Daylila

Gaming · Sunday, 7 June 2026

01 · Briefing · what happened

The platform takes its cut, and the live-service bet keeps coming apart

Gaming 5 min 80 sources

Court documents in Valve's antitrust case show how Steam guards its 30% fee, a California bill would force studios to keep dead games playable, and the live-service gamble that was meant to fund the industry keeps producing closures.

Key takeaways

  • Court documents in Valve's antitrust case detail a rule that stops developers selling games cheaper outside Steam — which is how the store protects its 30% cut of every sale.
  • A California bill (AB 1921) would force studios to keep server-dependent games playable after shutdown, attacking the model where a one-time purchase quietly depends on a server the publisher can switch off.
  • The live-service bet keeps folding — the PUBG creator's survival game shut down inside a year, and Sony's first-party sales nearly halved since 2020 — while a single-purchase indie sold 300,000 copies in three days.

This week’s games-business news has one thread running through it: who controls how a game makes money after it ships, and who pays when the bet goes wrong. A long-running lawsuit against Valve, owner of the Steam store, surfaced fresh detail on how the company protects its cut of every sale. A California bill advanced that would force studios to keep online games playable after they pull the plug. And another live-service game shut down — the model that was supposed to fund the industry keeps closing teams instead.

Steam’s 30% cut, and the rule that protects it

Valve has run the PC games store Steam for two decades. It takes 30 cents of every dollar a game earns there — the “platform cut,” the fee a storefront charges to host, sell, and deliver a game. A class-action lawsuit running since 2021 says that cut is too high and that Valve uses its size to keep it that way [44][6].

This week Bloomberg published documents from the case [6], and other outlets covered the detail [44][40]. The core allegation is a clause the suit calls a “Platform Most Favored Nation” rule: developers who sell on Steam can’t offer the same game cheaper on a rival store [44]. The mechanism matters. If a developer can’t undercut Steam anywhere, the 30% fee can’t be competed away — a rival store offering a smaller cut gives the developer no room to pass the saving on. Court filings say Valve told some developers, including the publishers behind Rainbow Six Siege and Middle-earth: Shadow of War, that games could be removed from Steam if priced lower elsewhere [44].

Valve’s founder Gabe Newell pushed back, saying players have “enormous choice” about where to buy games [21]. The case is unproven, and a parallel suit runs in the UK [44]. But the fight is the same one Apple and Google faced over their app-store fees — a single dominant gateway, a fixed cut, and a rule that stops anyone routing around it [6].

Keeping a game playable after it dies

On the same theme of control after the sale: California’s State Assembly passed AB 1921, the Protect Our Games Act, by 43 votes to 16 [33]. It would require companies to warn players before they shut down a server-dependent game, and to leave a way to keep playing it — offline access, community-run servers, or some other workable option [33][4].

The bill came out of the Stop Killing Games movement, which formed after Ubisoft pulled the racing game The Crew from its servers and from players’ libraries, leaving paying customers with nothing [33]. The system underneath: many modern games are sold as a one-time purchase but depend on a server the publisher can switch off whenever the game stops earning. When the server dies, the game dies — even for people who bought it. The bill tries to break that link. It still has to clear the State Senate [33].

Another live-service bet folds

Brendan Greene — “PlayerUnknown,” the designer behind the battle-royale hit PUBG — stopped development on his survival game Prologue: Go Wayback less than a year after launch and laid off staff [11][13]. “I have reached the limits of how far I can continue to fund this journey,” he wrote, saying his independent studio could no longer afford the work [13]. The game will be made free and buyers refunded [11].

This is the live-service model failing at small scale. A live-service game keeps selling things to the same players for years — battle passes, cosmetics, new seasons — instead of one purchase up front. It only works if a large crowd stays and keeps spending, so a studio has to fund the game for a long time before it knows whether the crowd will show. Prologue didn’t draw one, and the runway ran out [11].

The same pattern shows at the top. New data on Sony’s PlayStation business shows first-party game sales — the games Sony itself makes — fell from 58.4 million units in its 2020 financial year to 28.9 million in 2024, before a modest rebound to 32.1 million in 2025 [14][8]. Part of that is the COVID-era bump fading [14]. But part is Sony’s own live-service push, which aimed to launch 12 such games by 2025 and delivered a handful [14]. Its online shooter Concord was pulled two weeks after launch; other projects were cancelled and studios closed [14]. The bet meant to fund the future kept producing write-offs instead.

The other side: people still buying, and a different model that worked

For all the contraction, more people play than ever. The US industry body reported 212.3 million Americans — 67% of those aged 5 to 90 — play at least an hour a week, up 7.2 million on last year, even after console and PC price rises [24][31].

