Project: Phase 1

This page describes both an overview of the entire project and a detailed description of the first phase.

I. The chess program

Your project in the course is to produce a program for playing chess and chess variants with the following features:

• Users can design variants on chess, specifying new kinds of pieces and how they move, the board layout, the initial positions of pieces, how the game is won, and special rules.
• Users can play either as two humans at one computer, two humans at different computers through a network connection, or one human versus the computer.
• The AI for the "computer" player will be defined by plug-ins that users can write. (You will not be required to write any algorithms for automated chess playing apart from dummy plug-ins for testing.)

The range of possible "chess variants" is not well-defined. A large part of this project, especially at the beginning, is to find a reasonable range of chess variants for your program to handle. To do that you will be working with real, live potential users of your system: The Wheaton College Chess Club.

The project will be divided into three phases.

• Phase 1: Determine a reasonable range of variants for your system to model. Design a scheme for modelling and storing descriptions of these game variants. Make a bare-bones application as a driver for your model, showing the board configuration for a given game variant. You will work in 3 groups, sizes 3, 2, and 2. Each group will work on the same problem, and the best result will be used in subsequent phases. Assigned Feb 2, due March 4.
• Phase 2: Make a program for two human players at one computer to play a game of any variant. The user interface should be reasonbly good---there may be some expectation that there will be some changes to it in the next phase, but it should be readily extensible for those improvements. You will work in 2 groups, sizes 3 and 4, with teammates different from phase 1 as much as possible. Each group will work on the same problem, and the best result will be used in the final phase. Assigned March 14, due Apr 1.
• Phase 3: Add the remaining features: A user-friendly way to edit and define new game variants, distributed play, and the AI plug-in feature. You will work as a single group, though you will likely divide up into about 3 subteams, each working on a different aspect of the problem, all together producing one result.

Resources. There has been prior work on the problem of classifying and specifying chess-like games. You may look at various on-line resources (Wikipedia would be a reasonable place to start), and the following books:

• David Pritchard, Encyclopedia of Chess Variants, 2001
• D Pritchard and J Beasley, The Classified Encycolpedia of Chess Variants, 2007
• David Pritchard, Popular Chess Variants, 2003

II. Phase 1 details

By interviewing members of the Wheaton College Chess Club, determine a reasonable range of chess variants for your system to model. Design a scheme for modelling and storing these variants---both conceptually and using a set of Java classes. You will also need to be able to store this information in files.

Users will need a way to create games using your scheme, but this doesn't have to be easy at this point---it could be that they simply need to edit the files directly. Make a bare-bones user interface that acts as a driver for your model: the user should be able to select a game and have the initial set-up for that game shown, with some way to indicate how the pieces can move. This doesn't have to be very user-friendly at this point, only something that will prove the concept of the model.

In addition to actual software, in this phase you will also produce documentation describing

• A record of your interviewing of members of the chess club and how you determined the range of variations for your system to model.
• A formal description of how you model the information about the variant gmes.
• How the files are formatted and how to edit them.

I've made March 4 (the Friday before spring break) the official due date, but I strongly urge you to aim for the previous friday (Feb 25) as the working due date. The only reason I didn't make that the official one is that I figured you wouldn't use the week before spring break on phase 2 even if you had the option, so you may as well have that week as a buffer time. If, however, everyone does finish by Feb 25, I can reassign new teams and get the phase 2 specification earlier, so you can begin planning for that phase earlier.