A system to classify chess variants

I'm not sure a diagram would help. If the size of the board matters, and there is some ideal range of sizes for chess games, I think that range would be higher for hexagonal boards than square, based on my (limited) experience with both. I think it's because the hexagonal board has more corners, so the corners feel less isolated from each other, and closer to the center. Maybe it's just me

Actually the board is 5 faces of a 4x4x3 cube with 'rounded' edges and corners to provide an extra file of squares (or triangles at the cut-off corners). Flattened out to a planar configuration.

I used AutoCAD to create the CirSquare board in 1997. I wanted to fill in the hole in the center of the Byzantine chess circular board which is composed of 4 circular bands of 16 spaces each. My goal was to create a third direction of attack, which would be a frontal attack, to add to the two flanking attacks.

I felt that 16 spaces made the curve of the diagonal moves visually too tight near the center, so I added 4 more spaces to make the bands have 20 spaces. After many tries, I found that merging the perimeter of a 6x6 square board with the inner band of the circular board worked perfectly.

The initial layout was quite angular, thus not very attractive. I could not find a way to make the 4 triangles disappear without messing up the alternating color scheme of the spaces. An advantage of 20-space bands turned out to be the creation of 4 additional spaces for Pawns and 4 additional spaces for minor and/or major pieces. It also made the board 12 spaces wide across the battlefield while keeping the classic 4 ranks between the starting areas (castles). It also made the board have 96 spaces, which is the same as a 12x8 board, making it exactly 50% larger than a classic board and have 50% more pieces while maintaining the 1:1 ratio of pieces to open spaces at the beginning of a game. My goal of 3 directions of attack had been achieved in a way that created a circular board that still had a strong feel of play classic chess with two more directions of attack on a very dynamic board.

Because I did not initially know what to do with the 4 triangular spaces, I made them not playable--they were obstacles on the board. The game played terribly, so I had to figure out how to make the triangles playable. Requiring pieces to stop on them destroyed a player's tempo and requiring a piece to only go to the left or right out of a triangle was too pollical :-). So the only option left was to make them act like a fork in the road. That worked perfectly and made the triangles the most important and powerful part of the board. After recently playing Castle Siege Chess with me, Zied named the triangles the Bermuda Triangles and we renamed the game to Thrones Chess. In 1997 I had named the game Round Table Chess, but a furniture maker was already using that name for a chess table he was manufacturing, so I changed it to Castle Siege Chess.

Over the years as I continued to work on the game to improve it, I eventually found it important to increase the size of the triangles so that chess pieces could more adequately fit on them. I found that  changing from angular shaped spaces in the center area to more curved lines made the triangles larger, made all the center area spaces more equal in size, yielded more flowing lines, and also give an illusion of there being a hill in middle of the board. It was about that time that I added the castle walls.

It is really mind-bending for me to try to visualize the board as coming from 5 faces of a 4x4x3 cube with 'rounded' edges and corners. I could never have thought of it in that way. Very interesting. Thanks HGM.

This might help:

BTW, your name sounds Dutch, except that I would then expect it to be Van Deventer, rather than Van de Venter.

Kramnik suggesting a new variant here

https://www.chess.com/article/view/no-castling-chess-kramnik-alphazero

It is really 'VanDeventer' written as one word instead of two. Zied was trying to write my name from memory and maybe gave it a French twist. :-)

Ah, I see. For Dutch names 'vande...' would in 99% of the cases indeed be 'van de ...' (meaning 'from the ...'), and 'venter' is a Dutch word (meaning 'traveling salesman'). But in this case 'Deventer' happens to be a town in the east of the Netherlands.

Thanks HGM. My wife and I vacationed to southern England and Amsterdam several years ago. While we were in Amsterdam, we took a train to Deventer and spent a day there. Quaint little town. I asked a store owner if there were still any VanDeventers living there. She said no, that they had all moved away during WWII.

Have you or Evert run across any strategy game development programs that can handle circular boards since I talked with you last on TalkChess? My need to get Thrones Chess (p.k.a. Castle Siege Chess) playable online is growing urgent since I want to do a Kickstarter for it sometime next year.

Thanks!

Yeah, Deventer is nice. Usually people named 'Van <location>' are not actually living in that location. The surnames were forced upon us in most places by Napoleon, and when people picked them in 1811 the picked family names were supposed to distinguish you from neighbors that carried the same given names. Naming yourself after the town where you lived would not really achieve that. So mostly people that came from another town than where they lived called themself after their town of origin. So Deventer would probably be the worst place in the Netherlands to look for Van Deventers!

As to your question: 'Jocly' is a web applet for developing 2-player board games, which ise very versatile. The move generator of its AI just keeps a list of reachable squares for every square of origin, so you can do arbitrary board topologies. Indeed, some of the games that are already implemented use a the surface of a cube as board. The 2d graphics is impressive, but the AI is quite weak. And developing a new game for it is not very easy.

Thanks HG. As always, I appreciate you taking the time to pass along some of your wealth of knowledge. I would still like to send you a physical 'pre-production' copy of my game to get your opinion of it.

