Interesting, although the title is confusing to me. A chance of 1 in 50 million is only significant if you have a huge sample size. The odds might as well be 1 and 1 or 50 million to 1 because in the end it either ends or it doesn't and if it ends, the odds don't matter and if it doesn't they don't matter either.
Hmm, just wanted to vent my opinion on insignificant odds...
But thats a cool particle acellerator 'thingie' (I believe thats the technical term right).