How Common Are Science Failures?

Estimates the base rate of entire scientific fields getting things near-unanimously wrong (to calibrate how much doubt that licenses about climate change). Defines a strict reference class -- post-1900, a fundamental field paradigm, stubborn resistance to a then-known-correct alternative, and now-undisputedly-wrong -- to exclude cheap shots. Finds three qualifying failures: Lysenkoism (easy mode, but Stalin-driven), Freudian psychoanalysis, and (more controversially) strong behaviorism. Against ~250 paradigm-possibilities (~50 fields x ~5 paradigms), that's a ~1.2% failure rate, maybe up to ~3.6% counting undetected ones -- no more encouraging for climate skeptics than the '3% of papers' they dispute.

Quality 72 · Strong. Strong, upper (72). A clever, genuinely useful epistemics exercise -- turning 'science has been wrong before' into an actual base-rate estimate -- with a rigorous, well-defended reference class and well-chosen examples. Held below Excellent by the back-of-envelope denominator and its narrow framing (climate-skepticism calibration).

Claude’s paradigm shift 55 · Moderate. Notable-edge (55). The base-rate-of-field-wide-science-failure framing is a fresh, non-obvious analytical move.

Real-world impact 2 · Minor. 2 -- minor/within-blog. A meta-science/climate-epistemics argument; within-discourse, no material footprint.