Web3 Gaming and the Metaverse: Does the Token Serve the Game, or the Game Serve the Token?

Every Web3 game eventually answers one question, whether or not anyone asks it out loud: does the token exist to serve the game, or does the game exist to serve the token? The answer is rarely stated in a whitepaper. It shows up in design decisions, in where the money comes from, and in what happens when the token price falls.

Both outcomes genuinely exist. There are projects where blockchain elements sit inside something people would play regardless, and projects where the playable layer is a thin wrapper around a token-distribution scheme. Sorting one from the other is a skill, and it is mostly a matter of knowing which signals are load-bearing.

The two orientations

In a game-first design, the blockchain solves a problem the game actually has. Persistent item ownership across an ecosystem, a player-run economy with credible scarcity, interoperability with other applications, or a way for players to keep something when a publisher shuts down a server. Remove the token and the game still works; it just loses a feature.

In a token-first design, the game is the mechanism by which tokens are distributed and demand is created. Remove the token and there is nothing left to do. These projects are recognisable by what they optimise: onboarding funnels, referral structures, yield presentations, and time-to-first-earning, rather than the moment-to-moment experience of playing.

The categories are not clean. Plenty of sincere teams have started game-first and drifted token-first under financial pressure, because token markets pay attention faster than players do. The drift is usually visible in the changelog: features that increase earning velocity shipping ahead of features that make the game better.

Where does the money come from?

This is the most useful single question, and it cuts through almost all marketing.

In a conventional game, revenue comes from people who are net spenders: they buy the game, or cosmetics, or expansions, because they enjoy playing. The money enters from outside and is consumed inside.

In many token-based games, rewards paid to players are funded by new participants buying in — through token purchases, entry assets, or character purchases required to start. When that is the structure, player earnings depend on continued inflows rather than on value produced inside the game. Growth sustains payouts; a slowdown in growth reduces them; a reversal collapses them. This is not a claim about intent, and it does not require anyone to be dishonest. It is a description of where the cash comes from, and it can be assessed by reading the token flows rather than the pitch.

The honest version of the question to ask of any project: if no new participants joined tomorrow, and the token price stayed flat, would there still be revenue? If the answer is yes — because people pay for cosmetics, entertainment, competition, or spectating — the economy has an external input. If the answer is no, the economy is internally circular, and internally circular economies are only stable while they are growing.

Faucets, sinks, and why balance is hard

Game economies are conventionally described in terms of faucets, which create currency, and sinks, which remove it. Traditional online games have decades of institutional experience here, and they still get it wrong regularly.

Token-based games make the problem harder in three specific ways. Currency leaves the system: in a closed game, currency earned tends to be spent inside it, but a tradeable token can be sold immediately, so the faucet drains outward continuously. Emissions are often fixed by a schedule set before launch, which means the supply side cannot respond to actual player behaviour without a governance process. And rebalancing is politically costly, because reducing rewards to protect the economy directly reduces the income of the people who hold governance power.

The result is a recurring pattern: strong early rewards attract participants, emissions outpace genuine demand, the token declines, participants who joined to earn leave, and the remaining player base is whoever was there for the game. Whether that residual population is large enough to sustain development is essentially the whole question.

Sinks that work tend to be ones people accept willingly — cosmetics, competitive entry, crafting that consumes materials, repair and upkeep. Sinks that feel like taxes get routed around or drive players away.

Ownership: real, but narrower than advertised

“You own your items” is the standard pitch, and it is partly true. You control a token in a wallet, and no publisher can unilaterally take it from that wallet.

What ownership does not give you is control over the item’s meaning. The game server decides what a token does inside the game. A studio can patch, nerf, or stop recognising an item without touching your wallet. If the game shuts down, you retain a token whose in-game function no longer exists anywhere. Ownership of the record is real; ownership of the utility depends on someone continuing to honour it.

Interoperability claims deserve similar scepticism. The idea that an item earned in one game will be usable in another is technically possible and practically rare, because a second studio has no incentive to accept balance decisions and art direction it did not make. Cross-game asset use tends to work within ecosystems built by a single organisation, which is a much smaller claim than the one usually made.

The metaverse question

Metaverse projects extend the same tension into virtual worlds and land parcels. The pitch is that a parcel derives value from location and foot traffic, as physical property does.

The comparison has an obvious weakness: physical land is scarce because the planet is finite, whereas virtual land is scarce only because a project chose a supply number, and nothing prevents another project from creating an adjacent world with different numbers. Location value in a virtual world depends entirely on people choosing to be there, which depends on the world being worth visiting.

That makes population the metric that matters — concurrent users, returning users, time spent — rather than parcel prices or transaction volume. Worlds sold on future foot traffic while presently empty are making a forecast, not describing a state. Some virtual worlds do sustain real communities, and those are worth studying precisely because they are the exception that shows the model can work.

How to evaluate a project without a spreadsheet

A handful of practical checks separate most of the field. Play it, and notice whether you would keep playing with earnings switched off entirely. Read the community’s own conversations and see whether they are about strategy and content or about withdrawal rates. Look at the token distribution: how much went to insiders, over what vesting period, and whether unlock dates cluster near promised milestones. Check whether emissions can be adjusted and by whom. Look at the development record for whether shipped features improve play or improve earning. And check whether the studio has any revenue source that does not depend on token appreciation.

None of that requires believing the sector is worthless. Persistent player-owned economies are a legitimate idea, and the failures so far have been failures of economic design and incentive alignment rather than proofs of impossibility. What the record supports is caution about any project where the earning mechanism arrived before the game did.

Nothing here is investment advice. For terminology, our glossary defines the concepts used above; our learn section covers the underlying mechanics; and the wider NFTs and Web3 coverage follows how these designs develop.