To have interactive games with physics you need some kind of "physics solver" or "physics engine".
For games these needs to work in real time, but for movies they don't, as such there is a lot of techniques out there that isn't suitable for games. However some of these might be suitable in the future as new research improves on their efficancy, creates versions that can run on the GPU, and hardware in general gets improved.
By games "ready" I mean techniques where the paper shows that it can run in real time and potentially be used for games.
Most game engines uses some form of impulse-based dynamics I belive?
Extended Position Based Dynamics (XPBD)
Graph coloring for multithreading physics
By movie "ready" I mean techniques that aren't real time, usually somewehere between seconds and minutes per frame, and therefor suitable for pre-rendering but not for games.