Core Distinction

The Rules Are in Play

Game theory can model a game with no ending. It has no way to model a game where the rules change, because a fixed payoff structure is what makes the mathematics work at all. That is the seam between an infinitely repeated game and the Infinite Game. One runs forever on a frozen board. The other rewrites the board as a condition of continuing.

The distinction most long-game advice skips, and the reason a business can play forever and still be playing a finite game.

Updated August 2026

Two things called infinite

Game theory has a precise object called an infinitely repeated game. The same players, the same available moves, the same payoffs, repeated with no known final round.

Carse's infinite game is a different object. Players enter and leave. The rules change during play. What counts as a move is itself in play, and so is who the player is.

Both get called infinite. Only one of them is.

The first has an open horizon on a frozen board. The second changes the board as a condition of continuing. Collapsing them is the most common error in business writing on this subject, and it stays invisible because the vocabulary is identical on both sides.

Where the mathematics stops

Formal game theory needs fixed players, a fixed set of available actions and a payoff structure that is known or learnable. Those assumptions are what make a solution computable in the first place.

Change the rules mid-play and there is no longer a game to solve. There is a sequence of different games, with no principled way to say the same one continued.

Xabier Barandiaran makes this argument directly. Classical game theory and evolutionary game theory are both structurally aligned with finite games, because both assume static agents, fixed payoffs and closed boundaries. An infinite game theory would be the conjugate of the existing one rather than an extension of it.

None of this is a complaint about the tool. It is a description of what the tool is for. A hammer holds no opinion about screws.

The test that separates them

Ask what would have to change for you to still be playing.

If the answer is nothing, the horizon is open and the game is finite. You are running the same play forever, which is a fine thing to do and a different thing from this.

If the answer includes the rules, the roles, the definition of a good move or who you are while making it, you are standing somewhere the mathematics does not reach.

Most businesses describing themselves as playing the long game are describing the first one. The long game and the Infinite Game are not synonyms, and the gap between them is where nearly all of the interesting work lives.

What changes when the rules are in play

Strategy stops being the top layer. If the rules can change, then choosing which rules to play under is itself a move, and a larger one than any move available inside them.

Identity becomes a variable rather than a constant. A finite player protects who they are, because the identity is what the winning accrues to. An infinite player lets the play change them, because being changed is one of the ways the game continues.

Winning loses its meaning rather than its appeal. There is no arrangement of the board that ends it, so the question shifts from how to arrive to what is worth continuing.

This is the structural reason for an operating system rather than a strategy. A strategy assumes the rules hold. An operating system is what runs while they change.

Running it

Name the rules you are currently playing under, out loud and in specifics. Most of them were inherited rather than chosen, which is exactly what makes them feel like facts about the world.

Sort that list into the ones you picked and the ones that arrived with the room.

Then change one on purpose and watch what happens. Pick a small one. The point is to feel the difference between a move inside the rules and a move on them, and a small change teaches that as well as a large one.

The distinction stays useful for years afterward, because it turns a whole class of stuck situations into a different question. Not how do I win this, but who decided this was the game.

Frequently Asked Questions

Is an infinitely repeated game the same as an infinite game?

No, and the difference is the whole point. An infinitely repeated game keeps the same players, the same available moves and the same payoffs, with no known final round. Only the horizon is open. Carse's infinite game changes the rules, the roles and the identity of the player as a condition of continuing. One runs forever on a frozen board. The other rewrites the board.

What stops game theory from modeling changing rules?

A fixed payoff structure is what makes a game solvable. Formal game theory needs fixed players, a fixed action set and payoffs that are known or learnable, because those assumptions are what a solution is computed against. Once the rules change mid-play there is no single game left to solve, only a sequence of different games with no principled way to say the same one continued.

Is playing the long game the same as playing the Infinite Game?

Usually not. Playing the long game normally means running the same play on a longer horizon, which is a finite game with the ending pushed out of view. The Infinite Game puts the rules themselves in play. A useful test: ask what would have to change for you to still be playing. If the answer is nothing, the horizon is long and the game is finite.

Does this mean game theory is wrong?

Game theory is exact about what it models. It is a mathematics of winning, and it is the most rigorous one ever built. The point of the distinction is scope rather than error. Its results hold inside fixed rules, and a life or a body of work is not played entirely inside fixed rules.

What does it look like to change the rules on purpose?

It looks like changing what counts as a good move rather than making a better one. Deciding that a piece of work is measured by whether it is still useful in ten years rather than by how it performed in its first week is a rule change. So is deciding that a collaborator is a long-term relationship rather than a transaction. The move is small and the game afterward is different.

Related concepts

Core Distinction

The Shadow of the Future

A named result from game theory. Cooperation becomes rational the moment the final round disappears. The reverse is the sharper half. When a player can see the ending, taking on the last round makes taking rational on the one before it, and the logic runs backward until it reaches the first move. Most working relationships are damaged by a visible finish line rather than by bad character.

Core Distinction

Puzzles, Not Problems

A problem implies something broke and wants fixing. A puzzle implies something was placed, is waiting and is solvable, and the solving is the point. Same mechanism, entirely different posture. Stated precisely: a puzzle is a finite game nested inside the Infinite Game. Bounded, winnable, played to completion.

Philosophy

Joyful Sovereignty

An approach to playing the Infinite Game through joy, sovereignty and embodied play rather than strategy and corporate optimization. When your choices come from genuine alignment rather than conditioned obligation, there is a felt quality: power without performance, aliveness without effort. The philosophy. The compass state.

Archetype

The Pioneer

Someone who has already won the game most people are still trying to win. Accomplished, well-resourced, free by every external measure. Yet something does not match. Their outer life was built on conditioning they never chose. They want their decisions, work and relationships to finally match their inner knowing. Not seeking success. Seeking coherence.