Key Takeaways: Chapter 10 — Expected Value

Core Concept

Expected value (EV) is the probability-weighted average of all possible outcomes of a decision. It is the foundation of rational decision-making under uncertainty — not because it guarantees good outcomes, but because it produces a process that generates better average outcomes over time.

The Seven Essential Insights

1. EV Is a Process, Not a Prediction

Expected value describes what happens on average across many trials — it does not predict any single outcome. You can make a perfect EV-positive decision and lose; you can make a terrible EV-negative decision and win. Judging decisions by their outcomes (outcome bias) rather than their reasoning is one of the most pervasive and costly cognitive errors in human judgment.

The daily habit: After a significant decision resolves, ask not "did it work out?" but "was my reasoning process sound given the information I had?"

2. Variance Is as Important as Expected Value

Two bets can have identical expected values with vastly different variance — the spread of possible outcomes. High-variance bets are more dangerous when: - You have a limited number of opportunities (small sample size) - A catastrophic loss would remove you from the game entirely - You cannot survive the downside long enough for the law of large numbers to operate

The Kelly insight: Never bet your entire bankroll on even the best positive-EV opportunity. Optimal bet sizing balances growth rate against the risk of ruin.

3. Utility Is Not the Same as Dollars

A dollar gained or lost is not worth the same amount of utility to everyone in every situation. Because utility functions are typically concave (each additional dollar provides less additional satisfaction than the previous), the same dollar amount has different impact depending on your current position.

• $8,000 medical bill when you have $6,000 in savings = catastrophic utility loss
• $8,000 medical bill when you have $500,000 in savings = significant but manageable

The insurance corollary: Insurance has negative dollar EV but positive utility EV for most people, because the utility loss from catastrophic uninsured events vastly exceeds the utility loss from paying manageable premiums. This is rational — not a failure of math.

4. The Kelly Criterion: Optimal Bet Sizing

The Kelly Criterion (f* = (bp − q) / b) provides the optimal fraction of your bankroll to bet on any positive-EV opportunity. Key properties: - Kelly maximizes long-run wealth growth rate - Kelly treats ruin as infinitely bad (you can't play future rounds if bankrupt) - "Half Kelly" is often recommended in practice, because probability estimates are imprecise and conservative sizing protects against estimation error - Applies conceptually to non-monetary resources: job application effort, startup development time, creative energy

5. Sometimes Negative EV Is Rational; Sometimes Positive EV Is Not

Rational negative EV bets: - Insurance (when catastrophic downside has utility asymmetry) - Entertainment purchases (lottery tickets, if the entertainment value exceeds the dollar loss) - Option-value purchases (paying to enter a context or competition for the information, network, or optionality it provides)

Rational rejection of positive EV bets: - When variance threatens survival (the company-ending marketing bet) - When probability estimates are deeply uncertain (Pascal's Mugging territory) - When better opportunities for your resources exist (opportunity cost is real)

6. The Regret Minimization Framework as EV Heuristic

When probabilities cannot be reliably estimated and decisions are irreversible, the regret minimization framework asks: "At age 80, which choice will I regret more — trying and failing, or never trying at all?" This is a practical approximation of utility-weighted EV that captures the long-term psychological cost of inaction.

Best used when: - The decision is hard to reverse - Inaction has a large but hard-to-quantify opportunity cost - The downside of trying is bounded and survivable

7. EV Applies Far Beyond Money

Any decision with uncertain outcomes has an expected value structure: - Job decisions: Enumerate outcomes, assign probabilities, adjust for utility, compare - Creative projects: Low cost of failure + small chance of transformative success often = strongly positive EV - Social risks: Bounded downside (rejection) + asymmetric upside (connection, mentorship, opportunity) often = strongly positive EV - Networking outreach: 70% ignored + 25% helpful reply + 5% transformative = positive EV for 30 minutes invested

The systematic habit of reaching out, applying, and initiating is a compounding positive-EV strategy that produces what looks like "luck" from the outside.

Key Formulas

Expected Value:

EV = Σ (P(outcome) × V(outcome))
   = P₁×V₁ + P₂×V₂ + ... + Pₙ×Vₙ

Kelly Criterion:

f* = (bp - q) / b

where:
  f* = optimal fraction of bankroll to bet
  b  = net odds received (profit per dollar wagered on a win)
  p  = probability of winning
  q  = probability of losing = 1 - p

The Luck Connection

Expected value thinking is one of the most powerful luck-engineering tools available. People who consistently make positive-EV decisions — even when individual decisions don't work out — accumulate more lucky outcomes over time than those who chase outcomes or avoid all risk.

The perception that some people are "just lucky" often reflects a failure to see the long run of decisions that produced those outcomes. Behind most "lucky breaks" is a history of small positive-EV risks that no one else was watching when they didn't work out.

Common Mistakes to Avoid

Connections to Other Chapters

• Chapter 6 (Probability Intuition): Base rates and calibrated probability estimates are essential EV inputs
• Chapter 9 (Survivorship Bias): EV analysis corrects for the tendency to overlearn from visible winners
• Chapter 11 (Birthday Problem): The limits of probability intuition — why EV inputs need stress-testing
• Chapter 15 (Loss Aversion): The psychological mechanism that causes utility curves to be steep near losses
• Chapter 37 (Portfolio Thinking): Kelly Criterion applied to life decisions across multiple domains

The Luck Ledger

One thing gained: A framework for making decisions that wins more often than it loses over time — not by guaranteeing good outcomes, but by guaranteeing a good process. EV thinking is how rational people manufacture luck.

One thing still uncertain: Probabilities. The EV formula is only as good as the probability estimates you put into it. Building better calibration — the art of knowing how confident to be — is a skill that takes years to develop and is addressed further in Chapters 6, 11, and 27.