Scott Alexander, curated
← Back to curation

Response To Alexandros Contra Me On Ivermectin

Quality
77
Excellent
Claude Shift
52
Moderate
RWI
3
of 10

Follows up on

Ivermectin: Much More Than You Wanted To Know — MMTYWTK · Nov 2021

Summary

Point-by-point response to Alexandros Marinos' 21-part rebuttal of the Ivermectin MMTYWTK. Concedes real errors (Biber — an egregious one, with an apology; Cadegiani points 2-3; the wrong summary-statistics test, which DerSimonian-Laird makes stronger not weaker), holds ground on others (Elalfy's randomization ambiguity, Ghauri's non-RCT, Borody's undescribed 10x-hospitalization 'control group', Carvallo's fraud signals), downgrades the Strongyloides hypothesis (~50%->35%), upgrades publication bias, and updates to 95% confidence ivermectin doesn't work given new negative RCTs. Models the equal-rigor principle and the mainstream-vs-contrarian PR asymmetry.

Why this score

Quality 77 · Excellent. Strong — a rigorous, unusually honest rebuttal/update (apologies to study authors, explicit Bayesian updating, the equal-rigor discussion, the C+/A-for-effort self-grade) and a model of good-faith adversarial epistemics; narrower than the original MMTYWTK (88) as a follow-up. 77.

Claude’s paradigm shift 52 · Moderate. Notable — a sharp rebuttal-and-update; the equal-rigor and mainstream-PR points are incisive but not paradigm-shifting.

Real-world impact 3 · Moderate. Part of the influential ivermectin discourse. 3.