Regret

Origin domain

Information Theory

Subdomain

decision theory → Information Theory

Also from

Psychology, Behavioral Economics

Core Idea

Regret is the value gap an agent registers between the outcome it actually obtained and the better outcome that a forgone alternative would have produced — a comparison made after the fact against a counterfactual reference rather than against prior expectations. The defining structure is retrospective and relative: the same realized outcome can be coded as a triumph or a regret depending solely on which unchosen path it is compared against. Anticipated regret then feeds backward into choice, so the pattern couples a backward-looking evaluation to a forward-looking decision rule.

How would you explain it like I'm…

Wishing You'd Picked the Other One

Imagine you pick chocolate ice cream, and then you see your friend got strawberry and it looks even better. That sinking feeling — wishing you'd picked the other one — is regret. It only happens once you find out what would have happened if you'd chosen differently.

The Gap from What Could Have Been

Regret is the gap between what you actually got and what you could have gotten if you had chosen differently. The same outcome can feel great or terrible depending on what you compare it to. Winning $10 feels good if the other choice paid nothing, but bad if the other choice paid $1,000. Regret is not just an emotion; it also shapes future choices, because people often pick the option they think they will regret the least, even if it is not the one with the biggest possible reward.

Regret

Regret is the value gap an agent notices between what actually happened and what a different choice would have produced, judged after the fact against the unchosen alternative rather than against earlier expectations. The same outcome can be a triumph or a disaster depending only on the benchmark used. Anticipated regret then feeds back into decisions: people often act to minimize the regret they expect to feel, not just to maximize expected reward. This couples a backward-looking evaluation with a forward-looking decision rule, which is why regret-minimization is a recognized strategy in economics, decision theory, and machine learning, alongside expected-utility maximization.

 

Regret is the value gap an agent registers between the outcome it actually obtained and the better outcome that a forgone alternative would have produced, a comparison made after the fact against a counterfactual reference (the best unchosen option) rather than against prior expectations. The defining structure is retrospective and relative: the same realized outcome can be coded as a triumph or a regret depending solely on which unchosen path it is compared against. A bet that returns ten percent is a success measured against the bank but a regret measured against the stock that doubled. Anticipated regret feeds backward into choice, so the pattern couples a backward-looking evaluation with a forward-looking decision rule (regret-minimization), formalized by Savage's 1951 minimax-regret criterion and by Loomes and Sugden's 1982 regret theory of choice under uncertainty. Regret thus names a closed loop: outcome realized, compared against the best forgone alternative, scored as a shortfall, and that score fed forward to bias the next decision.

Broad Use

• Decision theory: regret-minimization criteria (Savage's minimax regret) select acts that bound the worst-case shortfall against the best act in hindsight.
• Behavioral economics: anticipated regret explains why people buy insurance, hold losing stocks, or avoid switching defaults.
• Reinforcement learning / online learning: "regret" is the formal objective — cumulative gap between an algorithm's reward and that of the best fixed action in hindsight; bandit algorithms are rated by regret bounds.
• Medicine: clinicians weigh decision regret (acting vs. waiting) when a missed-window outcome is irreversible.
• Auction/market design: bidders shade bids to avoid the regret of the winner's curse or of leaving surplus on the table.

Clarity

Naming regret separates two evaluations that ordinary "disappointment" conflates: shortfall against what you expected versus shortfall against what you could have chosen. It lets practitioners say precisely that a good outcome can still be regretted (a winning bet placed too small) and that minimizing regret is a different objective from maximizing expected value.

Manages Complexity

Regret compresses a whole counterfactual tree into a single scalar — the distance from the best forgone branch — turning "what else might I have done" into a comparable quantity. This bounds otherwise open-ended counterfactual reasoning to the one comparison that matters for evaluation and learning.

Abstract Reasoning

Once regret is recognized, minimax-regret and no-regret reasoning become available: an agent can commit to decision rules whose hindsight shortfall is provably bounded even under adversarial conditions, without needing a prior over states. It also licenses inferences about behavior that expected-utility models miss (default stickiness, omission bias).

Knowledge Transfer

The online-learning notion of bounding cumulative regret transfers directly to robust policy design: minimize the worst-case gap to the best benchmark you could not have known in advance. Conversely, the psychological finding that anticipated regret distorts choice transfers to mechanism design, where formats are engineered to reduce bidders' or consumers' regret exposure.

Relationships to Other Primes

Parents (3) — more general patterns this builds on

• Regret presupposes Comparison — Regret presupposes Comparison: it is the value gap measured between the realized outcome and a counterfactual alternative.
• Regret presupposes Decision — Regret presupposes Decision: there must have been a prior commitment among alternatives before retrospective evaluation can register a gap.
• Regret presupposes Modal Reasoning — Regret presupposes modal reasoning because the value gap it registers is a comparison between the actual outcome and a counterfactual alternative.

Path to root: Regret → Comparison

Not to Be Confused With

• Regret is not risk_aversion because risk aversion is a prospective preference over uncertain magnitudes, whereas regret is a retrospective comparison against a forgone alternative.
• Regret is not loss_aversion because loss aversion codes outcomes against a status-quo reference point, while regret codes them against the best unchosen option.
• Regret is not counterfactual_reasoning (its parent) because counterfactual reasoning is the general operation of considering alternatives, whereas regret is the specific valuation of the chosen-versus-forgone gap that drives evaluation and learning.