Retrospective on Predictions for 2016

I’m doing this a year late since data usually takes a while to update. In any case, here were my original predictions with remarks as to their accuracy edited in:

Here was my calibration curve:

As you can see, not very good, especially for 70 and 80%. Interestingly, my 60% confidence predictions were much more accurate than those I had 70% and 80% confidence in. I’ll try to avoid that next time. Predictions for my personal life were the least accurate (as they are for most people making predictions).

I didn’t grade the 50% ones (for obvious reasons), but my predictions for a general election HRC victory and Brexit (50% for each) were much more accurate than consensus at the time. I turned out to be very accurate for China’s, Mexico’s, and Russia’s economies, but was wrong on Belarus’ recording GDP growth. “There will be at least two IS-inspired terrorist attacks in the United States that will each kill at least five people” came close to being true.