How I almost outsmarted 3 pro forecasters using Bayes rule.
Trying to out-forecast Vox in 2020. What could go wrong?