A commuter on a platform taps Install on a game they saw in a thirty-second ad. A title screen fades. A tutorial level plays some of the moves for them, and the starter hero hits hard enough to feel good. Then the game hands over a pile of premium currency and something worth spending it on. Twenty minutes later, the currency is gone. What comes next is a chest the player cannot afford, and a small button that offers a reward video in exchange for a look at an ad.
It reads as a gift. It is closer to a sales call that happens to have a combat system.
That moment is the product.
Free-to-play is less a pricing decision than a financing structure: the studio buys an audience it has never met, then collects from that audience unevenly.
The buying happens on user-acquisition networks, where a studio bids for installs against every other studio bidding for the same thumb. What it pays per install is a cost settled up front, before the player has done anything at all. Everything that follows is an attempt to earn that number back. That is why free-to-play studios talk about their games in the language of funnels and cohorts rather than genres and reviews. Install prices vary by genre, country and season, so a studio losing money on one market will often stop bidding for players there, and the same player can be bought again later, at a higher price, if the game has proven it can hold them. A studio that cannot tell which ad produced which install is bidding blind, and blind bidding raises the price for everyone.
Most players never pay anything, and the model does not require them to. Free players fill matchmaking queues, populate leaderboards, keep a world from feeling abandoned, and generate the ad impressions that rewarded video turns into revenue. Ad inventory has a price too, set in auctions that run behind the screen, so a player who never spends a cent still earns something every time they choose to watch.
Rewarded video is a curious bargain: the player volunteers, the advertiser pays for a completed view, and the studio collects a fee for the introduction. Free players are sold to the paying minority twice, once as opponents and once as an audience. A game with no non-payers has no game for the payers to win, which is why the free tier is not charity but inventory.
So the opening session has a job, and it is measurable. Retention gates, the share of installs that return the next day and the next week, decide whether the studio can afford to buy more traffic. Someone who quits in the first half hour costs the full price of their install and returns nothing. The tutorial therefore doubles as a filter, sorting purchased attention into users worth keeping and users to stop paying for. Conversion is easiest early, which is why the first offer often arrives before the player has decided whether they even like the game.
Which is why the generosity is real, and why it stops.
Designers tune the exact step at which the tap runs dry. Gems that flowed freely during the tutorial turn scarce at the moment the player has learned enough to want something, and energy timers, build queues and difficulty spikes do the same work elsewhere. Push the wall too early and the player leaves before a habit forms; push it too late and the habit forms without ever being charged for. That trade-off is settled with data. Every tap in the opening session is logged. Studios run experiments on button placement, reward size and the timing of the first offer, so the version most players see is the one that performed best in a test. A newcomer is not being welcomed so much as measured, and the measurement exists to find the price at which they will flinch. A session that ends inside the tutorial is not treated as a failure of taste. It is treated as a failure of pricing.
Follow the money and the list is short. Store operators take a cut of every purchase. Ad networks take a share of every rewarded video. Publishers recover marketing budgets and royalties out of whatever the studio earns. What remains covers the install cost, the salaries and the servers, and the studio carries the creative risk while standing last in line. A small minority of players carries nearly all of it, and the non-payers pay in time, attention and patience. That is where the starter bundle comes in. A deep discount on a first purchase, offered once and expiring, anchors every later price against a number that will never return and sorts the audience into those who will pay something and those who will not. Bundles double as a segmentation test, since the size of the first purchase often predicts the ones that follow. After the first transaction the barrier is gone, and later offers address a customer instead of a stranger.
Concentration does the heavy lifting. A small share of players accounts for the large majority of spending, and the design aims at them without appearing to: leaderboards that reward spending, limited-time offers that expire, progression that can be bought past. The median player is subsidized by the extraordinary one. The game only needs to be pleasant enough to keep the many from making the few play alone. Spending has a shape, too: a crowd buys small conveniences, a handful buys large ones, and the design keeps both in the same world.
The arrangement shapes what gets built. Story, art and combat tuning all have to survive the funnel, which means they are judged by whether they move a player toward the ask. Studios hire monetization designers and live-operations analysts alongside artists, and the opening half hour becomes a shared research project rather than a creative statement. A designer who builds a beautiful first level that converts badly is not paid for the beauty.
The incentives are neither hidden nor sinister. They simply point somewhere other than craft.
Live operations staff spend their weeks on a calendar of events and offers, because a game with no fresh reason to return decays.
Three forces could reset the arithmetic. Privacy rules that limit how precisely advertisers can target and measure have made installs harder to price, which raises the cost of the model's raw material. Consumer-protection rules aimed at loot boxes and aggressive offers in some markets would remove specific levers.
Subscription catalogues, which pay a flat licence fee and hand players a library with no wall inside it, invert the funding entirely: the player pays before playing, so the first thirty minutes has no job beyond being good. That last model also moves the customer, from players to platform executives working to a fixed budget. None of these has replaced the existing arrangement, largely because none has solved discovery. A premium game still has to be found, and finding an audience without buying one at scale remains the problem that free-to-play answers with a budget.
What the model leaves open is not whether the wall is fair, but what it does to the thing standing behind it. A game can be designed for the moment a player runs out, and that design can be executed with real skill. Whether a first half hour built to test a stranger can also be the half hour that makes them care is a question the industry has found it can keep postponing.
The arithmetic still works without an answer.