Every Serious Poker Player Is a Believer

What the coldest game on earth can teach rational people about faith

Ask a professional poker player what they study away from the table and you will almost always hear the same three letters: GTO. It stands for "game-theory optimal." In heads-up poker (one opponent, no one else at the table)¹ the GTO is a way of playing that no opponent can beat in the long run, no matter how good they are or how well they read you. It is the yardstick against which the profession grades its decisions.

Nobody has ever seen it.

Mathematicians proved almost a century ago that a game like this must have an unbeatable strategy.² But nobody can write it down. Take the version of the game AI researchers study: Texas hold'em (the form of poker most people have seen on television) played heads-up, with no cap on how much you may bet. The number of distinct decision points in it (situations in which a player must act) is roughly a 1 followed by 161 zeros. The number of atoms in the observable universe is a 1 followed by about 80 zeros.³

So the modern poker profession is organized around something no one has seen. Players spend years studying it. They grade their decisions by how far they stray from it. They stake real money on it.

And no one calls this superstition.

And believer is the right word, in a functional sense, the only sense that matters for the game. Nobody at the table cares what you say you believe; they watch what you do when the money is in. What a player believes is what he acts on.

Real money, staked on a standard you cannot fully work out, on the strength of a proof you have not read and calculations you could not repeat: that posture is much closer to what the old religious thinkers meant by faith than to the caricature of believing without reasons.

A Standard You Cannot Compute

Look at what kind of thing the GTO is.

It is objective. The rules of the game fix it. It does not care who you are, where you grew up, or what you would like to be true.
It is binding. Play anything else and some opponent can beat you, whether or not you have ever heard the term.
It is out of reach. No human has computed it, and no computer yet built can.

Even the players who deliberately depart from it confirm its authority. A professional who notices that an opponent gives up his hand more often than he should will bet more often than the GTO would. But "more often than he should" means more often than the GTO would. The departure is measured with the ruler it departs from.

Now compare what Plato, Augustine, and Aquinas said about goodness. On their account, goodness is not a feeling, and it is not a local custom. Some acts are wrong the way throwing away a royal flush is wrong: not because anyone disapproves, but because of how things actually are. The standard binds you whether you believe in it or not. And no human mind grasps it completely.

That last claim is not a mystical flourish, and poker shows why it needs none. The rules of hold'em fit on a single page. The situations those rules generate run to that 1 followed by 161 zeros, and the right way to act in every one of them is the thing no one can write down. The traditions describe the good the same way. The precepts are short enough for a child to learn, and Aquinas said plainly where the difficulty lives: the general principles of conduct are known to everyone; it is in the descent to particular cases that certainty gives out.⁴ A short rule and an uncomputable application are not a paradox.

Objective, binding, out of reach. That is also how the traditions describe the good. Whether such a standard exists is a fair question. But if it does, steering your life by it is not a confused way to think.

What Tradition Actually Is

Players do not calculate the GTO at the table; nobody can. Away from the table, they study the output of solvers (programs that approximate the GTO for particular situations) and boil it down to rules of thumb: how much to bet on which kinds of cards, when having to act first should make you more cautious, which hands to give up and which to fight for. They follow those rules even though they could not reproduce the math behind them. The rule is a portable version of an answer too large to work out on the spot.

A commandment does that part of the job for the believer: a rule you can follow right now without being able to show, right now, everything that stands behind it. But a list of rules is not most of what a tradition is, any more than a list of rules is most of what a player has. Most of what a serious player knows is written down nowhere. It is in his hands, put there by repetition. The solvers confirmed some of what the best players had been doing by feel for years, and those players often could not say why it worked; a rule is what you get when someone articulates a pattern that was already working. Serious players run the same cycle over and over: study a situation, play it, go back over the hands afterward. Stated, enacted, remembered—until the right play is there under pressure without being computed.

The traditions state their pattern in a text, enact it in a rite, and commemorate it, week after week, year after year, and they do so for at least the reason the player drills: so that what is known will be there when it is needed, in the body and not only in the head. And they also set aside time to "go back over the hands afterward": prayer. If that is what a tradition is, G. K. Chesterton's warning follows: the man who cannot see what a fence across a road is for is the last man who should be allowed to tear it down; let him first go and find out why it was put up, and then he may.⁵ The Christian traditions claim much more for their rites and prayers than this, but for this argument those claims are not required.

Whether a particular tradition's rules and rites actually track the good is a fair question. But it is a different question from whether it can be rational to follow practices you cannot personally derive. It can. Professionals do it for a living.

So the charge that religion is "rules without reasons" misses how rules work, and misses that rules are the smaller part.

Losing with Aces

Poker forces you to separate the quality of a decision from how it turns out. You can put all your money in before a single shared card is dealt, holding a pair of aces against a pair of kings, a spot you win about four times in five, and lose. Right decision, bad outcome. Judging a decision by its result has a name among players, results-oriented thinking, and it is a defining amateur mistake. Skill and results line up only over enormous numbers of hands. Nothing guarantees the next one. The player's whole discipline is to make the correct decision now, on behalf of a long run no single night will show him.

