Hmmm, Mr. Subliminal was more correct than me in this case, because I was thinking along the lines of Bayesian statistics but my interpretation of it was wrong.
I figured the following:
We know for a fact that out of a sample of 500 people, 1 will definately be infected.
We also know...