Over the last few years, in part due to Less Wrong, I’ve stopped paying attention to short-term issues. I basically don’t consume news anymore as I don’t like wasting time learning facts that will be outdated in two-weeks. I think this is a pretty good strategy for information consumption but it means I have very few stable/reasonably distributed predictions with one-year horizons… and they’re almost all basketball predictions.
Is it worthwhile to read political and economic news just so I have more things to test my calibration on? Would reading more news even improve my predictions or just make me more/falsely confident in them?
Is it worthwhile to read political and economic news just so I have more things to test my calibration on?
No. You want to calibrate yourself? Run through the dozen or so trivia calibration datasets which have been linked on LW, or work your way through the >=136 predictions for 2011 on PredictionBook.com.
Would reading more news even improve my predictions or just make me more/falsely confident in them?
If I were to read the NYT like I used to, cover to cover, I think my predictions would improve. Hidden in obscure articles were many early indicators of things to come.
If you were to read it, or were to read USA Today, I think your predictions would get worse than the sort of dumb Outside View-like prediction you would otherwise have been forced to make (‘Well, I don’t know much about the Korea situation or this succession thingy, but I do know there hasn’t been an active war in, like, half a century so I’ll put a low chance of hostilities being renewed’).
Over the last few years, in part due to Less Wrong, I’ve stopped paying attention to short-term issues. I basically don’t consume news anymore as I don’t like wasting time learning facts that will be outdated in two-weeks. I think this is a pretty good strategy for information consumption but it means I have very few stable/reasonably distributed predictions with one-year horizons… and they’re almost all basketball predictions.
Is it worthwhile to read political and economic news just so I have more things to test my calibration on? Would reading more news even improve my predictions or just make me more/falsely confident in them?
No. You want to calibrate yourself? Run through the dozen or so trivia calibration datasets which have been linked on LW, or work your way through the >=136 predictions for 2011 on PredictionBook.com.
If I were to read the NYT like I used to, cover to cover, I think my predictions would improve. Hidden in obscure articles were many early indicators of things to come.
If you were to read it, or were to read USA Today, I think your predictions would get worse than the sort of dumb Outside View-like prediction you would otherwise have been forced to make (‘Well, I don’t know much about the Korea situation or this succession thingy, but I do know there hasn’t been an active war in, like, half a century so I’ll put a low chance of hostilities being renewed’).