Every disciplined fold, every hand thrown away because the numbers say so, is a small sacrifice: a pot you might have won, given up for a future you cannot see. You are not committed to a strategy until you have thrown away a hand you wanted to play.

Now look at the most familiar objection to religion: good people suffer, so the religious picture must be false. In one form, that is results-oriented thinking. It argues from a bad outcome to a bad decision, or from a bad outcome to "there is no standard at all." The traditions spotted this error long before game theory existed. The Book of Job opens by calling Job blameless, then strips him of his wealth, his children, and his health. His friends look at the wreckage and conclude he must have sinned. They are reasoning straight from the outcome to the verdict. Job's wife counsels the opposite mistake, the one every player on a losing streak has heard from inside his own head: "Curse God, and die."⁶ The standard has failed you; quit it. Job refuses both. He will not pretend he played badly,⁷ and he will not quit: "Shall we receive good at the hand of God, and shall we not receive evil?"⁸ God has questions for Job too, but the verdict falls on the friends: they are the ones who spoke wrongly.⁹

That is the weak form of the objection, though. The strong form is different: good people suffer, so there is no God who is both able and willing to prevent it. The GTO never promised anyone anything. The God in that argument is supposed to care. The poker analogy does not answer that, and this essay does not claim it does.

Tilt

What finishes a good player is usually not the losing streak itself. It is what the losing streak does to him. The bad stretch arrives. Give the game enough hands and it arrives for everyone. The player goes on tilt, the players' word for coming apart under loss. Tilt at its mildest is irritation and impatience. Tilt at its deepest is the thing that matters here: the losses convince the player that the standard is fake or the game is rigged, and he abandons correct play at exactly the moment discipline matters most. In that form, tilt is loss experienced as refutation. Job's wife supplied the formula. The best-known species of tilt has a name of its own, chasing: playing bigger and worse to get back tonight what tonight took. Genesis has that one too. Esau comes in from the field faint with hunger and sells his birthright for a bowl of stew.¹⁰ It is a sacrifice run backwards: the future traded for the present.

Keeping your decisions intact through the worst stretch of your life is not a technique. The player on tilt usually knows exactly what the correct play is; he cannot bring himself to make it. The old books have a description of what withstands this. In the Genesis flood story, Noah "found favor in the eyes of the LORD";¹¹ the Latin Bible's word is gratia. He is not spared the flood; he is carried through it. Grace, in that story, does not stop the rain. It is what carries a man through water that drowns the rest. Any player who has kept his game intact through a brutal month knows that white-knuckling only goes so far; the month outlasts resolve. Ask him how he did it and he has no good answer. He only knows how little of it felt like his own doing. Whether or not you believe in a giver, that much is real. The traditions take him at his word and go further: whatever carried him was, at bottom, not simply his. It was given. Their word is grace.

Where the Analogy Ends

Poker's rules were made up by people. But once the rules are fixed, the math guarantees that an unbeatable strategy exists. No theorem does that for the good. Whether reason can get there some other way is an old question, and it stays open here.

The GTO is also impersonal. It makes you unbeatable; it does not know your name. Christianity claims more: that behind everything is a personal Creator. A game-theory analogy cannot carry that claim, and this essay does not try.

And a poker player can check his rules of thumb. He can run a solver. He can look at his results over a million hands, and the numbers do not care what he hoped. That check works because poker's score is beyond argument: the chips are the verdict, and every school of play accepts the count. What counts as winning a life is precisely what is being argued. So there is no such machine for the good, and this essay has not built one. It proves no God and crowns no tradition; whether a tradition's rules really track the good is exactly what is in dispute, and nothing here settles it.

What is left is narrower, and harder to dismiss. If you think there is no objective moral order at all, this argument does not reach you; that is a different debate. There is a serious old argument about whether that denial can be lived; this essay does not press it. But if you think some acts really are worse than others, that cruelty is wrong whether or not anyone approves, then you already accept a standard that is objective and binding; and anyone who has ever had to weigh honesty against kindness knows how little the plain rules decide by themselves. And then taking your bearings from a standard you cannot fully compute, following inherited rules and practices you cannot personally derive, giving up the present pot for a long run you will not see, refusing to judge decisions by how they happen to turn out, and holding your game together through the losing streak is not superstition. It is what thinking clearly looks like when the game is too big to solve. The coldest game on earth shows it every day.

Most poker is played at full tables, where the math is messier and the guarantee no longer holds; the profession uses the GTO as its reference point anyway. ↩
Strictly, there may be more than one, but all are equally unbeatable, so players speak of it in the singular. ↩
A simpler cousin with fixed bet sizes was solved to within a rounding error in 2015. The version with no cap on the bets was not: the strongest programs for it only approximate the GTO, and compute their approximations on simplified versions of it. The solved cousin is heads-up limit hold'em; see Michael Bowling, Neil Burch, Michael Johanson, and Oskari Tammelin, "Heads-up limit hold'em poker is solved," Science 347 (2015). ↩
Summa Theologiae I-II, q. 94, a. 4. ↩
The Thing (1929), "The Drift from Domesticity." ↩
Job 2:9. ↩
Job 27:5–6. ↩
Job 2:10. ↩
Job 42:7. ↩
Genesis 25:29–34. ↩
Genesis 6:8. ↩