← /notes

Expected Value

Created Jun 6, 2026 pokerprobabilitydecision-makingriskgames

Expected value is the average result of a decision if you could make it many times: each possible outcome weighted by its probability, summed. A bet that wins $30 a quarter of the time and loses $10 the rest has an expected value of a few dollars in your favor, even though most single trials lose. The whole discipline of playing well under uncertainty reduces to one habit: choose the option with the highest expected value, again and again, and let the averages do their work over a long enough run.

Poker is the cleanest school for this because it forces the math into every hand. Suppose the pot holds $30 and it costs you $10 to call. You’re being offered “pot odds” of 3 to 1 — you risk one to win three. If your chance of completing your hand is better than one in four, calling is +EV; if it’s worse, calling is −EV, no matter how the single hand turns out. David Sklansky’s The Theory of Poker turned this kind of reasoning into a discipline, and his Fundamental Theorem follows from it: you play correctly when you play as you would if you could see your opponent’s cards — that is, when you make the highest-EV decision given what’s actually knowable.

The hardest part isn’t the arithmetic. It’s separating the decision from the result. A +EV call that loses was still the right call; a −EV call that happens to win was still a mistake. In a world with luck in it, good decisions lose all the time and bad ones get rewarded, and the only way to stay sane — and to keep improving — is to judge the process, not the single outcome. Poker players have a term, “results-oriented thinking,” for the error of grading a decision by how it happened to turn out. It’s the most common mistake in any field where chance intervenes between choice and consequence.

This is the engine under a lot of the garden’s risk ideas. The Kelly criterion tells you how much to bet once you know your edge; expected value is what defines the edge in the first place. Ergodicity is the crucial asterisk — EV is the right guide only when you’ll survive to play the long run, which is why a positive average doesn’t save you from a bet that can wipe you out. And fat tails are where naive EV goes wrong, because a rare enormous outcome can dominate an average that “looks” fine in every ordinary trial.

The transferable habit is to think in distributions instead of single stories. Before a decision: what are the outcomes, how likely is each, what’s the weighted average — and can I afford the bad tail? After: did I reason well, regardless of how the dice fell? Most people do the opposite. They tell a single confident story about what will happen, then grade themselves by whether that one story came true. Expected value is the antidote to both halves of that mistake.

Go Deeper

Books

  • The Theory of Poker by David Sklansky — Pot odds, expected value, and the Fundamental Theorem; the book that made poker quantitative.
  • The Mathematics of Poker by Bill Chen and Jerrod Ankenman — A rigorous, game-theoretic treatment for the deep end.