Decision Models Reference

Expected Value and Decision Models

Expected value is the long-run average outcome of a random process. It helps compare games, choices, and risk-reward tradeoffs.

Solved Examples

Jump directly to a worked example by expected-value task.

Fact Table

IdeaMeaningWatch For
Random variableNumerical outcome assigned to each event.List every possible value and probability.
Expected valueProbability-weighted average.It need not be a possible one-time outcome.
Fair gameExpected net gain is $0$.Include the cost to play.
Decision modelCompares expected outcomes of alternatives.Expected value measures long-run average, not a guarantee.

Content Formulas

Expected Value
$$E(X)=\sum x\,P(X=x)$$
Net Gain
$$\text{net gain}=\text{payout}-\text{cost}$$
Decision Rule
$$\text{compare expected net values under the same conditions}$$

Classic Examples

Find Expected Value from a Distribution

A game pays $0$ with probability $\frac12$, $2$ with probability $\frac14$, and $6$ with probability $\frac14$. Find the expected payout.

Solution Steps
  1. Multiply each payout by its probability.
  2. Add the weighted payouts.
  3. State the long-run average payout.
$$0\left(\frac12\right)+2\left(\frac14\right)+6\left(\frac14\right)$$$$0+\frac12+\frac32=2$$$$\boxed{E(X)=2}$$

Test a Fair Game

A game costs $2$ to play and has expected payout $2$. Is it fair?

Solution Steps
  1. Find expected net gain by subtracting the cost.
  2. Compare the result to zero.
  3. State the fairness conclusion.
$$E(\text{net})=2-2$$$$E(\text{net})=0$$$$\boxed{\text{The game is fair in the long run.}}$$

Compare Two Decisions

Game A has expected net gain $1$. Game B has expected net gain $\frac12$. Which is better by expected value?

Solution Steps
  1. Write both expected net gains.
  2. Compare the values.
  3. State the higher expected-value choice.
$$E(A)=1,\qquad E(B)=\frac12$$$$1>\frac12$$$$\boxed{\text{Choose Game A by long-run expected value.}}$$

Model an Insurance Decision

A $100$ loss has probability $0.01$. What is the expected loss before insurance?

Solution Steps
  1. Assign the loss as a negative outcome and no loss as zero.
  2. Multiply the loss by its probability.
  3. State the expected loss.
$$E(X)=(-100)(0.01)+0(0.99)$$$$E(X)=-1$$$$\boxed{\text{Expected loss }=\$1\text{ per event.}}$$