[jfengel]

The idea that the Romans had a compass does not seem absurd.

The connection to the dodecahedron seems arbitrary. The rotation of the earth means that any alignments last only for an instrument. You'd need a specific kind of mount to make any other orientation work, and the whole point of the points is that it lies on a flat surface. If it's meant to be hung, it would have an asymmetry that suggested where up and down would be.

He's just guessing that the angles add up to something. And it wouldn't surprise me if they did -- there are plenty of stars and a dozen faces. Something probably lines up somewhere.