I should probably get around to reading Black Swan one day or another. I know one thing for sure — statistics is garbage at predicting rare events. Exhibit A is Hurricane Harvey. If you try and back-figure what “return period” storm Harvey was in Houston, the answer is something like a 500,000 year storm. It’s so off the charts we don’t have good math to show what the charts are like out there, and the math we are using to plot the charts out to, say, 1000 years is clearly faulty. Harvey proved it.

There’s an article in there somewhere, but I’m afraid even as a professional hydrologist and licensed civil engineer I might not be qualified to do the topic justice. It would need some deep, deep math.