I think the overcoming of something we'd assign near zero probability is pretty cool, they get into that later in the article.
Gets into the difference between assigning random variables and actually knowing what's going on w/ causation "behind the scenes" of our mathematical designations.