Ha ha, glad I got at least one person to point out that computers can't be random (but of course how do you work that out - if it was truly random it couldn't be predicted which I don't think this can. Don't for one second think that random would mean 500 heads, 500 tails on 1000 tosses and no, it's not supposed to be a coin simulator!

Well...the rating of zero stars was a mistake, I'm not sure why it was rated zero, it was supposed to be no rating. Tearing it apart? I suppose I was a bit harsh. It appears to be a decent approximation of a coin - but only an approximation, which is very important to recognize. Mixing probability and randomness? Maybe a little, after all, this isn't a statistics course, more like a general discussion. I suppose it would depend on the direction one wants to approach the problem, ei…ther the theory of a fair coin or the analysis of a real coin. As far as developing a random algorithm, that was the point, it can't be done. More

First of all, no mind the negative comments regarding it not actually being random. So what?! If you think you can make it more random, develop one yourself! Eldon, thank you for sharing your widget. I had fun spinning for a few minutes. I'm sure you had fun developing it too!

I have no use for the widget, but it works well and seems random enough for my tastes.

@Physics Major, Math Minor: there are some valid points, but note that you're mixing the idea of an event being random and that event's probability. But leaving that issue aside, why rate the widget zero stars? If a computer can never get a truly random number, why tear apart a widget that comes as close as it can to being random, especially when no reasonable person is going to take the time to figure out the underlying mechanism?

were only able to determine that the numbers were not strongly correlated, which is more or less good enough for most people.

As I'm sure makes sense to a lot of people, flipping a coin usually isn't truly random either. Most coins aren't symmetric - there may be more mass on one side, aerodynamic factors, etc. It is a pretty good random event, though. Another excellent random event available to most people is a high-quality die. But you have to make sure to play on smooth felt …over a hard, smooth, level surface - the felt then should be smooth, and it prevents the die's corners from chipping.

The only random event I know of with any certainty is radioactive decay. Which also makes it appealing in the thought experiment with Schrodinger's cat. More