The idea of using "games" to learn mathematics, other disciplines, is nothing new. Puzzles, brain teasers... Martin Gardner territory.
Now that computer animation has invaded the game playing space, in arcades, on smartphones, on computer screens, a lot of skills building goes on with these projects. Sometimes the goals are simply mnemonic i.e. to help with recall of the relevant facts.
Any subject is amenable to some degree of rote learning and building efficacious "memory palaces" is time-honored component of rhetoric since before the Renaissance. Today it's called "spatial data management" or some such. Cite: 'The Art of Memory' by Francis Yates.
However development of these games requires yet other problem solving skills, just as using, designing and building a calculator each takes different skill sets.
A critical question is where to bring real human feedback into the process, versus "robot scoring" via machine administered tests (there's not much distinction between a scored game and a test -- practice practice then play it for "the record books".
Development is a collaborative process so human feedback is integral in both making the tools, and using them. There's typically a "solo mode" and a "multi-user" mode. Many games involve "role playing".
Solving puzzles together: we need games that are good at promoting such behavior. Games like Uru? So many yet uninvented. A vast territory.
Anyway, I have a head full of ideas, obviously. Business trips can be inspiring that way.
I represent the Python tribe in this context whereas others advocate for and/or apply other tools, oft times in a complementary fashion.
During my Chicago workshop, two of the geeks heard me pushing Visual Python as a favorite teaching tool and they assessed (a) the web site was too outmoded-looking to easily attract new users and (b) the installation mechanisms were too broken. They assiduously set about building a Linux installer and completed the project before the workshop was over. Wow. Now we just need to get the improvements accepted.
Then, after Chicago (Hyatt downtown, conference just finishing up), I drove south to Wolfram country. Great place. I'd never been to Champaign-Urbana before. Today I'm in Indiana.
Synergetics was a part of the Djangocon presentation, though not a big part. It's wasn't hard to weave in, given Chicago is all about architecture. I'd taken an "architecture boat" the night before and had my slides ready, something to look at as I got things set up. Mentioning Bucky Fuller when mentioning architects is a smooth segue, with Alexander Graham Bell another foreground figure.
However it's really just generic spatial geometry I'm talking about, in showing off Visual Python (vpython.org), VRML (x3d.org) and POV-Ray ( povray.org).
Reinforce the core math in the language of your choice, coding a model.
In model-view-controller architecture, we allow for various views based on a common model.
The same math outputs on various visualization formats. I don't do as much with sound as other teachers. Not my forte.
Scott, my former boss (he's posted here, tried to join the threads about calculus -- his forte), strongly advocates for Mathematica in particular. That's his heritage as a Uhl guy.
I'm quite respectful of Mathematica and look forward to more results coming in from the various classrooms using it, but given my Python affiliations, I'm more likely to talk about SAGE or iPython Notebooks.
Given we all share a commitment to Diversity, it's not that big a deal when two camps, or schools of thought, develop different branches of the technology. There's no assumption of a "winner take all" theme.