Anchor on the base rate before adding inside-view details

Start your forecast from the class median, then adjust — do not start from your narrative and adjust to the base rate.

Why it works

Anchoring research shows that adjustments from an initial number are insufficient — people consistently end up closer to their starting point than the evidence warrants. If you start from the inside view (which is optimistic) and try to adjust downward toward the base rate, you will not adjust far enough. Starting from the base rate and adjusting upward for genuine distinguishing features of your project produces much more accurate final estimates.

How to do it

1. Set the base-rate median as your initial estimate.
2. List only specific, verifiable features that make your project genuinely different from the class average.
3. For each feature, estimate the directional and magnitude adjustment it warrants — and be conservative.
4. Treat the final estimate as a point on a distribution, not a point, and communicate the range.

Evidence

Anchoring and insufficient adjustment is one of the most replicated findings in judgment research (Tversky & Kahneman). Applied to forecasting, it predicts that base-rate anchoring will outperform inside-view anchoring with adjustment, which is what Flyvbjerg’s empirical work on projects confirms. (observational)

In domains where the base rate is poorly estimated or the population is genuinely heterogeneous, base-rate anchoring can be as misleading as inside-view anchoring.

Sources

• Tversky & Kahneman (1974), "Judgment under uncertainty: Heuristics and biases," Science
• Flyvbjerg (2008), "Curbing optimism bias and strategic misrepresentation," European Planning Studies

Common mistake

Using the base rate only as a sanity check at the end rather than as the starting anchor — at that point the inside-view estimate is already committed and the base rate is rationalized away.