The 0.5 derives from 'in a hundred year interval, there's a 50% chance to observe the event', which was my intuitive interpretation of a 'hundred year flood'. B

The 0.5 derives from 'in a hundred year interval, there's a 50% chance to observe the event', which was my intuitive interpretation of a 'hundred year flood'. But as so often in statistics, that intuition was wrong -- 1/100 is indeed correct. (I'm at least not alone in that erroneous interpretation: "A common misunderstanding is that a 100-year flood is likely to occur only once in a 100-year period."[0]) The better way of looking at it is probably: In a 100*n year window, you expect to see n such events, which makes it pretty immediately obvious that the yearly incidence probability is 1/100. [0]: https://en.m.wikipedia.org/wiki/100-year_flood#Probability https://en.m.wikipedia.org/wiki/100-year_flood#Probability