Subjective expected utility
Rational choice under uncertainty using your own probabilities and values.
What it means
The normative theory that a rational agent facing uncertainty chooses the option maximizing the sum, over possible outcomes, of each outcome's personal utility weighted by the agent's own subjective probability of it occurring. It generalizes expected utility by replacing objective, known probabilities with degrees of belief, and it was given axiomatic foundations showing that anyone whose preferences satisfy a few consistency conditions acts as if maximizing SEU with some coherent probabilities and utilities. The framework is the bedrock of decision analysis and Bayesian rationality, but it is descriptively violated in systematic ways — the Allais and Ellsberg paradoxes show people's choices cannot be captured by any single coherent probability-and-utility pair. It matters as the benchmark against which behavioral departures are measured, and as the engine of prescriptive decision-making even where human intuition strays from it.
Examples
Deciding whether to carry an umbrella by weighing your personal sense of how likely rain is against how much you'd dislike getting wet versus the bother of carrying it.
Weighing a job offer, you multiply your own hunch about whether the startup survives by how much the upside would mean to you, against how badly a collapse would hurt.
A patient deciding on elective surgery weighs their personal sense of how likely the complication is, and how much they would hate it, against the relief they expect to gain.
First described in Leonard J. Savage (1954); building on Ramsey and de Finetti.