Microsoft moves to close the award-winning studios it bought — and the rest of the industry shows why

Gaming 4 min 80 sources

Xbox is negotiating spin-off-or-shutdown deals for Compulsion, Double Fine, and Ninja Theory; Sega writes down its Angry Birds buy; a nine-person team is laid off a month after a praised launch. A $200bn industry is contracting at the top.

Key takeaways

Microsoft is negotiating to spin off or shut down award-winning Xbox studios including Compulsion, Double Fine, and Ninja Theory — the same studios it once bought.
Buying a studio or a famous franchise buys its past output, not the conditions that made it; Sega's $200m Angry Birds write-down is the same lesson at scale.
The games market hit a record $200 billion in 2025, yet studios keep closing — because the money is concentrating in a few live-service hits.

The biggest games company on Earth spent years buying acclaimed studios. This week it started preparing to close them.

Xbox is unwinding the empire it built

Microsoft is in active talks to spin off or shut down several Xbox studios, including Montreal’s Compulsion Games — maker of the Peabody Award-winning South of Midnight — San Francisco’s Double Fine (Psychonauts), and Cambridge’s Ninja Theory (Hellblade) [1][5][7]. The studios may be allowed to buy back their own independence, but many staff will lose their jobs regardless [1]. Bloomberg reports “several other studios across the portfolio” are also at risk [1].

The same week, Craig Duncan — head of Xbox Game Studios — left after 15 years, along with his chief of staff [4]. It’s part of a “reset” run by new Xbox CEO Asha Sharma, who took over in February. In a staff memo, Sharma and content chief Matt Booty admitted the division had “over extended” its studio system: “we have not adequately funded them to compete and win” [7].

Here’s the mechanism worth holding. Microsoft bought these studios for the work they could do. But a studio’s output isn’t a thing you can warehouse — it comes from a small team with the freedom and time to make something strange. Folded inside a company that needs every project to clear a corporate revenue bar, that freedom is the first thing to go. The studio underdelivers against a target it was never built to hit, and closing it becomes cheaper than funding its next game. Microsoft is now reportedly weighing whether to spin off all of Xbox [39]. One former PlayStation executive called the tactics “a basic misunderstanding of how the interactive entertainment world moves” [63].

Sega already ran this experiment

If you want to see where Microsoft’s logic ends, look at Sega and Rovio. In 2023 Sega paid $776 million for the maker of Angry Birds [14]. This February it wrote down $200 million of that, saying Rovio’s profit had “fallen below the initial forecast” [14].

Analysts told GamesIndustry.biz the deal was priced on an assumption that no longer held: that mobile would keep growing and that buying a famous name buys you a future. “Leveraging IP tends to provide a false sense of security,” said analyst Joost van Dreunen — Sega “acquired a fading franchise” whose sales had fallen from €309m to €181m [14]. Sega couldn’t even bolt Rovio’s ad technology onto its own games; CEO Haruki Satomi said the “operational and marketing methods” simply differed too much [14]. A library of IP, it turns out, isn’t a machine for making money. It’s a snapshot of money already made.

A month from praise to redundancy

The contraction isn’t only happening to giants. Kwalee Labs — a nine-person team — shipped Luna Abyss, a well-reviewed bullet-hell shooter, in late May [3][9]. It launched day-one on Game Pass and scored 81 on Metacritic [3]. On June 16, the whole team was laid off [3][9]. CEO Hollie Emery said the decision was “completely outside of our control” [9].

The detail that explains it: Luna Abyss peaked at 317 concurrent players on Steam [3]. A day-one Game Pass deal pays a studio a fee, but it also means many players try the game “free” inside their subscription rather than buying it — so the launch can earn praise without earning a sustaining audience. Seven years of work by nine people ended a month after release [3]. France’s Don’t Nod, maker of Life is Strange, is in similar trouble: auditors warned it could run out of cash by November without new financing [20][40].

The number that makes it strange

All of this is happening inside a growing market. Global games revenue passed $200 billion in 2025, with PC leading the growth, according to analysts Newzoo [6][34]. The industry has never sold more.

