For fun, I had a chat with a street preacher at lunchtime. He told me a story which I thought I'd examine critically. He said that he once flicked through the bible at random and came up with a page that spoke to him with huge relevance at that time. His wife said it was a coincidence, so he showed her what he'd done and came up with the same page the second time. He then did it a third and a fourth time, in a book with 2000 pages (or thereabouts). This was proof to him that there was some force at work.

I won't deny or agree with the putative divinity in that story. Let's just do some stats, though. Let's ignore the idea that the book has a flaw in its spine, predisposing it to open at that exact page. Let's use the idea that he was truly random in his page selection, first of all.

Probability of choosing the same page four times is 1/2000 to the power of 4. This is one in 160 billion. Quite improbable. Wow.

Then let's assume, as I have done, that, in fact, he replicated the act of searching at random a bit less randomly than he imagined, going to approximately the same 300 page region of the book each time. Somewhere near the middle, maybe, or maybe near the end. A 300 page range wouldn't be a bad guess for one's flicking-at-random-target-area. That would make it a one in 810 million, which is very long odds, but comparable to the sole-winners of the one in 14 million national lottery draws.

Now, let's use a quick experiment I did with my Uri Geller book. I tried to flick to about the same page each time. I got a range of 40 pages. So, let's imagine that this guy repeated his steps really well. The probability of hitting the same page each time is one in 2.5million. Lottery winners beat those odds every week.

So, even if the book wasn't biased, the significance of reaching the same page each time seems to be statistically less universe threatening than at first imagined.