Expected Value in Poker: The Only Number Your Decisions Have
Expected value is the average result of a decision if you made it infinitely many times: EV = (win probability × amount won) − (lose probability × amount lost). Poker skill is choosing the highest-EV option every street while variance decides single hands. Players who judge decisions by results instead of EV learn the wrong lessons at full speed.
Poker players love talking about reads, heart, and running good. Underneath all of it, the game is one question asked over and over: of the actions available right now, which has the highest expected value? Everything else, hand reading, position, bet sizing, is input into that single calculation.
EV is also the concept that separates players who improve from players who just accumulate hands. Here is the working version, table-speed included.
The formula and a worked hand
Expected value is the average result of a decision repeated forever:
EV = (P(win) × amount won) − (P(lose) × amount lost)
Concrete spot: the pot is $200, your opponent is all-in, and it costs $50 to call with a flush draw you estimate at 30% equity.
| Option | Calculation | EV |
|---|---|---|
| Fold | Nothing risked, nothing gained | $0 |
| Call | (0.30 × $200) − (0.70 × $50) | +$25 |
The call loses seven times out of ten and is still worth $25 on average, because the pot pays four times the call price against odds of a bit over two to one. You will feel terrible 70% of the time making a play that prints money forever. That emotional gap is where most poker goes wrong.
Estimating the 30% is its own skill: count outs, apply the rule of 4 and 2, compare to pot odds. Our pot odds and equity guide drills that pipeline, and the poker equity calculator lets you check any matchup exactly, the fastest way to calibrate your in-game estimates.
Folding is zero, and zero is a high bar
The fold option always has EV of exactly zero, which makes it the benchmark: every call or raise must beat doing nothing. Framed that way, a leak has a precise definition, any repeated action with negative EV, and the most common leaks become visible arithmetic:
- Calling river bets “to see it” with hands that beat nothing: pure negative EV purchased for curiosity.
- Chasing draws at bad prices: 20% equity facing a price that needs 33% is lighting money, however pretty the draw, the discipline our starting hands guide starts building preflop.
- Bluffing frequencies that ignore fold equity: a bluff risking $100 to win $100 needs folds more than half the time; against a station, that number is fantasy.
- Playing the wrong game at all: table selection is an EV decision made before the first card, and usually the biggest one of the night.
EV shapes bets, not just calls
The same arithmetic prices your aggression. A half-pot bluff risks 0.5 units to win 1, so it profits whenever opponents fold more than a third of the time; a pot-sized bluff needs folds half the time. That single relationship explains most professional sizing behavior: small bets need little fold equity and work as cheap information, large bets are priced for hands or opponents that fold too much. Value betting inverts it, you want calls, so the question becomes how much a worse hand will pay, not how often a better one folds. Every sizing decision at the table is one of these two EV equations wearing different clothes, and players who can name which one they are solving stop guessing at bet sizes entirely.
Results are noise; EV is signal
Over a session, even a big winner’s edge is a whisper against variance: correct play loses sessions constantly, and terrible play wins them often enough to teach terrible lessons. The practical consequences are mindset rules that sound like philosophy but are just statistics. Judge decisions, not outcomes. Review hands by asking what the EV comparison was, not whether the river was kind. And size your bankroll, conventionally 20 to 40 buy-ins for cash play, so variance cannot fire a winning strategy before the long run shows up; the same survive-to-compound logic behind our money hub and the compound calculator: edges only pay those who last long enough to collect.
The professional definition of running good is not winning flips. It is making the highest-EV choice so habitually that the long run has no choice but to pay you, and being bankrolled well enough to still be seated when it does. That skill is buildable, one priced decision at a time, starting from the fundamentals on the poker hub.
Frequently asked questions
What is expected value in plain terms?
The long-run average of a decision. Calling $50 to win a $200 pot with 30% equity earns on average (0.30 × 200) − (0.70 × 50) = +$25 per call, even though most individual calls lose. EV prices the decision, not the outcome.
How do I estimate EV fast at the table?
Count outs, convert to equity with the rule of 4 and 2, then compare against pot odds: the call price divided by the total pot after calling. Equity above the price is positive EV. The full pot-odds walkthrough is in our dedicated guide.
Why do correct decisions still lose so often?
Because positive EV usually comes with win probabilities well below 100%. A 30% equity call at great pot odds loses seven times in ten by design. The profit lives in the price, not the frequency, which is why results over small samples say almost nothing about decision quality.
What is the EV of folding?
Zero, always: you invest nothing further and win nothing further. That makes folding the benchmark every other option must beat. Calls and raises with negative EV are literally worse than doing nothing, which is the cleanest definition of a leak.
Does EV apply beyond calling decisions?
Everywhere: bet sizing, bluff frequencies, tournament spots, even game selection. A bluff needs to work often enough that fold equity covers the times it is called. Choosing a softer table is the highest-EV decision most players never think to make.
How is EV different from equity?
Equity is your share of the pot if hands ran to showdown now; EV prices a specific action, folding, calling or raising, given equity, pot size and future betting. Equity is an input; EV is the answer.
What bankroll does EV-based play require?
Enough buy-ins that variance cannot bankrupt a winning strategy before the long run arrives, conventionally 20 to 40 buy-ins for cash games, more for tournaments. EV math tells you the play is profitable; bankroll depth is what lets you survive to collect.