Sink placement: front-loaded or back-loaded progression

Every progression curve has to decide where the costs sit. Cheap early and expensive later, expensive early and flatter later, or something in between. The choice is usually made by instinct and it determines more about the game's economy than most of the tuning that follows.

What front-loading does

Front-loading means the early steps cost meaningfully relative to what a new player earns. The first upgrades are a real decision.

The argument for it is that it establishes the economy as consequential from the start. A player who has to choose between two early upgrades learns that choices matter and that currency has value. Games where the first ten upgrades are effectively free teach the opposite — that currency is abundant and decisions are not real — and then have to unteach it later.

The argument against is that it is a filter. Every early cost is a place a player can decide the game is asking too much before they have any investment in it. Front-loaded curves lose players who would have stayed under a gentler start.

The resolution most games land on is front-loading the decision without front-loading the cost. Early choices that are mutually exclusive but cheap teach the same lesson without the filter. The player picks one of two paths, both affordable, and learns that choosing has consequences.

What back-loading does

Back-loading means the early game is generous and the costs rise steeply later.

This retains better early, obviously. Players get momentum, feel progress, and become invested before the game asks for anything significant.

The cost lands on the long-tenured players, and it lands hard. If the curve rises steeply enough, players eventually reach a point where each step takes an implausible amount of play. That is where the game stops being a progression and becomes a grind, and where the paying decision changes character — from "this would be convenient" to "this is the only realistic route".

The second cost is on the economy's stability. A back-loaded curve means that a player's earning power grows as they progress while the things they want get much more expensive. Small changes to earn rates then have very large effects at the top of the curve, which makes the economy hard to tune without breaking something.

The shape that usually works

Most durable economies are neither, and the reason is that they have several curves running at once.

A gentle main progression that keeps players moving, plus a steeper optional track for players who want depth, plus a set of purely lateral things that cost currency without being on any curve. The gentle track handles retention, the steep track handles the players who stay for years, and the lateral spending is what absorbs surplus currency from players who have finished the tracks.

That third element is what most economies lack and what causes the classic late-game problem: players who have progressed past everything and have currency with nowhere to go. A curve alone cannot solve this, because any curve eventually ends. Something that consumes currency without being progression — cosmetics, consumables, anything horizontal — is what keeps the economy functioning after the curve runs out.

The corollary is that lateral sinks should be introduced before players need them, not after. A sink added in response to an accumulation problem arrives to a population that already has large balances, and it gets consumed instantly without changing the ongoing dynamic.

See how we look at where costs sit in a progression →

Testing the placement without shipping it

The useful check on a curve is not whether the numbers look right in a spreadsheet but what a player's balance does over time under it.

Simulating a typical player's earning against the curve, over a few months of play, shows whether the balance climbs, stays flat, or drops. A curve where the balance climbs steadily means the costs are too far back — the player is out-earning the things to spend on. A curve where the balance is pinned at zero means the costs are too aggressive and the player is permanently short.

The pattern you generally want is a balance that oscillates: rises for a while, drops when something is bought, rises again. That means the player is regularly reaching a decision point, which is where both engagement and purchases come from.

Running that simulation for several different player types — someone who plays daily, someone who plays weekly, someone who is very efficient — shows which players the curve serves and which it does not. Most curves are accidentally tuned for one play pattern, and finding out which one before launch is much cheaper than after.

The reason this is worth doing carefully is that a progression curve is one of the hardest things to change later. Players are distributed along it, and any adjustment advantages some of them and disadvantages others. Getting the shape roughly right at the start is worth considerably more than any amount of subsequent tuning.

Talk to us about progression and economy shape →