Scott Alexander on Bayesian analysis is the right framewor…

Scott Alexander

Author and psychiatrist

The six estimates span twenty-three orders of magnitude. Even if we remove Peter (who’s kind of trolling), the remaining estimates span a range of ~7 OOMs. And even if we remove Saar (limiting the analysis to neutral non-participants), we’re still left with a factor-of-50 difference.

AI Verified source (Mar 28, 2024)

3mo ago

Policy proposals and claims

Bayesian analysis is the right framework for settling the COVID-19 origins question

votes Against

Statement relation verification history AI Verified Report this

AI Verified Yes. In context, the author is explicitly assessing a Bayes-based attempt to resolve COVID origins: he says Rootclaim asks what happens if you 'directly apply Bayes' to hard problems, then says 'it doesn’t work,' and cites the huge spread among multiple Bayesian estimates as evidence that the method failed to resolve the dispute. That makes the quote directly relevant and makes an opposing stance substantially more likely. ([astralcodexten.com](https://www.astralcodexten.com/p/practically-a-book-review-rootclaim)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

AI Verified Relevant: in context, this quote is part of the author’s argument that different participants’ Bayesian analyses of COVID origins produced wildly divergent results, and the author then says that directly applying Bayes here "doesn’t work" because the evidence is too hard to quantify. That makes a determinate stance on whether Bayesian analysis is the right framework for settling the COVID-19 origins question substantially more likely. ([astralcodexten.com](https://www.astralcodexten.com/p/practically-a-book-review-rootclaim)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

Vote inference verification history Latest opinion AI Verified Report this

AI Verified Against: the quote emphasizes that the Bayesian-style estimates 'span twenty-three orders of magnitude' and still differ by a 'factor-of-50' even after excluding some participants; in the surrounding article, the author says 'it doesn’t work' and calls that only 'faint praise' for a method that was supposed to resolve the disagreement. That implies he rejects Bayesian analysis as the practical framework for settling COVID origins, even if not Bayes in the abstract. ([astralcodexten.com](https://www.astralcodexten.com/p/practically-a-book-review-rootclaim)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

AI Verified Against — the quote presents the Bayesian attempts as failing to converge: the estimates 'span twenty-three orders of magnitude' and still show a 'factor-of-50 difference.' In context, the author says this kind of full Bayesian treatment 'doesn’t work' here because the evidence is too hard to quantify, and calls that only 'faint praise' for a method that was supposed to resolve the dispute. So while he allows Bayes in principle, he is opposing it as the right practical framework for settling COVID origins. ([astralcodexten.com](https://www.astralcodexten.com/p/practically-a-book-review-rootclaim)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

Quote authenticity verification history

Report this

AI Verified Verified: the source page contains the exact quoted text verbatim at lines 388-388, and the same page identifies the post as by Scott Alexander and dated Mar 28, 2024. The stored author, date, source URL, and quote text all match. ([astralcodexten.com](https://www.astralcodexten.com/p/practically-a-book-review-rootclaim)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

AI Verified Verified: the three-sentence passage appears verbatim in the cited Astral Codex Ten post at lines 388-389, and the same page identifies the post title, author Scott Alexander, and date Mar 28, 2024 at lines 7, 13, and 15. The stored quote text, author, date, and source URL match the source. ([astralcodexten.com](https://www.astralcodexten.com/p/practically-a-book-review-rootclaim?hide_intro_popup=true)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

replying to Scott Alexander