Hacker News Reader

> a 1% chance per year, that implies a 63% probability per century.

I'm not sure I'm following the math here?

1 - (1 - 0.01)^100

Thanks, now I understand how they reached the number but not why it's so convoluted. I only took one stats course in college.

So it's

P(bomb exploding in a century) = 1 - P(bomb not exploding in a century) = 1 - (P(bomb not exploding in a year)) ^ 100 = 1 - (1 - P(bomb exploding in a year)) ^ 100

Why is it not something simpler, like P(bomb exploding in a year) ^ 100 ?

> Why is it not something simpler, like P(bomb exploding in a year) ^ 100 ?

That would be "bomb exploding EVERY year (= this year AND next year AND ...)".

What we actually want is "bomb exploding this year OR bomb exploding next year OR bomb exploding this year and next year OR bomb exploding in two years OR ...". But that's too complicated, so we use the trick with 1 - P(bomb not exploding in a century).

That makes sense. Thanks for taking the time to clear up a stranger's confused mind! :)

Maybe cause nuclear weapons only existed for 70 years ...?

Who said otherwise?

Putting it in terms of "probability per century" is intended for forward-looking projection, not backwards.

Reply on news.ycombinator.com