zlacker

I wrote this on 2020-01-24 (when that speculation was fairly new), and wonder why that sort of simple statistical argument has rarely been made explicit:

It certainly is an interesting coincidence that the only lab in China that can deal with it happens to be in Wuhan. The question is, how big of a coincidence. If the disease hit a random person randomly uniformly anywhere in China, the probability that it would have happened in Wuhan is a bit less than 1% (as there are about 10+m people in Wuhan, and 1400+m people in China).

If you think it might have struck randomly any city above a million people in China uniformly, it’s also roundabout 1% (as there are about 100 of those).

So this is by no means proof that something fishy happened, but it is significant enough to warrant investigation.

If you assume that this could only have happened in a city with, say, more than 5m people, Wuhan is one of about 15 to 20 of those (so we're just above the "usual" 5% significance threshold).

Still, an independent investigation of that lab seems warranted. Of course it’s China, so unlikely to happen...

(I must say that I think the comment has stood the test of time, so far.)