EV Is Not Math, It's How You See the Future

Most players have heard the phrase "expected value" a thousand times. They have nodded at it as if they understood. But the first thing I want to say about expected value — before we touch a number or a formula — is that it is a way of relating to the future. That is what it is underneath the math. Get that part wrong and the equations never mean anything. Get it right, and the math becomes the most honest language you have ever spoken at the table.

The future is a fan, not a fact

Every decision you make at the table will, depending on which cards come and which actions your opponent takes, give you many possible futures. Each one has its own probability and its own outcome. You do not know which future you will get. You cannot know — the cards have not been turned, your opponent has not yet moved. The future is a fan of possibilities, not a single fact.

Expected value is very simply the weighted average of all the possible futures that flow from a given action, weighted by the probability of each. It is the average of the fan. If you took this exact decision a thousand times against this exact set of opponent strategies, with random cards each time, the expected value is the average outcome you would receive across all those thousand attempts.

That is what the number means. It is not a prediction of what will happen this time. It is what would happen on average if this exact moment played out a thousand times with the universe shuffling cards differently each time, everything else held fixed. Expected value is the language we use to talk about the long average of a moment, even though the moment itself only happens once.

What changes when you look this way

There is something strange and beautiful here that most players never sit with. The moment only happens once. You will only see this turn card once. You will only face this river decision tonight, in this hand, with these stakes, with this opponent, with this exact prior history. And yet the right action is determined, in some deep sense, by what would happen on average across many imagined repetitions of this moment that will never actually happen.

The math is talking about a thousand parallel universes of this exact moment. Your life is only going to traverse one of them. The action that is best on average is not necessarily the action that gives you the best outcome in this particular universe. It might. It might not. But over many such moments stacked together across your whole career, the action that is best on average is the action that wins. The decisions summed over time become destiny.

A river call, on the ground

Let me put this on the ground so the abstraction has a body. You are heads up on the river. The pot is 100 big blinds. Your opponent has just bet 60 into that pot of 100, so you are facing a call of 60 to win the existing 160 in the pot. The question is whether to call.

To answer with expected value, you estimate two things. First, the probability your hand is best when she has bet — your equity against her betting range. That is different from your equity against her checking range, because she chose to bet, and that choice tells you something about the kinds of hands she might be betting with. Suppose, looking at her betting range carefully, you think your hand beats her bets about 35% of the time. Second, the size of the prize and the size of the risk. The prize, if you call and win, is the pot plus her bet — 160. The risk, if you call and lose, is 60.

So the expected value of calling, in big blinds, is 0.35 × 160 minus 0.65 × 60. That is 56 minus 39, which is +17 big blinds. The expected value of folding is zero, because folding gives up your claim to a pot you have already contributed to but costs you nothing more from this moment forward. 17 is bigger than zero, so calling is the higher-expected-value play, and that is the right action.

The question EV replaces

Now sit with what just happened, because it is more important than the number itself. We took a decision you are about to make in real time, with real money on the line, and we replaced the question what should I do? with the question what is the average outcome of each available choice across all the ways this moment could go?

We did not ask whether your specific hand is good or bad. We did not ask whether you are feeling confident or scared. We did not ask whether she is the kind of player you like or do not like. We asked: given everything we believe about her range, what would happen on average if we called many times in this exact situation, and what would happen on average if we folded — and which average is higher? That is expected value reasoning. It is cold. It is in a sense indifferent to this specific moment. And it is the only reasoning that, over enough moments, actually wins.

A currency for comparing options

What expected value gives you, even as an approximation, is something nothing else gives you: a way of comparing different actions on a common scale. Without it, when you face that river decision, you have only feelings — intuitions, vague senses of which action seems right. Those feelings might be calibrated by years of play, but they have no common currency, no way of being checked against each other or against reality.

Expected value gives you a currency. It says this action averages +17 big blinds, that one averages zero, this other one averages −8. Now you can compare. You can say this one is better, and the comparison is grounded in a quantity that, even if approximate, has a direct connection to your bottom line.

A useful fiction — lean on it, don't worship it

One piece of caution underneath all this. The numbers I gave you — 35% equity, 60 into 100 — were assumed for the example. In a real spot at the felt, you do not know those numbers. You are estimating them. Your read on her range is a model, and the model may be wrong. The whole computation you did in your head in a few seconds is built on guesses stacked on guesses, and the final +17 is only as good as the guesses that went into it.

So expected value as a piece of math is exact. Expected value as a thing you can compute in real time is always an approximation — a sketch, a guess. The discipline is not in computing the math precisely. The discipline is in being honest about the guesses, updating them as new information arrives, and not pretending to a precision you do not have. The number is a useful fiction. Lean on it. Just do not worship it.

This article is drawn from the audio lesson "How to View Poker Outside of a Single Universe."

Poker Outside of a Single Universe