programming and technical issues.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
RichV
Posts: 16
Joined: Sun Apr 14, 2019 3:33 pm
Full name: Rich VanDeventer

Re: Chess engines for circular chess boards

Post by RichV » Mon May 06, 2019 3:19 am

Thanks Gunther! I am new to the list and don't know all the posting tricks yet. Your help is greatly appreciated.

RichV
Posts: 16
Joined: Sun Apr 14, 2019 3:33 pm
Full name: Rich VanDeventer

Re: Chess engines for circular chess boards

Post by RichV » Tue May 21, 2019 2:49 am

Evert,

Based on your earlier response which included a discussion on the relative power of the Archer and Catapult based a few different types of calculations, I rooted around in Zillions of Games and discovered that it too has a power rating system for game pieces. It uses numbers that are in the thousands of points and calculates the power for each piece based on its current position on the board. The values for each piece type do not vary substantially from position to position. It seems a bit different from the other systems you referenced.

For traditional chess, its approximate values are (I divided the value for each piece by the Pawn's value to get values that would approximate those of traditional chess):
P - 1,850 (1)
R - 8,598 (4.6)
N - 4,991 (2.7)
B - 5,921 (3.2)
Q - 14,011 (7.6) (seems to be a point or so undervalued)

For Castle Siege Chess using the Archers and Catapults:
P - 3,864 (1) (compared to a traditional chess P, it is 2.1 times more powerful, probably due to 12 Pawns and promoting one rank earlier)
R - 44,251 (11.5) (2.5 times more powerful than a traditional chess R, probably due to the larger board, 3 directions of attack, and the triangles.)
N - 18,577 (4.8) (1.8 times more powerful than a traditional chess N, probably due to the larger board and slightly due to the triangles.)
B - 22,358 (5.8) (1.8 times more powerful than a traditional chess B probably due to the larger board and the triangles.)
Q - 64,697 (16.7) (2.2 times more powerful than a traditional chess Q, probably due to the larger board, 3 directions of attack, and the triangles.)
A - 16,941 (4.4) (slightly less powerful than the N.)
C- 23,188 (6.0) (slightly stronger than a B; I think the value is about 3/4 of a point or so too high and the value of the B is about a point or so too low.)

Just thought you might be interested to see how ZoG calculates the power of the chess pieces on the CirSquare 96 board.

The power calculations you described for the A and C made me think about having an alternate A and C that are a little more powerful. I could change the name of my current Archers to Longbowman and create a new optional piece called the Crossbowman that moves the same way as the current Archer but captures only on the 3rd space, so he can attack from a position further away from the target piece. I could change the name of the current Catapult to Mangonel (a small catapult that was not very accurate) and create a new optional piece called the Tribuchet (a larger, more accurate catapult) that moves the same as the current Catapult but captures only on the 3rd space. I have not tested the new pieces in action yet, but I have visualized their moves on the CirSquare 96 and think they might work rather well, but I am a little concerned that they may be a little too powerful. What do you think?

For new pieces, I like to use piece names that relate to the battlefield warriors and weapons of European medieval armies instead of using animal names like many other chess variants do.

RichV
Posts: 16
Joined: Sun Apr 14, 2019 3:33 pm
Full name: Rich VanDeventer

Re: Chess engines for circular chess boards

Post by RichV » Tue May 21, 2019 3:26 am

Evert,

I just edited the ZoG script for CSC to make the A and C capture on the 3rd space instead of the 2nd space. It made virtually no difference in their power rating. But I can't help but feel like they would be more dangerous by being able to strike from one square further away from the target. So then I changed the script so that the A and C move 1, 2, 3, or 4 but capture only on the 3rd space. This time the power of the Archer increased from about 16,941 to about 19,410, an acceptable 15% increase in power; and, the Catapult increased from about 23,188 to about 27,255, which is also an acceptable 18% increase in power.

Now, I feel like the new optional pieces and renaming of the existing A and C is worthwhile to play test. What do you think?

Post Reply