Learning Movement

This isn’t so much of a game idea as a possible use in games. Or just a theoretical experiment for mapping movement.

By creating a model of the human body (not exact, but approximately), or even of some sort of entity, and then using RL (I love RL) or some other method of testing to test out how it moves, semi-realistic movements could be created. For example, this algorithm/program could be applied to a computer-generated martial art, dedicated to causing maximal pain with minimal force. It’d be interesting to see what it comes up with and see which current martial arts use these CG techniques.

This model could be used in all sorts of things and the numbers that are returned from doing certain movements could be used in damage calculations in games. However, modelling the human body would require massive computation. But then again, technology is adavncing pretty fast. Who knows, it might be a viable project.