> The assignment was this: the student either had to flip a coin a hundred times and write down the sequence of results, OR make up the results of the coin tosses and write down a "random" sequence of a hundred heads or tails.
>
> The papers were turned in the next day, and for EVERY assignment, the teacher correctly guessed whether the student had used a real coin or faked the results.
>
> The secret turned out to be some property of random sequences that human nature misunderstands. Unfortunately I don't remember the exact numbers, but it turns out that the probability of having something like six or seven heads in a row in a sequence of one hundred coin tosses is around 99.9%. But a student faking a sequence of a hundred coin tosses is highly unlikely to include a string of that many heads.
>
> I wish I remembered the exact numbers.

My maths is very rusty but, if you go for a sequence of six in a row, then when you throw the first there's a 1 in 32 chance that the next five will agree. Same for the second, all the way to the 95th. So 95 attempts at a 1 in 32 chance - um - certainly well over evens! Anybody out there study probability?