Post #558: What are the odds …

Posted on March 17, 2020

… that a randomly-chosen seemingly-healthy adult in Virginia is infected with coronavirus.  All other things being equal.

For example, suppose that a stranger accidentally brushed up against your hand at the store.  Is that tantamount to attempted murder?  Or is it more likely just a harmless social faux pas?

Here’s my estimate, based on facts I’ve presented in various recent posts.

Virginia currently has 67 confirmed cases of COVID 19 (5 PM 3/17/2020, per Johns Hopkins).  Assume those are all adults.

Per the Chinese experience, about 80% of all cases are non-severe cases.  Let’s be conservative and assume that none of those non-severe cases got tested, so that the 67 cases reflects just 20% of the true total.  So the true total is five times higher.

On average, there’s about a six-day latency between being infected and showing symptoms.  And almost everyone who is infected will show symptoms — the Chinese report that true asymptomatic cases are extremely rare.

Finally, in the US, the count of diagnosed cases has been doubling roughly every three days or so.

So an estimate for the total number of infected adults, not yet showing symptoms, in Virginia would be six days’ growth in five times the current case count.  Or growth (2 x 2) times baseline count (67 x 5), or 1,340 individuals.

From Census data, there are about 6.6M adults (18+) in Virginia.  That makes the odds roughly 1 in 5000.  At the outside.  The likelihood that any random adult you might meet, here in Virginia, is currently infected but showing no symptoms, is almost certainly no more than 1 in 5000.

But before you had that person arrested for assault, you’d have to figure the probability of getting infected based on that incidental contact.  Presumably, it’s fairly low, or we wouldn’t have (e.g.) Amazon delivery guys, store clerks, mail carriers, and similar professions going on about their business.