Check whether you are estimating frequency or vividness
When making frequency or probability estimates, pause and ask: 'Am I estimating actual frequency, or how easily I can picture this?' then look up the base rate before deciding.
When This Fires
- Estimating project failure probability ("most projects like this fail")
- Assessing risk of rare events (security breaches, market crashes, illness)
- Deciding how much to worry about edge cases versus common cases
- Any time you say "this happens all the time" or "this almost never happens" without data
Common Failure Mode
Trusting your frequency estimate because you can recall multiple examples. Recalling three vivid examples of a project failing feels like strong evidence for a high failure rate — but your memory is biased toward dramatic, unusual, and recent events. The hundreds of projects that succeeded quietly left no memorable trace. Your sample is biased toward what's vivid, not what's frequent.
The Protocol
When you catch yourself estimating a probability: (1) Pause and ask the diagnostic question. (2) If your estimate is based on how easily you can picture the outcome, flag it as availability-biased. (3) Look up the actual base rate — how often does this class of event actually occur? (4) Anchor your decision to the base rate, adjusting only for factors specific to your situation that the base rate doesn't capture.