Sure-thing principle
Also known as: Savage's sure-thing principle
If you'd choose the same thing whether or not an event occurs, knowing it shouldn't change your choice.
What it means
The sure-thing principle is a foundational axiom of rational choice under uncertainty: if a person would prefer option A to option B both when some event happens and when it does not, then they should prefer A to B even when they do not know whether the event will happen. It expresses a kind of dominance and is closely tied to the independence axiom that underpins expected-utility theory, formalizing the idea that a state of the world which yields the same outcome under both options should be irrelevant to the choice between them. Its descriptive failures are landmark findings: the Allais paradox shows people violate it by treating a common consequence inconsistently when it is bundled with certainty, and the disjunction effect shows people defer a decision until an uncertain event is resolved even though they would make the same choice either way. These violations helped motivate prospect theory and models of ambiguity. It matters because the principle marks a boundary between normative rationality and how people actually reason under uncertainty.
Examples
A student says she would take a Hawaii trip whether she passed or failed an exam, yet chooses to wait and pay to learn the result before deciding — violating the principle, since the outcome doesn't change her preference.
A manager says she would hire the candidate whether or not the budget expands, yet holds the offer until the budget meeting — paying for news that cannot change her decision.
Drivers will pay heavily to cut a small risk to zero but shrug at halving a larger one, though the reduction is the same size. Certainty gets a premium the axiom forbids.
First described in Leonard Savage (1954); disjunction-effect tests by Tversky & Shafir (1992).