Probabilistic Thinking
Think in ranges of outcomes with probabilities, not certainties.
Key Principle
Assign probabilities to outcomes and update them as evidence changes.
Understanding Probabilistic Thinking
Probabilistic thinking means abandoning the binary "it will work" or "it won't work" and instead thinking in terms of likelihoods: "There's a 60% chance of outcome A, 30% chance of outcome B, and 10% chance of outcome C."
This matters because the world is uncertain, and decisions made as if outcomes are certain tend to be poorly calibrated. A 60% probability is very different from 90%, but binary thinking treats both as "likely."
The goal isn't perfect prediction—it's better calibration. Over time, your 70% predictions should come true about 70% of the time. If your 70% predictions come true 90% of the time, you're underconfident. If they come true 50% of the time, you're overconfident.
Real-World Examples
- Instead of "This startup will succeed," think "40% chance of significant success, 35% chance of modest outcome, 25% chance of failure."
- Instead of "The market will go up," think "55% chance of positive returns this year, 30% chance of flat, 15% chance of significant decline."
- Instead of "This candidate will be great," think "60% chance of strong performer, 30% chance of adequate, 10% chance of bad hire."
- Instead of "This feature will work," think "70% chance of positive reception, 20% chance neutral, 10% chance of backlash."
How to Apply This
Write down your probability estimates before learning outcomes to calibrate over time
When new information arrives, explicitly update your probabilities (Bayesian updating)
Make decisions based on expected value: probability × outcome value
Track your predictions to identify areas where you're systematically over or underconfident
Common Mistakes to Avoid
- Treating 60% and 90% as equivalent ("both are likely")
- Not updating probabilities when new evidence emerges
- Overconfidence in precise probability estimates
- Ignoring low-probability, high-impact outcomes
Notable Quotes
"The future is not a single path but a range of possibilities."
— Philip Tetlock
"Being wrong is not a crime. Failing to update when you're wrong is."
— Julia Galef