Statistically Speaking

I’m in a class this quarter called “Design and Manufacturing,” where over the course of ten weeks I’ll design and, um, manufacture an object more or less of my choosing.  I’m in the thick of the brainstorming process currently, and have spent the last twenty minutes drawing clocks.  That’s a discussion for another post, and eventually I’ll decide on a project and post about that (presumably).  But for now, the last sketch is of a cube with a different clock on each face — one for LA, one for London, one for Tokyo, etc.  A desk toy for the jet-setting executive.

This, in turn, made me wonder: if I roll this dice-clock multiple times, will the average time tend towards Greenwich Mean Time?

I call it the Central Minute Theorem.



  1. I suppose the question is what’s the probability distribution function of time. Picking a random time zone would be more or less a discrete random variable with 24 possible outcomes, which means that the average time anywhere on the planet is noon. So if it happens to be 12 in London, this is totally true. Where’s my Fields Medal?

    (The fallacy here, of course, is that the GMT is a value, not a distribution. Like anywhere else, the central limit theorem is going to hold true. Hooray Gaussians!)

Comments are closed.