Predicting Puzzles and Imagining Mysteries
Art has a strong post up discussing what can and cannot be predicted. His thesis distinguishes between events that can be forecast given enough aggregated information (which are roughly equivalent to “puzzles”) and events that cannot be forecast because they actively seek to break past patterns (which are roughly equivalent to “mysteries’). The former can benefit from tools like prediction markets, computer simulations and actuarial techniques; the latter require imaginative strategic thinking.
