Hacker News new | ask | show | jobs
by vkou 528 days ago
> You’re assuming “it was tracked to there” = “it originated there” but that’s a big leap.

If the market were one of multiple second-generation infection spreader sites, we'd have had significantly faster growth in the rate of infections.

With the power of hindsight, the rate of spread of COVID is well-understood, and the timelines of people getting sick is consistent with one initial outbreak site. There just aren't any known cases where someone came down with COVID on day 1 of the outbreak, without also having gone to that market.
1 comments
danenania 528 days ago
All of that is subject to detection bias as well. There’s a lot of uncertainty involved in tracking early cases. It’s not anything close to solid evidence.
link
vkou 528 days ago
It would be a damn weird form of bias when every early case that could be accounted for was somehow tracked to only the market.

The location bias hypothesis would work if they looked at all patrons of the market, and checked if they got sick. That's not what they did, though, because they didn't have a list of all the patrons of the market. It doesn't have a guestbook.

They looked for people who got sick, and asked them where they went before they fell sick. And all of them turned out to have gone to the market - and not on the same day.

link
danenania 528 days ago
The exact methodology of how they found these people is very important, as there are many potential sources of bias, not to mention conflicts of interest.

I did some searching and it seems that you are perhaps basing your conclusions on this article/the studies it points to? - https://www.thelancet.com/journals/lanmic/article/PIIS2666-5...

> A geospatial analysis reports that 155 early COVID-19 cases from Hubei Province, China, in December, 2019, significantly clustered around a food market in Wuhan, China.

However, I couldn't find details on how these 155 cases were selected or what exactly "significantly clustered" means. 155 is a small sample, so the details are important.

Also important is whether the data on these cases was provided by government sources. If so, I would question how reliable or representative that data is.

link