── ── Cognitive bias
Representativeness Heuristic
The representativeness heuristic is judging probability by how closely something resembles a prototype — overriding actual base rates. Named by Tversky & Kahneman (1972); produces three systematic errors: base rate neglect, the conjunction fallacy (A-and-B feels more likely than A), and insensitivity to sample size.
Run Representativeness Heuristic on a real problem
Bring something you're actually deciding — free, in the browser.
How it works
Step 1 — Identify judgment + profile: What probability estimate is being made? What profile information is driving it?
Step 2 — Reference class + base rate: What is the base rate of this outcome in the relevant reference class? Was it stated before the profile assessment? (If N while a profile is present, representativeness is active.)
Step 3 — Conjunction test: Does the judgment involve multiple attributes (X and Y and Z)? P(A AND B) ≤ P(A). If the conjunction was rated more probable than a component, the conjunction fallacy is active.
When to use it
- user says 'this person/startup looks like a winner', 'I know one when I see it', 'they have the profile of a great hire', 'this reminds me of [famous success]', or any judgment where a vivid profile drives a probability estimate without an explicit base rate
- Also activate when auditing for the conjunction fallacy or when pattern-matching is cited as the basis for a high-stakes decision
When not to use it
When the decision is routine and reversible, applying a formal method costs more than it returns.
Worked example
"The Next OpenAI" — Representativeness in AI-Startup Investing (2023–2026)
After ChatGPT's launch in late 2022, venture investing tilted heavily toward AI. By 2024–2025, a recurring pitch-meeting sentence became the tell: "This founder looks like the next OpenAI" or "this company has the profile of the next Stripe." The judgment was driven by resemblance to a celebrated prototype — ex-frontier-lab researchers, a Stanford/MIT pedigree, a demo that feels magical, a fast-rising usage chart — rather than by the base rate at which such companies actually become durable winners. This is the representativeness heuristic operating at…
Install this skill (free, MIT)
npx skills add deciqAI/knowledge-skillsUseful? Star the repo — stars help other builders find it.
FAQ
What is the representativeness heuristic?
Judging probability by resemblance to a prototype instead of by base rates. If a description sounds like a librarian, we judge the person likely to be one — even when the population contains vastly more of the alternative. Named by Tversky and Kahneman in 1972.
What is the conjunction fallacy?
Rating 'A and B' as more probable than 'A' alone because the combination resembles the story better — famously, judging 'Linda is a bank teller and a feminist' more likely than 'Linda is a bank teller.' A conjunction can never be more probable than either of its parts.
How does this bias hurt startup judgment?
Pattern-matching is the investor's and founder's default: 'this looks like the next Airbnb' substitutes resemblance for base rates, and small-sample wins get read as repeatable. The correction is always the same question — what's the base rate for things that look like this, and how big is my sample?
Related mental models
Self-renewal is the structured practice of identifying which mental models have passed their expiry date and genuinely updating — not merely annotating — them.
Status quo bias is the systematic preference for the current state over available alternatives — even when alternatives are objectively superior by the person's own values.
Most communication failure is a failure of sequence and mode — not information.
Survivorship bias is drawing conclusions from a sample pre-filtered by survival — treating survivor traits as the cause of survival when non-survivors (absent from data by…