That’s the contradiction worth sitting with. Record revenue, and a wave of closures at the studios that make the games. The money is real — it’s just flowing to a shrinking set of enormous live-service hits and away from the mid-sized and acclaimed-but-small. Epic CEO Tim Sweeney described it as a “tidal wave sweeping over the AAA game business” [37]. A big market and a brutal one are not opposites. They can be the same week.

02 · Lesson · why it matters

You can buy what a team made, but not the thing that made it

A company can own a studio's games and still destroy the studio, because the part worth having was never on the balance sheet.

The thing on the receipt isn’t the thing you wanted

When Microsoft bought Compulsion, Double Fine, and Ninja Theory, it paid for studios that had made unusual, admired games. When Sega paid $776 million for Rovio, it paid for Angry Birds. Both buyers wrote a cheque for a track record.

But a track record is a list of things that already happened. The buyer doesn’t actually want the old games — those are already made and sold. It wants the next ones. And the next ones don’t exist yet. What the buyer is really paying for is the team’s ability to make something good again.

That ability is the one thing a purchase can’t transfer cleanly. It isn’t stored in the studio’s name or its back catalogue. It lives in a particular group of people, working a particular way, under particular conditions. You can buy the building. You can’t buy the weather inside it.

The conditions are the product

Why did a small studio make a strange, award-winning game in the first place? Usually because it was small. A team of a few dozen people could chase an odd idea without a committee. They had slack — time to throw work away and start over. They could fail on a project and survive, because their costs were small enough to carry.

That looseness looks like waste to a large company. So the large company, reasonably, tightens it. Every project now has to clear a revenue bar set by a business that earns billions. Headcount gets justified against forecasts. The odd idea now needs a spreadsheet attached.

None of those are stupid decisions on their own. But together they remove exactly the conditions that produced the thing the buyer admired. The studio that made a weird, beloved game can no longer afford to be weird. So it makes something safer, which does worse, which confirms it was a bad bet — and closing it becomes cheaper than funding it. The buyer destroyed the capability by managing it well.

The same shape, twice

Sega ran this experiment to the end. It bought a famous franchise on the assumption that a famous name keeps earning. But Angry Birds had already peaked; its sales had nearly halved before the deal closed. Sega even found it couldn’t connect Rovio’s own technology to its other games — the two studios simply worked in different ways. A $200 million write-down later, the lesson is plain: a library of hits is a photograph of capability, not the capability itself.

And it isn’t only about acquisitions. A nine-person team shipped a praised shooter and was laid off a month later, after the game found only a few hundred players a night. The work was good. The condition that would have let the team try again — enough revenue, or a patient owner — wasn’t there. Good work and a surviving studio are two different things, and the second is the one nobody can buy outright.

Who this reaches

It’s tempting to read this as an industry story about executives and balance sheets. It isn’t only that. The strange, small, admired games are the ones most exposed to this pattern — and those are the games that make the medium interesting. When the conditions that produce them get optimised away, the loss travels to the player who never hears the studio’s name, who just notices, years later, that fewer odd things get made.

It reaches the people who made the work, too. “Completely outside of our control,” the laid-off CEO said — and she was right. The team did the part it could control: it made something good. The part that decides whether a studio lives sits a level up, in conditions set by people who were never in the room when the game was made.

On the whole

Almost everything good is made under conditions, not just by people — slack, trust, the freedom to fail small. Those conditions are easy to admire and easy to miss, because they don’t show up on any receipt. We keep treating the output and the conditions that produced it as the same thing, buying the first and quietly dismantling the second. If you find yourself holding something good — a team, a habit, a thing that works — it’s worth asking what conditions are keeping it alive, and whether the sensible improvements you’re about to make are the same ones that would kill it. The hardest part to see is that you can be doing everything right by the numbers and removing the thing the numbers can’t measure.

03 · Lab · your turn

Run the Studio

Rehearse how each sensible management choice trades the unmeasurable conditions that make hits for the number corporate can see.

04 · Hope · carry this

The teams that get closed don't vanish — they scatter, and the strange instinct that made one good game tends to surface again somewhere smaller, where it's free to be strange. A game made by nine people in two months earning millions on Steam this week is proof that the conditions which make something worth playing can never be fully owned, only borrowed, and they keep finding new rooms to grow in.