Decision Theory: Expected Values, Trees and Uncertainty Criteria
Commerce2 min readNov 21, 2026

Decision Theory: Expected Values, Trees and Uncertainty Criteria

Decision Theory: Expected Values, Trees and Uncertainty Criteria
2 min read · 275 words

Quick Answer

Decision theory separates certainty, risk and uncertainty: expected-value analysis for risk, decision trees for sequential choices, and maximin/maximax criteria for uncertainty. Sensitivity analysis tests how robust the chosen option is when probabilities and payoffs move.

Visual Summary

  • Risk — probabilities known; expected value decides.
  • Uncertainty — probabilities unknown; criteria fill in.
  • Decision tree — maps sequential choices and chances.
  • Sensitivity — how the choice survives changing inputs.

Risk and Expected Values

Under risk (known probabilities), each option’s expected monetary value weights payoffs by probabilities; the rational choice maximises EMV — adjusted for risk attitude (risk-averse decision-makers prefer lower-variance options at equal EMV). Decision trees map sequential decisions (squares) and chance events (circles), folding back expected values from the tips to choose at each decision node.

Uncertainty Criteria

With unknown probabilities: maximax (optimist — best possible outcome), maximin (pessimist — guard the worst), equally likely (average all), and minimax regret (minimise the maximum opportunity lost — the regret table). Sensitivity analysis varies probabilities and payoffs to find where the choice flips — the robustness check every recommendation needs.

Exam Angle

NET passages compute EMVs, build regret tables, or ask which uncertainty criterion a decision-maker used. Practise in the UGC NET Commerce mock.

FAQ

What does minimax regret minimise?
The maximum opportunity loss — the regret of choosing wrong against each state.

Why fold back a decision tree?
To replace chance nodes with expected values, deciding sequentially from the tips inward.

Related reading

Exam angle

How exams use this: CA/Foundation papers mix recall, calculation and application in one sitting — train all three under +1/−0.25. Test it under exam conditions →

Quick revision

  • Risk: — probabilities known; expected value decides.
  • Uncertainty: — probabilities unknown; criteria fill in.
  • Decision tree: — maps sequential choices and chances.
  • Sensitivity: — how the choice survives changing inputs.
  • Operations Research: Linear Programming and Its Algorithms
  • Econometric Inference: Standard Errors and Assumption Breaches
ShareTelegramX

Have a doubt on this topic?