Decision frameworks · Reference-class forecasting

Reference-class forecasting: base rates for decisions

Updated 20 July 2026 · Advisor: Strategist

Reference-class forecasting predicts how your decision will turn out by looking at how a class of similar decisions actually turned out — the base rate — and adjusting from there. It's the antidote to the optimistic "inside view", where you reason only from your own case and quietly assume you'll beat the odds.

Cerno is a private AI boardroom that runs consequential decisions through a structured process and hands you an Executive Decision Brief. Reference-class forecasting is one of the lenses it applies; here's the method itself.

What is reference-class forecasting?

Developed from the work of Daniel Kahneman and Amos Tversky, it corrects a predictable bias: when we estimate, we take the inside view — building up from the details of our specific plan — and consistently underestimate cost, duration and failure. The fix is to take the outside view: find a reference class of similar past cases and start from how they actually went.

If comparable expansions took 40% longer than planned, your expansion probably will too, however special yours feels.

When should you use it?

Use it for any estimate that matters — timelines, budgets, adoption, success odds — especially when you're tempted to think "our situation is different." The more confident and specific your inside-view number, the more you need the base rate as a check.

How to run it in 5 steps

›Define the decision as a prediction. "This project will take X months / cost £Y / succeed with Z% odds."
›Pick a reference class. A set of genuinely comparable past cases — yours and others', broad enough to have a real base rate.
›Find the base rate. How did that class actually turn out on the dimension you care about?
›Start from the base rate. Make it your anchor, not your optimistic plan.
›Adjust — cautiously. Move off the base rate only for specific, evidenced reasons, and less than instinct tells you.

Worked example

Considering a market entry, the base rate reframes the confidence:

Executive Decision Brief

Should we expand into a new market this year?

Recommendation

Enter, but with a lightweight test first — the base rate for cold-start market entries argues against a full commitment up front.

Confidence

—

Risk

Med-High

Vote

3–2

Illustrative brief · Evidence Mode on

Which Cerno advisor uses this?

The Strategist

Strategist

Cerno's Strategist advisor reasons from base rates and reference classes: it asks how comparable decisions actually turned out and starts there, so the brief's confidence reflects the outside view rather than optimism.

About to trust an optimistic estimate? Put the decision through the boardroom and let the Strategist check it against the base rate.

Check the base rate in Cerno

Frequently asked

What's a 'base rate'?

The rate at which something happens across a group of similar cases — for example, the share of comparable projects that finished on budget. Reference-class forecasting starts from the base rate, then adjusts, rather than starting from your own optimism.

Why not just trust my own judgement?

Because the 'inside view' — reasoning only from the specifics of your case — systematically underestimates cost, time and risk. The base rate (the 'outside view') corrects for the optimism you can't see in yourself.

Inversion: the mental model, explained

Inversion means solving a problem backwards — instead of asking how to succeed, ask what would guarantee failure, then avoid it. Here's how the model works, with a worked example and the five-step method.

Second-order thinking, with examples

Second-order thinking asks 'and then what?' — tracing the consequences of the consequences, not just the first result. Here's how to do it, why first-order thinking misleads, and a worked example.

Pre-mortem: the template, with worked examples

A pre-mortem imagines a decision has already failed, then works backwards to find why — surfacing risks before they happen. Here's the 5-step template, worked examples, and how to run one properly.