summary - A methodological critique, by two uninvolved spatial-statistics researchers, of Worobey et al. 2022’s central geospatial argument. Argues the statistical case is invalid on two grounds: (1) the assumption that the centroid of early case locations marks an epidemic’s true origin point is unproven for the class of spatial-epidemic models it would need to hold under, and (2) the Monte Carlo significance test used to rule out other candidate locations does not correctly represent the null hypothesis it claims to test. First posted to arXiv in August 2022, shortly after Worobey 2022’s publication; Worobey and coauthors posted a reply defending their method (arXiv 2403.05859, “Confirmation of the centrality…”). Separately, Science issued a 2024 erratum correcting an unrelated distance-calculation error in the original paper’s data — not a response to this critique, but corroborating that the original analysis contained at least one acknowledged error.
relevance_note - The most direct methodological challenge found to Worobey 2022’s statistical “proof,” discovered while snowballing per the brief’s explicit request; does not itself argue for lab-leak, only that the single most load-bearing zoonosis-side epi paper’s central statistical claim is weaker than presented.