I think they are pretty clear that they are presenting it as the probability of being caught and not the correlation. It appears they're trying to show the dramatic increase between testing weekly (52 times) vs daily (260 times), and why it would be cost effective for you (the employer) to test daily.