GameFi tokenomics

What changes the moment the currency has a price

A designer running a closed economy controls both ends of the pipe. Gold enters when a monster dies and leaves when a player repairs armour, and if the balance drifts the consequence is that prices in the auction house look silly. There is no exit. Nobody converts gold into rent.

List the currency and three things change at once. The exit is now permanently open, so any imbalance leaves the game as sell pressure instead of sitting in the economy. The currency acquires an external price, so players optimise for dollars per hour rather than for progression. And the designer loses the ability to devalue quietly, because a reward nerf is now a public repricing that players can trade ahead of.

Naavik and Delphi Digital wrote the retrospective on what that did to the category. Their January 2023 review put it as a widespread focus on player earnings, expressed in the term play to earn, taking precedence over fun and sustainability, which led to a financial bubble.⁵ The mechanics were not new. Attaching a market price to them was.

Faucets and sinks, and why the ratio is the entire job

The governing constraint is old and simple. Terry Chung's 2022 survey of virtual economies for 1kx states it as stuff in is roughly equal to stuff out: for an economy to hold, the rate of asset issuance has to be balanced against the rate at which assets are consumed or removed.¹ A faucet mints. A sink destroys. Everything else is decoration on that identity.

Chung's survey is useful because it catalogues sinks that were tested on real players before crypto existed. Utility taxes such as travel costs, armour repair and stat resets. Rare collectibles auctioned for currency. Gambling mechanics with negative expected value. Crafting that consumes both items and currency. Taxes on trade and land. Content gates that charge for access. Each one has a track record, including the ones players hated.¹

Two of those have measured results worth stealing. Transaction taxes were the largest single ISK sink in EVE Online for two consecutive years. In RuneScape, the 2% Grand Exchange trade tax is not merely destroyed, it funds algorithmic buybacks of items that are experiencing supply bursts. Axie ran a 4.25% marketplace tax on top of its currency loop.¹ A marketplace tax is the sink that scales with activity rather than with the designer's guesswork, which is why it belongs in the design before launch and not after the first inflation crisis.

Axie's ledger: three faucets, one sink

Naavik's November 2021 deep dive, published while the game was still growing, counted the flows precisely: three main sources of SLP and one sink. SLP entered through daily quests, through PvE adventure mode and through PvP arena wins, and left only through breeding.³ The same report noted that the scholar population, players borrowing assets from managers in exchange for a share of earnings, cashed out SLP as a daily wage and therefore sank almost none of it into breeding.³

The Verge described the same structure from the outside in April 2022. The economy rested on three resources, SLP, the Axies themselves and the AXS governance token, and the game produced two of them in constantly increasing quantities, with SLP available from quests and single player grinding in a way the article compared to printing money.⁴

That is the whole failure in one sentence: the only sink required players to spend on a long-term asset, while the dominant cohort was there to convert today's output into today's income. When the design's single exit valve is one the largest user segment has no reason to use, the faucet rate is the economy.

The arcade token alternative: bound the price on purpose

There is a version of this that does not depend on a market. a16z crypto's 2025 work on arcade tokens describes a token type whose value is programmatically bounded inside one product ecosystem, held stable by two mechanisms rather than by trading.² A faucet sells the token at a published price, which sets a ceiling, because nobody buys from a speculator at a dollar what the vending machine dispenses at a quarter. An optional buyback below that price sets a floor.

Run the numbers on a game currency. Suppose a studio sells its soft currency at $0.010 and commits to buy it back at $0.008. A player who grinds 10,000 units has produced something worth at most $100 and realisable at $80, and the band between those two figures is fixed by policy rather than by order book depth. On the other side, the studio's liability is bounded at one cent per unit outstanding, which is a number an accountant can put on a balance sheet. a16z makes exactly that point: issuers who print these tokens must still track their shadow value against future redemption.²

The tradeoff is honest and it is not free. A bounded currency will not appreciate, so it cannot be used as the growth story, and the issuer takes on a redemption obligation it has to fund. What it buys is a currency the design team can retune without triggering a market event, which is the single capability play to earn economies gave away.

Where the design usually breaks

Four patterns account for most of it. Sizing faucets to acquisition targets, because paying players to arrive works until the payment has to come from somewhere other than the next arrival. Building the only meaningful sink into an asset the earning cohort has no reason to buy. Treating the reward token's price appreciation as evidence that the economy works, when during growth it is evidence of nothing except growth. And shipping without a marketplace or transaction tax, which leaves the designer with no sink that scales automatically.

The correction is not subtle. Faucet rates get modelled against a flat player count and a declining one, not against the plan. Sinks are sized as a percentage of issuance and monitored weekly. And the reward loop is designed so that a player who never sells anything still has a reason to play, because if the only reason to play is the payout, the game is a payments product with graphics.

What to settle before the token lists

Name every faucet and every sink, then write the expected daily quantity of each at three player counts: your plan, flat, and half. If total sinks are under half of total faucets at flat, the economy is already inflating and the launch will simply reveal it faster. Decide whether the game currency needs to be tradeable at all, because the arcade token structure is available and it removes the exit problem by construction.² Separate the currency players earn from the asset you want held, and make sure the earning currency is not required to acquire the held one, or the separation is cosmetic.

Then anchor the thing underneath. A game economy is a distribution mechanism for a product people want to use. If the game is not worth playing without the token, the token is not fixing that, it is postponing the discovery. The 2023 reset in this category was builders arriving at that conclusion collectively.⁵

This page is design reference. It is not investment advice and it is not a recommendation on any token.