Probabilistic thinking means estimating the likelihood of different outcomes rather than assuming any single prediction is certain, and updating those estimates as new information arrives.
The idea
Most real-world questions don’t have a single guaranteed answer — they have a range of possible outcomes, each with a different chance of happening. Probabilistic thinkers assign rough odds to outcomes instead of thinking in binary “will happen / won’t happen” terms, and they revise those odds as evidence comes in. Charlie Munger listed this among the core mental models worth carrying in a decision-maker’s “latticework,” alongside basic ideas from statistics such as expected value and base rates.
When to use it
- Forecasting outcomes with incomplete information (business bets, investments, planning)
- Weighing risks where the cost of being wrong varies by outcome
- Evaluating claims or predictions made with false certainty
How to apply it
- List the plausible outcomes, not just the one you expect.
- Assign a rough likelihood to each, using base rates where available.
- Weigh outcomes by both probability and impact (expected value), not just probability alone.
- Update your estimates as new evidence arrives instead of anchoring on the first guess.
Watch out for
- False precision — a rough probability is still more honest than false certainty, but don’t over-trust a specific number.
- Ignoring base rates in favor of a compelling story (a common source of misjudgment).
- Ignoring how a single bad outcome, even if unlikely, can be catastrophic (fat-tail risk).
Related models
- Decision Making on Short, Medium and Long Term — probabilistic thinking feeds into weighing choices across time horizons.
- First Principles — another core reasoning tool from the same mental-models toolkit.
- Regret Minimization — a complementary way to decide under uncertainty.
Sources
Charlie Munger’s talks on a “latticework of mental models,” collected in Poor Charlie’s Almanack.