Game Theory & Decision
Intermediate

Regret

Difference between the best possible payoff and the one you got.

Formula

R=u∗−uR = u^* - u

Live Calculator

Members

Variables

RRegret
u*Best possible payoff
uActual payoff

Example

Minimax-regret decision rules minimise the worst possible regret.

How it works

Regret measures how much you missed out on by not choosing the best option for the situation that actually occurred. It is found by subtracting the payoff you received from the highest payoff you could have received, so a regret of 0 means you made the best possible choice in hindsight. This idea comes from comparing each decision against a perfect benchmark, which helps decision-makers judge how costly an error really was.

When to use it

Use regret when comparing choices under uncertainty, especially if you want to measure the opportunity cost of a decision after the outcome is known. It is common in game theory, economics, operations research, and decision analysis.

Worked example, step by step

A company had three bidding strategies for a contract. If demand turns out high, the best payoff would have been $120,000 and the company’s chosen strategy gives $95,000. Regret = u* - u = 120,000 - 95,000 = 25,000. So the regret is $25,000. This means the company earned $25,000 less than the best possible outcome for that demand scenario.

Common mistakes to avoid

  • Subtracting in the wrong order and getting a negative regret
  • Using the best payoff from all scenarios instead of the best payoff for the actual outcome
  • Confusing regret with total loss rather than the gap from the best available payoff

Did You Know?

Regret minimisation drives some of the strongest poker-playing AIs ever built.

Share this formula

Reviewed & maintained by the FormulaWon editorial teamLast reviewed September 2026