Predictions feel like insight, but they are often extrapolations from a small, vivid set of cues. Intuition overreacts to the story in front of you.
A better approach is to start from the outside: ask what typically happens in similar cases, then adjust—carefully and modestly—for the unique details.
This method fights two temptations at once: ignoring base rates and ignoring regression. It replaces “I can see it happening” with “How often does it happen?”
The slow system is required here. It must build a reference class, consult it honestly, and resist the urge to over-adjust toward a compelling narrative.
Good forecasting is rarely dramatic. It is disciplined, a little boring, and usually less confident than the prediction your intuition wanted to make.