The CIA has concluded that Covid-19 probably began as a leak from a laboratory in China in a new assessment of the origins of the pandemic that killed millions of people.

https://www.ft.com/content/9880273c-8517-4502-abf0-e667319ea6bd?s=09

I've done a different statistical test: place candidate "epicenters" randomly in Wuhan weighted by the population density and measure the summed distances to residences of cases. As you can see, the p-value is ~0.01 for how close Huanan market is to case residences vs how close random locations are to case residences. It remains statistically significant to p<0.05 if I use a non-linear weight for placing candidate epicenters; important, since Huanan market isn't in one of the low density parts of Wuhan (it's not in an incredibly high-density area either, though).

The X-axis starts at zero because I'm subtracting the summed distance from the geometric median location of cases to each case, which is the coordinate that minimizes that sum.

Here's what 999 randomly placed candidate epicenters looks like.

One of those things where the eye test says it all IMO.

From the erratum: "The authors therefore generated new versions of these three files with n = 1000 and re-ran all statistical tests in which any of the three files were used. All results remained the same as previously reported."

This is not the kind of thing they'd retract over.

@WilliamGunn

1) Yes, you are correct on that point. I'm more curious if there really is a "vibe shift" happening.

2) FYI, this is a year old, but in the errata, Lisewski claims: "The statistical claims and corresponding epidemiological interpretations of lineage A and B cases in Worobey et al. are not supported by their statistical data."

3) I think this market will probably resolve NO, but I am uncertain, so I created it.