The Practice

By the end of the project, you are expected to have in hand one or more physical dice that can be rolled and experimentally evaluated to determine their discrepancies. Each face must be flat, and the dice must be convex. To achieve varied face probabilities, your dice will probably be irregular, with nonuniform face areas/shapes and/or weight distributions.

The way you build the dice is up to you. You could try 3D-printing, sculpting (e.g., from clay or wood), molding, assembling from pieces made of plastic or metal, etc. The tricky part will probably be iterating your designs as you learn about what it takes to achieve various probabilities, and what kind of physical change it takes to achieve a desired change in probabilities. There is some experimental work (e.g., http://dicephysics.info/) that discusses what happens when you shave dice by small amounts or load faces with a small extra weight.

You may find it helpful to use a physical simulator (such as bullet or PhysX) to iterate through some initial designs, but there are no guarantees that the simulations will give accurate probabilities for physical dice.

If you choose to do 3D printing, the Maker Space can be used to build your dice. The Maker Space also has equipment for laser cutting if you want to assemble your dice from specially designed face templates. (Note: some training is required to use the Maker Space -- see the TA.)

As you refine your designs you'll probably discover that not every theoretically good set of face probabilities is easy to achieve in practice. How will you refine your desired set of face probabilities in light of your physical design experience?