👁 Preview — Study, Practice and Revise are open; mock tests and the rest of the syllabus unlock on subscription. Unlock all · ₹4,999
← Back to Logical Reasoning
Study mode

Decision Analysis

Study Practice Revise 🔒 Test

Introduction to Decision Analysis

Decision Analysis is a systematic approach to making choices among several alternatives. In logical reasoning and competitive exams, it helps you evaluate options carefully to select the best possible outcome. Understanding decision analysis equips you with tools to handle uncertainty, weigh risks, and maximize benefits effectively.

Before diving deeper, let's clarify some key terms:

  • Decision: A choice made between two or more alternatives.
  • Alternatives: Different options or courses of action available.
  • Outcomes (States of Nature): Possible results that can occur after making a decision, often uncertain.
  • Payoff: The value (usually in monetary terms like INR) or benefit received from an outcome.

For example, imagine you want to invest INR 10,000. You have two alternatives: invest in a fixed deposit or buy stocks. Each option has different possible outcomes with varying returns. Decision analysis helps you choose the best investment by considering these outcomes and their likelihoods.

Decision Trees

A decision tree is a visual tool that maps out decisions and their possible consequences, including chance events and payoffs. It helps you organize complex problems into a clear, step-by-step structure.

Here's how to construct a decision tree:

  1. Start with a decision node: Represented by a square, this is where you choose among alternatives.
  2. Add chance nodes: Represented by circles, these show uncertain events with associated probabilities.
  3. List outcomes: At the end of each branch, write the payoff (e.g., profit or loss in INR).
  4. Calculate expected values: Work backward from the outcomes to find the expected payoff for each decision.

Expected value (EV) is the weighted average payoff, calculated as:

Expected Value (EV)

\[EV = \sum (P_i \times V_i)\]

Weighted average payoff considering all outcomes

\(P_i\) = Probability of outcome i
\(V_i\) = Payoff of outcome i

Let's look at a simple decision tree example:

graph TD    D[Decision: Choose Investment]    D --> FD[Fixed Deposit]    D --> ST[Stocks]    FD --> FD1[Return: INR 5000]    FD --> FD2[Return: INR 5000]    ST --> ST1[Market Up (0.6): INR 12000]    ST --> ST2[Market Down (0.4): INR 2000]

In this tree, choosing Fixed Deposit yields a fixed return of INR 5000. Choosing Stocks has two outcomes: market up with 60% chance (INR 12,000) and market down with 40% chance (INR 2,000). Calculating expected values helps decide the better option.

Payoff Tables

A payoff table summarizes the payoffs for each decision under different states of nature in a tabular form. It helps compare alternatives quickly.

Here is an example payoff table for investment decisions (payoffs in INR):

Decision Market Up (60%) Market Down (40%)
Fixed Deposit 5000 5000
Stocks 12000 2000
Mutual Fund 9000 3000

This table shows payoffs for three investment options under two market conditions. Using this, you can apply different decision criteria to select the best decision.

Decision Criteria

When probabilities are known or unknown, different criteria help select the best decision:

  • Maximax Criterion (Optimistic): Choose the decision with the maximum possible payoff. It assumes the best-case scenario will happen.
  • Maximin Criterion (Pessimistic): Choose the decision with the best worst-case payoff. It protects against the worst outcome.
  • Minimax Regret Criterion: Choose the decision that minimizes the maximum regret. Regret is the loss from not choosing the best decision in hindsight.

To understand regret, consider this formula:

Regret

\[Regret = \text{Maximum payoff in state} - \text{Payoff of chosen decision}\]

Loss incurred by not choosing the best decision in a state

Maximum payoff in state = Best payoff achievable in that state
Payoff of chosen decision = Payoff for the decision under consideration

Let's illustrate these criteria using the payoff table above:

Decision Max Payoff Min Payoff Maximax Value Maximin Value
Fixed Deposit 5000 5000 5000 5000
Stocks 12000 2000 12000 2000
Mutual Fund 9000 3000 9000 3000

Using Maximax, Stocks (max payoff 12,000) is preferred. Using Maximin, Fixed Deposit (min payoff 5,000) is safest. Minimax Regret requires calculating regrets first.

State Max Payoff Fixed Deposit Regret Stocks Regret Mutual Fund Regret
Market Up 12000 12000 - 5000 = 7000 12000 - 12000 = 0 12000 - 9000 = 3000
Market Down 5000 5000 - 5000 = 0 5000 - 2000 = 3000 5000 - 3000 = 2000

Maximum regret for each decision:

  • Fixed Deposit: max(7000, 0) = 7000
  • Stocks: max(0, 3000) = 3000
  • Mutual Fund: max(3000, 2000) = 3000

Minimax regret criterion selects Stocks or Mutual Fund (both with minimum max regret 3000).

Worked Examples

Example 1: Choosing an Investment Option Medium
You have INR 10,000 to invest. You can choose Fixed Deposit (FD) with a guaranteed return of INR 5,000 or Stocks with a 60% chance of earning INR 12,000 and 40% chance of earning INR 2,000. Which option should you choose based on expected value?

Step 1: Calculate expected value (EV) for Fixed Deposit.

Since FD has a guaranteed return, EV = 1 x 5000 = INR 5,000.

Step 2: Calculate EV for Stocks.

EV = (0.6 x 12,000) + (0.4 x 2,000) = 7,200 + 800 = INR 8,000.

Step 3: Compare EVs.

Stocks have a higher EV (8,000) than FD (5,000).

Answer: Choose Stocks for higher expected return.

Example 2: Maximin Criterion Application Easy
Using the payoff table below, select the safest decision using the maximin criterion.
DecisionState 1State 2
A30001000
B20002000
C4000500

Step 1: Identify the minimum payoff for each decision.

