You start with the story
Kahneman and Tversky called this the inside view: you judge a case by its own vivid details and build a plausible narrative. It feels rigorous, but it quietly assumes things go to plan.
News from Now On · Lab Knowledge Byte
Ask the base rate first
The single question that most improves a forecast: how often does this kind of thing usually happen?
Start with the record
The outside view asks a different question first: what happened to the last hundred cases like this? That historical frequency—the base rate—is your anchor before any story adjusts it.
Stories run optimistic
Across 258 major transport projects, nine in ten went over budget, with an average cost overrun of 28%. The inside-view estimates weren't unlucky—they were systematically too rosy.
It became official policy
The UK Treasury now requires reference-class forecasting: base the estimate on how similar past projects actually turned out, then adjust. Do the same with any headline—anchor on frequency, then move.