Every forecast comes with a story: a new product, a huge market, a brilliant founder. The more detail, the more believable it feels, but detail isn't evidence.
The check is simple and rarely done: how often has this kind of thing actually happened before?