When I programmed Castle Siege Chess (now Thrones Chess II) in Zillions of Games (which is expensive and seems to have been abandoned by its creators) I placed the 2d layout of the CirSquare board in quadrant I of the XY plane to determine the coordinates of each playable space on the board. I used zones and links for designating the legal paths for piece moves. Would that be what I would do in Jocly? Do you know how I could obtain the Jocly file for Byzantine (Circular) Chess so that I could use it as an example of a circular chess game to leverage in the development of my game? I am pretty good at reverse engineering one game to create another. That is what I did with Zillions--I reverse engineered a classic chess ZRF file to create my CSC variants.

The only thing that worries me about using Jocly is knowing that the Jocly server was destroyed by a hacker and Jocly has not come back yet. Is Java and JavaScript so vulnerable to hackers that I should not put any Java or JavaScript files accessible to the public on my server?

*doesnt read any of the posts*

We should start a petition for Chess.com to include Monster King as one of its variants 

https://en.wikipedia.org/wiki/Monster_chess

To start with the last point: JavaScript should not be any more dangerous to have on your website than normal HTML pages. Seen from the server it is just a data file like any other; it will only run as a program in the browser client. So malicious JavaScript might be able to do damage on the client side, but could never be used to attack the server. If there was a hacker attack on jocly.com it must have been through vulnerabilities of the server code used to have the varius client-side Joclys communicate, (which was not part of the open-source Jocly project), or through completely unrelated faults in the hosting software.

The Jocly source code can be found at GitHub. The game definitions of the various chess variants can be found in the sub-folder /src/games/chessbase/, and is divided between a 'model' and a 'view' .js file. The former implements the rules, the latter is for display-related issues (type and size of the pieces, and such).

Chess variants are similar enough to require a lot of common code, and general support for them is provided in the files 'base-model.js' and 'base-view.js'. Somewhat less general, but still common traits of chess variants are handled in files for certain piece sets (staunton-set-view.js, fairy-set-view.js, xiangqi-set-view.js) board topologies (grid-geo-model.js, hex-geo-model.js, cubic-geo-model.js, cylinder-geo-model.js) or board looks (grid-board-view.js, shatranj-board-view.js, circular-board-view.js). There is a file 'index.js' that for each defined variant contains instructions on which of those to include.

There are files byzantine-model.js and byzantine-view.js which should contain the implementation of Byzantine Chess, and the definition in index.js specifies what type of common code it uses for rules (base-model.js, cylinder-geo-model.js) and looks (base-view.js, circular-board-view.js and nishapur-set-view.js).

Your board is rather special, though, so you would probably have to write your own geo-model for it, taking care of the forking of slider paths on the triangles. I don't think any of the supported variants has a similar problem. (The variant listed in the 'more Jocly games' overview as '360 Chess Authoring', for which the source files carry names cubic-*, could also be inspirational, although I think that the cube has barriers on some edges that make it possible to map it on an ordinary flat board without irregularities, but with just some peculary periodic boundary conditions.)

HG,

I could not get to the GitHub sub-folder /src/games/chessbase/, it gave me a 404 error.  I entered it on the browser address line: https://github.com/src/games/chessbase/.  I also tried it in the search field. What is the correct way for me to access the sub-folder?

Thanks, Rich V

It is a sub-folder of the Jocly project at Github (to which I also gave the link). The full URL would be:

https://github.com/mi-g/jocly/tree/master/src/games/chessbase

If you go to the Jocly project page first, you can navigate through the entire source-code tree on-line.

Of course if you eventually want to add your own variant you would have to download the source (e.g. as a tar ball, although it would be preferable to install 'git' on your own machine and then 'clone' the Github repository localy. You will have to add the necessary JavaScript files (at least a *-view.js and a *-model.js), add them to the project description in the index.js file, and then 'compile' the lot by running 'gulp' to get a Jocly 'library'. Which you can then host on a website.

If you want to hack on Jocly source code it is better to git clone it to your local computer.

First install git from https://git-scm.com/downloads then

git clone https://github.com/mi-g/jocly.git

If you don't want to use git you can download Jocly source clicking on green "Clone or Download" button on https://github.com/mi-g/jocly

Another possibility can be to contact https://greenchess.net/ author Uray M. János and ask him to implement your variant on his site.

You still have to post after that

Thanks Zied, HGM, and btami!! As a first time GitHub'er, I need all the help I can 'git'. :-)  If I need to do a lot of downloading, I hope my laptop has enough hard drive space and horsepower. First, I will try HGM's link and advice of going to the Jocly project page.  Thanks!!

Recently discovered your chess variant and I must say that I love it. My issue with it, however, is the following:

The existence of the Archbishop, Chancellor and Dragon/Empress as pieces. While these three seem logical follow-ups to the already existing chess pieces (R+N), (Q+N) and (B+N), they are too powerful, and I predict that, should Musketeer chess become more popular among pros/masters/grandmasters etc, these three will see far more usage than weaker pieces such as the Leopard or the Spider. After all, why go for a weaker piece, which might also move in a peculiar way than for the simple force of nature that is the fairy piece known as the Dragon?

I think that in chess variants with extra pieces, the way to go is to have pieces with a limited range compared to the bishop and rook, but with a wider "area-of-effect", if you will forgive my borrowing from RPG terminology. A good example of this is your Spider, or a piece such as the Paladin from other variants (K+N).

I may be wrong due to my inexperience with chess, but this is my two cents regarding musketeer chess, and variants in general.

It would be interesting if the Chancellor's orthogonal range was limited to two squares, sort of in the same way as the Leopard is the 'nerfed' version of the Archbishop.