Unwinding currency inflation after a generous event

An event pays out more than intended. Maybe a multiplier stacked in a way nobody modelled, maybe participation was far higher than forecast, maybe the rewards were simply set too high. Either way, the event ends and the player base is holding considerably more currency than the economy was designed around.

The instinct is to correct downward — reduce rewards for a while until things normalise. That instinct is usually wrong, and understanding why points at what does work.

Why cutting rewards does not fix it

The surplus is a stock. It already exists, sitting in wallets. Reducing the rate at which new currency enters does nothing to the stock — it only slows what is added on top.

So a reward cut takes a long time to matter and is felt immediately as a takeaway. Players experience reduced rewards as a punishment for an event they enjoyed, which is a bad exchange for a correction that will take months to have any effect on the surplus.

Worse, the cut applies to everyone including players who did not participate in the event and are not holding a surplus. Those players are being penalised for someone else's windfall, which is both unfair and visible.

The general rule is that stock problems need stock solutions. A surplus of currency is drained by giving it somewhere to go, not by slowing the tap.

Draining without a takeaway

The effective response is to add things worth buying, priced against the new balances rather than the old ones.

This works because it is voluntary and it feels like content rather than correction. Players with large balances get something to do with them, which they generally want — a large unspendable balance is unsatisfying. Players without large balances are unaffected, since the new items are simply out of reach for now and they were not the problem.

The pricing matters. Items priced at the old scale get bought instantly by everyone holding a surplus and drain very little. Items priced against the surplus — meaningful relative to what the wealthy players actually hold — drain properly and remain aspirational for everyone else.

The second lever is time-limited sinks. A window in which something desirable is available for currency creates urgency to spend, which converts a passive balance into an active drain. This works particularly well immediately after the event, while the currency is still front of mind and before players have mentally banked it.

The third is consumables. A recurring use for currency that is consumed rather than owned keeps draining indefinitely rather than once, and is the mechanism most economies use for exactly this reason.

What to do about the ongoing rate

Separately from the stock problem, if the event revealed that reward rates were too high in general, that is worth fixing — but as a deliberate rebalance rather than a reactive cut.

The distinction is in how it is done. A quiet reduction applied immediately reads as a reaction to the event and generates the takeaway response. A rebalance announced with reasoning, applied alongside changes to costs so the relationship is preserved, reads as maintenance.

The other option, often better, is to leave rates alone and raise the cost of new content instead. This resets the relationship going forward without reducing anything anyone currently receives. It is slower and it avoids the whole category of takeaway complaints, and for most games the slower path is fine because the surplus is not an emergency.

See how we model what an event does to currency supply →

Preventing the next one

The specific thing that produces runaway events is multiplicative effects that were modelled additively.

An event with a participation bonus, a streak bonus, and a purchased multiplier can pay several times what any individual component suggests, and the modelling frequently checks each component rather than the maximum combination. The player who stacks everything is the one who defines the top of the payout, and there is always one.

The check that catches this is to compute what the most efficient possible player earns, not what the typical player earns. If that number is uncomfortable, the event needs a cap regardless of how reasonable the average looks.

The second check is a total issuance limit for the event as a whole. If the event is expected to distribute a certain amount across the population, a hard stop at a multiple of that catches the case where participation or efficiency was badly misjudged, and it fails safely rather than continuously.

Neither of these is sophisticated. They are the two numbers that get skipped because event modelling naturally focuses on the experience of a typical participant, and the problems come from the tail.

One more thing worth doing is recording what actually happened, in enough detail to reference later. The specific combination that produced the overshoot — which bonuses stacked, what participation rate was assumed versus observed — is the input to designing the next event of that shape. Without it, the correction gets made, the surplus gets drained, and the same structure ships again in six months with the same flaw, because nobody remembers which of the three components was the one that compounded.

Talk to us about event economy modelling →