Forecasting in a Probabilistic World: Why Reasoning Beats Prediction
There is a particular reproach I hear again and again, and it always arrives dressed as common sense. If you analysts live in a “probabilistic world,” the reader asks, then what is the point of your analysis at all? What can you actually predict? Or is probability just a convenient hiding place — a way of saying, after the fact, that the world is fifty-fifty, the forecast didn’t pan out, and nobody is to blame? It sounds devastating, and it is meant to. But underneath the sarcasm lies a misunderstanding so deep that untangling it tells us almost everything worth knowing about how to think about the future, the war, and our own helplessness before events. ...