Course contentsShow
← Data Science
Lesson 152 of 2,1454. Discrete Probability DistributionsFree lesson

Decision Making Under Uncertainty

Use expected value to evaluate decisions, assess risk, and compare alternatives in business scenarios.

Decision Making Under Uncertainty

What you'll learn: How to use expected value to make smarter choices when outcomes are uncertain and evaluate which option is truly best.

The Core Idea

Life rarely offers guarantees. Should you launch a new product? Accept a risky contract? Choose between marketing strategies? When you can't predict the future with certainty, expected value becomes your decision-making compass.

Expected value tells you the average outcome you'd get if you could repeat a decision many times. By calculating the expected value for each option, you can compare apples-to-apples even when the actual outcomes differ wildly.

How It Works

For each possible decision:

  1. List all possible outcomes and their probabilities
  2. Calculate the expected value: E(X) = Σ [outcome × probability]
  3. Compare expected values across alternatives
  4. Consider variance to assess risk

Real Business Example

Imagine choosing between two marketing campaigns:

Campaign A: 70% chance of 50,000profit,3050,000 profit, 30% chance of 0
E(A) = 0.7(50,000) + 0.3(0) = $35,000

Campaign B: 40% chance of 100,000profit,60100,000 profit, 60% chance of -10,000 loss
E(B) = 0.4(100,000) + 0.6(-10,000) = $34,000

Campaign A has a higher expected value and lower risk (no losses possible). But if Campaign B had E(B) = $40,000, you'd face a trade-off: higher expected return but greater variance (risk of loss).

Beyond Just Expected Value

Smart decision-makers also consider:

  • Risk tolerance: High variance might be unacceptable even with better expected value
  • Worst-case scenarios: Can you survive the potential losses?
  • Opportunity cost: What are you giving up by choosing one option?

Key Takeaway: Expected value transforms uncertain choices into comparable numbers, letting you rationally evaluate alternatives—but always consider both the average outcome and the spread of risk.