Mantic Monday: Mantic Matt Y

A Mantic Monday on Matt Yglesias's 2021 prediction set (25 questions) and Metaculus's parallel forecasts, framed around the long-running project of holding pundits accountable via prediction. Key epistemics: the same-questions problem (you can't rank two pundits who answer different-difficulty questions; part of being a good pundit is knowing your lane); the artificiality gap (Yglesias's testable predictions only tangentially track his actual value-add, like explaining flaws in on-paper wealth stats); and the 'pundit's claim as a bid -- how would a portfolio buying each of their claims do?' framing, ideally against a prediction market. Scott's own practice: embed the predictions a post implies at its bottom. Wishes commentators who aren't really in the prediction business would just admit it.

Quality 67 · Strong. Strong (low): a substantive forecasting/pundit-accountability essay with the good 'claim-as-a-bid / market-omnipresence' framing, but partly a roundup of Yglesias's predictions.

Claude’s paradigm shift 48 · Moderate. Moderate/Notable: the 'score a pundit as a portfolio of their claims' and 'admit you're not in the prediction business' points are useful, non-obvious framings.

Real-world impact 2 · Minor. Minor/within-discourse: a forecasting-methodology essay; no material footprint.