Mantic Monday 11/15

A substantive Mantic Monday anchored by a real deep-dive on the Karger/Monrad/Mellers/Tetlock reciprocal-scoring paper — the endogeneity-in-the-future problem with conditional prediction markets, the low-quality-forecaster convergence failure mode, the collusion worry, and the empirical result that reciprocal and traditional scoring performed about equally. Then Scott's own 'iterated prediction markets' riff (chaining five-year markets to amplify small edges, and why it might beat a single 80-year market), the Virginia-election markets-vs-media contrast, and the usual market roundup (street-votes zoning, Starlink, Metaculus calibration).

Quality 66 · Strong. Strong: pushes above the typical Mantic Monday on the reciprocal-scoring exposition and the genuinely clever iterated-markets thought experiment, which is Scott thinking out loud in a useful way. 66.

Claude’s paradigm shift 44 · Moderate. Notable-shift lower edge: good synthesis and one semi-original mechanism sketch, but built on others' paper and market ideas. 44.

Real-world impact 1 · Negligible. Forecasting-community discourse; no concrete material change. RWI 1.