drethelin comments on Open thread, Feb. 23 - Mar. 1, 2015 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (161)
I think homogenous mass prediction gets you better calibrated at predicting the kind of questions in the sample group but doesn't help a lot otherwise. If you have ever played the the calibration game that someone made a couple years ago it's a lot like that. After using it for a while you get better and better at it, but I don't thikn that made me better calibrated in general.