From Outside View to Intelligence Estimate - Using Analytic Confidence to Integrate Base Rates and Expert Judgment
Intelligence analysts routinely make probabilistic judgments under uncertainty, yet often give insufficient weight to base rates — the statistical frequency of relevant outcomes in comparable cases. The outside view (Kahneman, Tversky, Lovallo, and others) corrects intuitive forecasting by anchoring judgment in reference classes rather than in narrative salience. However, base rates alone cannot replace analysis: analysts often hold case-specific information justifying departure from a statistical baseline, and intelligence customers require a single usable probability. This paper proposes an intelligence-specific operationalization of reference-class forecasting that treats the base rate as the normative anchor, the analyst's inside-view judgment as a proposed departure, and analytic confidence as the moderating variable governing how much departure is warranted. The result is a single estimative probability combining three familiar inputs: a baserate probability, an inside-view probability, and a confidence assessment grounded in evidence quantity, evidence quality, domain expertise, and phenomenon predictability. The contribution is not mathematical novelty but a tradecraft rule — translating weighted regression toward a reference class into a communicable form for decision-makers.
Authors
- Alexander Jackisch
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-20
- DOI
- https://doi.org/10.5281/zenodo.24391929
- Primary Topic
- Intelligence, Security, War Strategy
- Type
- preprint