I think Chris jumped from 24-hour probability to cumulative probability (pointing out that what seemed like a small probability, would, in a matter of days, increase to a somewhat significant probability), and I followed right along with him. You are right that the 24-hour probabilies would have increasingly large denominators with the passage of time. If the probabilies did not fall off, then the denominators would remain the same for 24-hour probabilities, and the cumulative probabilities, I think, would simply be the sum of those fractions for X number of 24-hour periods. So, over a 10-day period, the probability would be 10:1000 = 1:100.

But there seems to be something wrong with that: in 1000 days, there would be a certainty . . .

Mike W.