Expected Value in Poker: How to Calculate EV

Expected value is what a decision is worth on average if you could make it a thousand times over. Multiply each outcome by how often it happens, add the results together, and you have the number. A play is correct when that number beats every alternative, whatever the river does to you this particular time.

What EV actually measures

Every poker decision has a price and a set of possible outcomes. EV collapses all of them into a single figure in dollars or big blinds.

You call $50 to win a $150 pot. Sometimes you take the $150, sometimes you lose the $50. EV asks what the weighted average of those two results comes to. If it is positive, that call makes you money across your career even though most individual attempts feel like a disaster.

That is the shift EV forces on you. You stop grading decisions by whether they worked and start grading them by whether they were worth making. The board runs out how it runs out and you have no say in it. The only thing you control is whether you chose the line with the highest number attached.

Folding is part of that comparison too. A fold is always worth exactly zero: you put in nothing more and you win nothing more. So the real question at every decision point is not whether a call is good, but whether it beats zero — and whether raising beats it.

The formula

EV = (how often you win x what you win) − (how often you lose x what you lose)

Three inputs, and two of them you can read straight off the table.

• What you win is the pot as it stands plus any bet you are calling.
• What you lose is only the amount you are about to put in. Money already committed is not yours and does not belong in the calculation.
• How often you win is the sole estimate, and it is where all the difficulty lives.

That last term is your equity. The comparison of equity versus pot odds is simply this formula set to zero: pot odds hand you the win rate at which EV breaks even, and everything above that line is profit.

Worked example: the river call

The river bricks. The pot is $100 and your opponent bets $50. You hold a bluff-catcher — a hand that beats every bluff in his range and loses to every value bet.

You are calling $50 to win $150, so your break-even point is $50 ÷ $200, or 25%.

Now put a number on how often he is bluffing. Say the line he took makes you think it is about a third of the time.

EV = 0.33 x $150 − 0.67 x $50 = $49.50 − $33.50 = +$16

Sixteen dollars, every single time this spot appears. You will lose the hand twice in three attempts and it is still one of the more profitable buttons you press all session.

Your win rate	EV of calling $50 into $100
15%	−$20
20%	−$10
25%	$0
33%	+$16
40%	+$30
50%	+$50

Two things fall out of that table. Being slightly wrong is cheap: misreading his bluffing frequency by five points shifts the answer by about $10. Being reflexively tight is not cheap at all, because folding surrenders that $16 in every spot where you were actually ahead — and against anyone paying attention, the number gets worse, since he simply starts bluffing every river.

Every bet has two ways to win

Calling has one route to profit: showing down the best hand. Betting has two, and that is why aggression prints money.

Flop. The pot is $100, you hold a flush draw, and you bet $60. With two cards to come your draw gets there roughly 35% of the time by the river.

Split the outcomes:

• He folds. Say that happens 40% of the time. You win the $100 on the spot.
• He calls. The other 60%. The pot becomes $220, so you win $160 with your 35% and lose $60 with the other 65%: 0.35 x $160 − 0.65 x $60 = +$17.

EV = (0.40 x $100) + (0.60 x $17) = +$50

Forty of those fifty dollars arrive from hands that fold before a card is dealt. That is the fold equity you gain by betting instead of checking, and it is the biggest single reason a draw plays better as a bet than as a call. Check it back and only the 35% remains.

Ranking lines, not grading one call

EV is not a pass or fail test applied to a single action. It is a ranking of every action available to you.

Folding is the zero. Calling might be worth +$16. Raising might be worth +$40. All three are correct in the weak sense that none of them sets fire to money, and two of them are still mistakes.

• A fold banks nothing and risks nothing. It is the baseline every other line has to beat.
• A call freezes the pot where it is. No leverage on later streets, and you win only by having the best hand.
• A raise stacks folds on top of showdown wins. More variance, usually more EV, occasionally far less when only better hands continue.

Build one habit out of this: after any non-obvious decision, name the second-best line and say why it lost. If you cannot name it, you did not choose — you defaulted.

This is also where most preflop money leaks away. Flatting a raise on the BTN with a hand that plays fine postflop is rarely a losing action on its own, but it can still be several big blinds worse per hundred hands than three-betting the same holding, because the flat gives up every pot the three-bet would have taken down uncontested. Two lines can both be profitable while one of them quietly costs you a stack a week.

Why a +EV call still loses the pot

A 35% draw misses 65% of the time. That +$16 river call loses two times in three. AA gets in against KK as roughly an 82/18 favourite and still goes down once in six, and AK against QQ is a slight dog at about 43/57 — a race you lose more often than you win and should take every time the money says so.

None of that makes those decisions wrong. It makes the results noisy.

The gap between your EV and your actual bankroll is variance, and over any stretch shorter than tens of thousands of hands it runs wider than most players expect. A genuine winner can sit below break-even for a month without anything being broken. Getting comfortable with standard deviation in poker is what stops you from rebuilding a sound strategy around one bad week.

The defence is to grade the decision, not the outcome. Review the hand at the moment you acted, using only the information you had then. If the number was positive and larger than the alternatives, the hand was played correctly and the river owes you nothing.

Making the estimate fast at the table

You are not running three-decimal arithmetic in thirty seconds. You need the sign of the number, not its exact value.

1. Price it first. Bet divided by the final pot. Half pot needs 25%, two-thirds needs 28.6%, a pot-sized bet needs 33%.
2. Estimate equity with the rule of 2 and 4. Outs times two for one card, times four for two. Nine flush outs is about 18% on the next card and roughly 35% by the river.
3. Add fold equity only when you are the aggressor. Calling never generates any.
4. Compare against folding and against raising, pick the biggest number, and move on.

Away from the table, do it properly. Take three hands from your last session, guess the EV before you work it out, then run the real ranges through a poker equity calculator and see how far off you were. That gap between your instinct and the truth is the actual leak, and it closes quickly once you start measuring it.

Most of the most common beginner mistakes are EV errors in disguise: folding a hand that was being offered a price, calling a river against a range containing no bluffs, checking back a draw that wanted to bet. Fix the arithmetic underneath and a whole category of them disappears at once.