When to sunset a currency in a live game's economy

Currencies accumulate. A feature ships with its own token because that was simplest, an event introduces a seasonal currency, a system gets its own resource to avoid touching the main economy. Two years in, the wallet screen has seven entries and most players understand three of them.

At some point somebody proposes removing one. That is usually the right instinct and it is a considerably more delicate operation than the addition was.

Why the tidy-up is worth doing

The cost of surplus currencies is mostly cognitive and it falls on the player.

Every currency is a thing the player has to learn: what it is, how to get it, what it buys, whether it expires. A wallet with seven currencies asks a new player to build seven mental models before the economy makes sense. Most of them do not, and instead treat the whole thing as opaque — which means they stop making deliberate decisions about spending, and deliberate decisions are what drive purchases.

The second cost is on you. Every currency is a separate faucet-and-sink pair to balance, a separate thing to price against, and a separate axis in every analysis. Seven currencies is not seven times the work but it is well over one times, and most of the extra buys nothing.

The third is that dead currencies are worse than live ones. A currency that a discontinued feature used, which players still hold and cannot spend, is a standing reminder of something that was taken away. It sits in the wallet doing nothing except reminding them.

The three ways it goes wrong

The first is deleting a currency players still hold. Whatever the balance represents, they earned or bought it, and removing it without compensation reads as confiscation regardless of how little it was worth. The amount does not govern the reaction — the fact of it does.

The second is converting at an unfavourable rate. If the retired currency converts into the main one at a ratio players consider poor, the conversion reads as a devaluation. Players who held a large balance are hit hardest, and those are usually your most engaged players. A conversion rate that saves you a small amount of currency issuance and annoys your top thousand players is a bad trade.

The third is announcing too late. Players who find out at the moment of the change have no chance to spend what they hold, which removes their agency. Players told several weeks ahead can decide for themselves, and most of the resentment evaporates when the decision was theirs.

The pattern in all three is the same: the economic cost of retiring a currency generously is almost always smaller than the trust cost of retiring it efficiently.

The sequence that works

The version that goes smoothly has four steps and takes longer than anyone wants.

Announce first, with a date far enough out that a player who plays normally can spend down a typical balance. What counts as far enough depends on your game, and it should be derived from actual balances rather than picked as a round number of weeks.

Stop issuing it immediately at announcement. Continuing to hand out a currency you have said is going away puts players in an absurd position and guarantees larger balances at the deadline.

Give it somewhere good to go. The spend-down window is much better received if there is something genuinely worth buying with the retiring currency — ideally something exclusive to the window. This turns a takeaway into a last chance, which is a completely different experience of the same event.

Convert whatever remains at a rate that is at least fair and preferably slightly generous. The residual balances at the deadline belong to players who did not engage with the wind-down, which includes players who were away. Being generous to them costs very little in aggregate and avoids the specific bad outcome of a returning player finding their balance quietly reduced.

See how we plan economy changes with a rollback path →

Deciding which one to cut

Not every surplus currency should go. The test is whether it is doing a job the main currency could not.

A currency exists legitimately when it needs to be non-fungible — when the point is that this resource cannot be converted into general spending power. Event currencies that gate exclusive rewards work this way, and merging them into the main economy would break the gate.

A currency exists illegitimately when it is just the main currency with a different name, used in one feature for historical reasons. Those are the candidates. Merging them costs nothing conceptually and removes a line from the wallet.

The middle case is a currency that was meaningful and no longer is, usually because the feature it served has faded. These are the ones worth retiring most, because they are pure overhead — players hold them, cannot use them well, and the game maintains them.

The order to work in is the dead ones first, since nobody defends them, and the redundant ones second. Leave the ones doing real work, even if the wallet stays untidy. A clean wallet is not the goal; a comprehensible economy is, and a currency that means something specific is easier to understand than one more use of a general one.

Talk to us about simplifying a live economy →