And a counter-example to the live-service squeeze: Mina the Hollower, a single-purchase action game from a small studio, sold 300,000 copies in three days at $20 [29][19]. One game, bought once, no years-long server bill. It reportedly secured the studio’s finances in a week [19]. The demand is there; the question every studio is now testing is which way of charging for it survives.

Where the contraction shows up: AI and jobs

Underneath all of it is cost. Development budgets keep rising, and studios are reaching for two levers: layoffs and AI. The US tech sector announced 38,242 job cuts in May, the most since August 2024, and 123,653 so far this year — up more than 65% on the same period in 2025 [52]. Games studios are part of that wave; mobile developer Metacore cut 159 staff and closed offices this week [10].

The AI lever drew open conflict. The new Tomb Raider: Legacy of Atlantis disclosed “AI-assisted assets” later “refined by humans,” to cut development cost [3][2]. The voice actor Jennifer Hale, who played the lead in Mass Effect, called on studios to own the choice: “Ain’t nobody making you do it” [1]. The mechanism is plain — AI is a way to make games with fewer people, which is exactly why it lands hardest on the people who make them.

02 · Lesson · why it matters

Why the toll on the bridge never falls

When one gateway controls how everyone gets to the customer, it can set a price no rival is allowed to beat — and the toll stays high not because it's fair, but because there's no road around it.

Steam takes 30 cents of every dollar a PC game earns. That number hasn’t moved in twenty years. This week, court documents in a lawsuit against Valve, the company that owns Steam, showed one reason why. Set aside whether Valve broke any law — the case is unproven. The interesting part is the mechanism. It explains a shape you’ll see far beyond games.

A gateway is not the same as a seller

Most businesses sell you a thing. A gateway sells access to the people who want your thing.

Steam doesn’t make most of the games on it. It sits between the people who make games and the people who play them. To reach the millions of players who buy on Steam, a developer almost has to be on Steam. That position — standing in the one doorway everyone uses — is worth more than any single product. The developer needs the doorway more than the doorway needs any one developer.

That’s the source of the 30% cut. It isn’t a charge for work done. It’s a charge for the doorway.

A fixed fee only holds if no one can undercut it

Here’s the part that’s easy to miss. A high fee, on its own, invites competition. If Steam takes 30% and a rival store takes 12%, a developer could sell cheaper on the rival, pass the saving to players, and pull the crowd across.

So the fee alone isn’t enough to protect itself. Something has to stop that escape.

The lawsuit points at a rule it calls a “Platform Most Favored Nation” clause. In plain terms: if you sell on Steam, you can’t sell the same game cheaper anywhere else. Court filings say some developers were told their games could be removed from Steam if they broke it.

Read what that does. It doesn’t force a developer to leave the rival store. It removes the reason to. If the price has to be the same everywhere, a smaller fee elsewhere buys the developer nothing they can hand to players. The rival’s cheaper cut can’t become a cheaper game. The competition is real on paper and dead in practice.

The toll holds because there’s no road around it

Now the 30% can sit still for twenty years. Not because everyone agrees it’s right — many developers say it’s too high — but because the one move that would force it down has been closed off.

This is the quiet engine under a lot of dominant gateways. The fee feels fixed, almost natural, like the price of gravity. It isn’t. It’s held in place by a rule that prevents the undercut. Take the rule away and the fee has to defend itself in the open, against rivals free to be cheaper. Keep the rule and the fee never has to.

You’ve seen this bridge before

Apple and Google faced the same fight over their app stores — the same 30% cut, the same one dominant doorway, the same rules that kept developers from routing around it. The detail changes; the shape repeats.

It repeats outside software too. A booking site that forbids a hotel from offering a lower rate on its own website. A marketplace that won’t let a seller be cheaper elsewhere. A delivery app whose contract bars a restaurant from charging less for walk-ins. Different industries, one move: control the gateway, then forbid the undercut. The toll on the bridge stays high for exactly as long as there’s no road around it.

What the pattern lets you see

Next time a fee feels permanent — a platform cut, a service charge, a commission everyone grumbles about but pays — ask the second question, not the first. The first question is “is this fee fair?” That’s a debate with no end. The second question is sharper: “what stops someone from offering it cheaper?”

If the answer is “nothing — rivals just haven’t shown up yet,” the fee is exposed, and time will probably pull it down. If the answer is “a rule that bans undercutting,” the fee is protected, and it can sit still for decades no matter what anyone thinks of it. The price you can see is set by the road you can’t.

03 · Lab · your turn

The Toll You Can't Route Around

Rehearse a studio's pricing move against a dominant store and feel how a no-undercut rule, not the fee itself, keeps the cut high.

Across the beats