One habit was a lot like the analogical thinking that helped the venture capitalists and movie enthusiasts in chapter 5 make better projections of investment returns and film revenues. Basically, forecasters can improve by generating a list of separate events with deep structural similarities, rather than focusing only on internal details of the specific event in question. Few events are 100 percent novel—uniqueness is a matter of degree, as Tetlock puts it—and creating the list forces a forecaster implicitly to think like a statistician.
