SuperforecastingPhilip E. Tetlock, Dan GardnerProblem 1Human judgment about future events is often driven by intuition, narratives, and overconfidence, leading to systematically poor predictions. Replace intuition-driven certainty with probabilistic thinking that treats predictions as estimates with explicit uncertainty. Forecasting accuracy improves when judgments are expressed as calibrated probabilities rather than categorical claims, allowing continuous refinement and comparison against outcomes.Problem 2Individuals tend to anchor on initial impressions and fail to sufficiently update beliefs when new information emerges. Adopt an iterative updating process where beliefs are continuously revised in response to incremental evidence. Effective forecasters treat prediction as a dynamic process, using small, frequent adjustments rather than large, infrequent revisions, thereby maintaining alignment with evolving data.Problem 3Complex events are often oversimplified into single explanations, reducing analytical accuracy. Decompose problems into multiple contributing factors instead of relying on a single dominant narrative. Breaking questions into smaller components and evaluating each independently allows aggregation of partial insights into more accurate overall forecasts.