4,448 followers
RT @Aaroth: @nikete Also true of any other empirically testable feature of the forecasts! https://t.co/EM2ZP3ky9Y Its possible the Manifold…
RT @Aaroth: @nikete Also true of any other empirically testable feature of the forecasts! https://t.co/EM2ZP3ky9Y Its possible the Manifold…
@nikete Also true of any other empirically testable feature of the forecasts! https://t.co/EM2ZP3ky9Y Its possible the Manifold forecasts are great, but yes, that calibration plot is non-responsive to the midwits' concern.
This result is from a short (9 pages!) and sweet paper by Alvaro Sandroni from 2003 called "The Reproducible Properties of Correct Forecasts": https://t.co/EM2ZP3k0kq