Although I think the point Gzus was alluding to is that anything with a finite probability can happen.
If there is a 1 in a million chance if something that doesn't mean it can't happen first time or even every time ... there is a difference between empirical probability and statistical probability anyhow.