Statistics for Everyday Life
Risk vs Uncertainty
A famous distinction (often credited to the economist Frank Knight) separates situations where you can put numbers on the odds from those where you genuinely cannot — and they call for different tools.
| Aspect | Risk | Uncertainty |
|---|---|---|
| The odds | Known or reasonably estimable | Unknown, and often unknowable |
| Example | A dice roll; an insurable house fire | A brand-new technology; an unprecedented event |
| How to handle it | Probability, expected value, insurance | Scenarios, robustness, margin of safety, capping the downside |
| Can you compute an average outcome? | Yes, meaningfully | Not reliably — the inputs are guesses |
When to use risk
Treat a decision as risk when you can assign trustworthy probabilities — then expected-value reasoning and insurance work well.
When to use uncertainty
Treat it as uncertainty when you cannot honestly quantify the odds — then focus on surviving the worst case rather than optimising an average.
Frequently asked questions
- What is the difference between risk and uncertainty?
- Under risk the probabilities are known or estimable, so you can reason with expected values. Under uncertainty they are not, so precise probability calculations give false confidence and robustness matters more.
- Is investing in the stock market risk or uncertainty?
- Both. Short-run volatility can be modelled statistically (risk), but rare, unprecedented events and an unknown future contain genuine uncertainty. That is why capping the downside matters as much as chasing the average return.
- How do you make decisions under uncertainty?
- Not by pretending to know the odds. Use scenarios, keep a margin of safety, avoid ruinous bets you cannot recover from, and prefer options that hold up across many possible futures rather than optimising for one.