This is a Lexicon entry: the mechanism, when it shows up, and the countermeasure. The full correction, with real-world cases and the audit prompt, hasn't been written yet.
Definition
A mental shortcut where people judge the probability of an event based on how easily recent examples come to mind. Plane crashes feel more common than car crashes because they are more vivid and memorable - not because they are more frequent.
When it shows up
Estimating frequency of events soon after exposure to a vivid case: plane crash coverage, a friend's cancer diagnosis, a recent shark attack. Any probability judgment where examples come to mind unevenly. Modern context: After a high-profile AI system failure makes headlines, product managers across an industry overestimate the probability of identical failures in their own systems and over-invest in that specific failure mode, under-investing in statistically more common but less salient ones. High-stakes instance: Policy response - regulatory and legislative responses to rare but vivid events (plane crashes, shark attacks, terrorist incidents) systematically over-resource low-probability high-salience risks relative to high-probability low-salience ones. Failure mode: Recent, emotional or widely reported examples load the mental sample. The estimate tracks memorability rather than frequency, producing systematic overestimates of rare-but-vivid events and underestimates of common-but-mundane ones. Estimated cost: High - often irreversible where this bias dominates unchecked decisions.
Failure mode
Recent, emotional or widely reported examples load the mental sample. The estimate tracks memorability rather than frequency, producing systematic overestimates of rare-but-vivid events and underestimates of common-but-mundane ones.
Countermeasure
Before estimating, ask: 'What is the base rate in the relevant population?' If you cannot answer without looking it up, the memory-sample is what you are using. Look it up. When correction costs more than the bias: When the vivid recent example is actually representative and the historical base rate is stale. In fast-moving domains - technology, markets, geopolitics - recent data genuinely predicts better than long-run averages. Anchoring mechanically on base rates in a discontinuously changing environment can produce worse estimates than availability-driven judgment.
Related traps
Also connected in the map19 more, locked
Credit & first seen
DiscoveredTversky, A.&Kahneman, D. (1973).'Availability: A heuristic for judging frequency and probability.'Cognitive Psychology 5(2), 207-232 source ↗
Also revealed
Entry #5 of 631 in The Lexicon · see the full Lexicon