Book Review: Superforecasting

Review of Tetlock's Superforecasting. Summarizes the fox/hedgehog finding, the IARPA-spawned Good Judgment Project (superforecasters beat CIA analysts with classified data by ~30%), and what makes them good: probabilistic thinking, inside/outside view, granular probabilities (rounding hurts accuracy), willingness to update, reduced cognitive bias / scope insensitivity. Scott's sharp original notes: 'superforecaster' just means top-2%, not a natural kind ('2% of forecasters were in the top 2%'); and the whole thing is the LW/CFAR rationality tradition in respectable book form -- useful as high-status validation more than as news to insiders.

Quality 70 · Strong. Strong, upper (70). A lucid, useful review of an important book with a couple of genuinely sharp original observations (the top-2%-isn't-a-discontinuity point; 'a Bayesian who doesn't use Bayes' theorem'). Slightly below the SSC book-review average (~71-72) because it's a brisk summary-plus-critique Scott himself frames as not-too-useful-if-you-know-the-tradition, rather than a deep-dive.

Claude’s paradigm shift 48 · Moderate. Moderate (48). The ideas are Tetlock's; Scott's value-add is the top-2% debunk and the 'it's the rationality tradition' framing -- a modest, fresh reframing within a review, not new in its moment.

Real-world impact 3 · Moderate. 3 -- moderate. Superforecasting/GJP have real reach (forecasting tournaments, EA/Metaculus forecasting culture, even IC practice); the review itself is a niche-sphere amplifier that channeled rationalist readers toward that world -- niche/professional influence, the broad reach credited to Tetlock.