When I first learned of the Monty Hall problem in an undergrad Discrete class, I wrote a program to simulate it. It was pretty easy to see it's statistically correct, at least.

You'd think those PhDs would have at least tried it a few times, to see if it appeared correct in practice, before writing a signed letter to a magazine.

But many people tried performing similar experiments on computers, fearlessly programming them in hundreds of different ways. Not surprisingly, they fared a little less well.

Kinda surprises me. My app was 50 (non-boilerplate) lines of Java. If you can't write an app to demonstrate 3 simple logic rules, which a university sophomore can (did) in an hour, I uh, hope you don't make a career out of it.