There really is not much at all to the physics of this minigame, but still I am reading up a little to refresh on the Verlet method.
Coding game physics in general can be an all-consumptive process, so these days I'm preferring to keep my eyes on task, and to follow the game design plan. I remember assembling before me the Chris Hecker articles on rigid-body mechanics - spending so many hours figuring out what it all meant, drawing out huge matrices, putting together a general rigid-body physics simulation, setting up input files, editing LinPack, and cleaning up compile issues. At the end of the process came the realization that, try as I might, there is really no good use I could come up with for ragdoll physics in a game. This was at the same time that a certain physics engine claimed to have secret, novel, and more-stable algorithms for doing all these things in 3D. Around that time, I was deciding to eschew floats completely from my code anyway. Ever since it's been only simple wham-bam physics and completely active character control for me. If there's an inexpensive or freely available integer-based dynamics library that works well with my primitives, I will probably use it and like it, but for now I'll stay away from all general-purpose high-precision engines until given a reason to do differently. Server-side physics might enter the creative domain eventually, but hopefully not.
I should now have a solid block of two protected hours per night to work on development projects, without stealing from the sandman. I'm happy about that.