  • A: min(3000, 1000) = 1000
  • B: min(2000, 2000) = 2000
  • C: min(4000, 500) = 500

Step 2: Choose the decision with the maximum of these minimum payoffs.

Maximin value = max(1000, 2000, 500) = 2000.

Answer: Decision B is safest under maximin criterion.

Example 3: Minimax Regret Calculation Medium
Given the payoff table below, calculate regrets and select the decision minimizing maximum regret.
DecisionState 1State 2
X70003000
Y90001000
Z60004000

Step 1: Find maximum payoff in each state.

  • State 1 max: max(7000, 9000, 6000) = 9000
  • State 2 max: max(3000, 1000, 4000) = 4000

Step 2: Calculate regrets for each decision.

  • X: State 1 regret = 9000 - 7000 = 2000, State 2 regret = 4000 - 3000 = 1000
  • Y: State 1 regret = 9000 - 9000 = 0, State 2 regret = 4000 - 1000 = 3000
  • Z: State 1 regret = 9000 - 6000 = 3000, State 2 regret = 4000 - 4000 = 0

Step 3: Find maximum regret for each decision.

  • X: max(2000, 1000) = 2000
  • Y: max(0, 3000) = 3000
  • Z: max(3000, 0) = 3000

Step 4: Choose decision with minimum maximum regret.

Answer: Decision X minimizes maximum regret (2000).

Example 4: Expected Value with Multiple Outcomes Hard
A company is deciding whether to launch a new product. The payoffs (in INR lakhs) and probabilities for market conditions are:
  • Good Market (probability 0.5): Profit INR 50 lakhs
  • Average Market (probability 0.3): Profit INR 20 lakhs
  • Poor Market (probability 0.2): Loss INR 10 lakhs
Alternatively, the company can avoid risk and earn a fixed profit of INR 15 lakhs. Which option should it choose based on expected value?

Step 1: Calculate expected value for launching the product.

EV = (0.5 x 50) + (0.3 x 20) + (0.2 x -10) = 25 + 6 - 2 = INR 29 lakhs.

Step 2: Compare with fixed profit option.

Fixed profit = INR 15 lakhs.

Step 3: Choose option with higher EV.

Answer: Launching the product is better (EV = 29 lakhs) than fixed profit (15 lakhs).

Example 5: Real-Life Decision Analysis Problem Hard
A farmer must decide between two fertilizers for a 1-hectare field. Fertilizer A costs INR 2,000 and yields either 3,000 kg/ha (60% chance) or 2,000 kg/ha (40% chance). Fertilizer B costs INR 1,500 and yields 2,500 kg/ha (100% chance). The market price is INR 20 per kg. Which fertilizer should the farmer choose based on expected profit?

Step 1: Calculate expected yield for Fertilizer A.

EV yield = (0.6 x 3000) + (0.4 x 2000) = 1800 + 800 = 2600 kg.

Step 2: Calculate expected revenue for Fertilizer A.

Revenue = 2600 kg x INR 20 = INR 52,000.

Step 3: Calculate expected profit for Fertilizer A.

Profit = Revenue - Cost = 52,000 - 2,000 = INR 50,000.

Step 4: Calculate revenue and profit for Fertilizer B.

Revenue = 2500 kg x INR 20 = INR 50,000.

Profit = 50,000 - 1,500 = INR 48,500.

Step 5: Compare profits.

Answer: Fertilizer A yields higher expected profit (INR 50,000) and is the better choice.

Tips & Tricks

Tip: Always list all possible outcomes and their probabilities before starting calculations.

When to use: When constructing decision trees or payoff tables to avoid missing scenarios.

Tip: Use the maximin criterion when you want to be risk-averse and ensure the worst-case scenario is acceptable.

When to use: When the exam question emphasizes safety or conservative decision-making.

Tip: Calculate regrets by comparing payoffs across each state to quickly identify the minimax regret decision.

When to use: When the problem asks for minimizing potential losses or regrets.

Tip: Convert all units to metric and currency to INR before solving word problems to maintain consistency.

When to use: In real-life or applied problems involving measurements and costs.

Tip: Practice drawing decision trees neatly and clearly to avoid confusion during exams.

When to use: When solving multi-step decision problems under time constraints.

Common Mistakes to Avoid

❌ Ignoring probabilities and treating all outcomes as equally likely.
✓ Always include the given probabilities in expected value calculations.
Why: Students often overlook probability weights, leading to incorrect expected values.
❌ Confusing payoff values with regret values.
✓ Calculate regret by subtracting payoffs from the maximum payoff in each state, do not mix them up.
Why: Misunderstanding the concept of regret causes errors in minimax regret problems.
❌ Not considering all possible alternatives or states of nature.
✓ List and analyze every alternative and state given in the problem before deciding.
Why: Incomplete analysis leads to suboptimal or incorrect decisions.
❌ Mixing units or currencies in calculations.
✓ Convert all measurements to metric units and all monetary values to INR before calculations.
Why: Inconsistent units cause calculation errors and confusion.
❌ Rushing through decision tree construction leading to missing nodes or outcomes.
✓ Draw the tree step-by-step, verifying each branch and outcome carefully.
Why: Haste causes incomplete or incorrect decision trees, affecting final answers.
✨ AI exam tools — try them free (included in every plan)
Tip: select any text above to Explain / Example / Simplify it.
Curated videos per subtopic
Top YouTube explainers, AI-ranked for your exam and language. Unlocks with subscription.
Unlock

Try Practice next.

Progress tracking is paywalled — subscribe to mark subtopics as understood and save your streak.

Go to practice →
Ask a doubt
Decision Analysis · 10 free messages
Ask me anything about this subtopic. You have 10 free messages this session — chat history isn't saved in preview.