Built on Facts

Not much meat here at the Built on Facts table today. Our research group is embarking on a new project with the Office of Naval Research, and today was the official all-day meeting with the officer in charge of coordinating the various projects in his part of the ONR funding structure. We've spent a lot of time getting everything ready, and based on what I saw today I think the research is going to open up a lot of new, interesting, and important territory. Plenty of good blog post material, too.

I'll fill in some of the details later on, though of course a lot of it will have to wait until we publish results later on down the line. The fall semester begins next Monday, and will bring with it a more regular schedule for me which should help get this blog humming along a little better than the somewhat irregular last few weeks.

To hold us over, I'd like to point out (via Swans on Tea) a very neat article about coin flipping. While it's the canonical example of a a random process of probability 0.5 for each of two events, in reality the situation is not so clear. The actual probability of reaching a given side tends to be more like 51-49 for a "random" toss, or as much as 80-20 for spinning on edge, and essentially 100% if you have a machine that can accurately flip a coin with the same force and angle each time.

The article doesn't mention this, but it's also interesting that it's very difficult for humans to make up fake results of coin tosses. The Hs and Ts are almost always much more uniform than they actually are under true randomness, with runs of 5-6 heads occurring with relative frequency in real randomness but almost never in fake data by a person trying to imitate randomness.

All right, that's all for now. See y'all tomorrow!