Base rates
Anchor on how often something happens in the reference class before adjusting for the specific case.
The classic Kahneman-Tversky finding: when judging the probability of something specific, humans almost completely ignore the base rate of how often it happens generally. We jump straight to the specific story.
The fix is to start with the base rate, then adjust. The base rate of new restaurants failing is roughly 60% within five years. Your friend's specific restaurant might be better or worse than average, but the conversation should start at 60%, not at "this one looks great because of the chef."
For operators, base-rate thinking matters for:
- Estimating likelihood of project success (most projects of this type take how long? have what success rate?)
- Hiring (what fraction of similar candidates work out at this seniority?)
- Strategic decisions (how often do similar acquisitions create value? how often do similar pivots succeed?)
Where you have base-rate data, use it as the prior. Then adjust based on what's actually distinctive about this case. Most strategic mistakes happen because the specific case overrides the base rate without earning the right to.
Examples in the wild
Most M&A creates negative shareholder value. That's the base rate. Any specific deal needs strong evidence to overcome it, but most deals get approved on "this one is different" without any rigorous case.
The base rate of active equity managers underperforming the index is roughly 80-90% over 15+ years. Any specific manager has to clear that base rate before any case for hiring them makes sense.
Worry about plane crashes vs. car crashes: the base rates are 1 per 4M flights vs. 1 per ~100M km driven. The car is much more dangerous per hour spent. Most people get this exactly backward.
Base rates is one of the mental models we apply through real cases inside the Pareto MBA — a part-time program for professionals who want to think clearly about business.