One of the two judges’ written decisions in the Rootclaim vs. Peter Miller debate. Van Treuren describes his process (giving each side benefit of the doubt on uncontested claims, relying only on debater-cited sources, ~750 slides and ~70 hours reviewed) and reports two convergent methods, both favoring zoonosis. His explicit Bayesian calculation, modeled on Michael Weissman’s framework, gives a prior odds ratio P0(LL)/P0(ZO) = 1.7x10^-3, an evidential update of 2.125, and a final ratio of 3.6x10^-3 — i.e. roughly 1-in-300 (~0.3%) probability of lab leak. He identifies the probability WIV would undertake DEFUSE-style gain-of-function work, and the outbreak’s epicenter at the Huanan Seafood Market, as the two factors weighing most heavily against lab leak. This is one of the case’s “six independent [Bayesian] analyses.”
relevance_note: one of the two judge decisions the case explicitly anchors on (“two expert judges ruled decisively for zoonosis”), and one of the six point estimates behind the “23 orders of magnitude” spread.