Your feed is not your player base: Disney's licensing win, Wolverine's backlash and the GTA 6 leak

Three of the biggest stories in games this week ran on the same broken instrument. Insomniac's developers went public about the abuse aimed at Wolverine while the sales tracking pointed the other way. Netflix says its games reached a billion people and is cutting the division that got them there. The GTA 6 leak dominated every feed in the industry, amplified hardest by the outlets condemning it. Only one company in the week's news was clearly reading the actual numbers, and it is the one nobody thinks of as a games company.

Disney stopped buying game studios and started renting Mickey

D23 marked Disney's first real embrace of gaming in roughly a decade, and the interesting part is what it embraced. The owned-studio era produced a run of write-offs: Avalanche Software, which Disney let go and which went on to make Hogwarts Legacy for Warner Bros, Club Penguin, shuttered, and Disney Infinity, discontinued. The licensing era produced Kingdom Hearts and Disney Dreamlight Valley, the two games that got real stage time at the event, the second of them built by Gameloft. Disney's own messaging around D23 put its gaming licensing revenue in the billions.

What makes the model work is not generosity toward partners. It is leverage.

Licensing is not outsourcing when you keep the mandates

The obvious objection came up on the show: outsourcing usually degrades the work, so why is Disney's version producing hits? The answer is that Disney gave up capital risk, not creative control. Justin Ruiss relayed secondhand accounts from Marvel vs. Capcom 3, where Doctor Strange reportedly had to float rather than walk in his animation, and trailers could not show heroes losing to villains. Those are industry stories rather than confirmed fact, but they are consistent with everything else Disney does with its characters.

The economics are the other half. A licensing deal puts no capital on Disney's balance sheet, no marketing spend on its P&L, and no studio on its org chart. The upside is capped at a revenue share, and the downside is a brand ding it can walk away from. Nintendo is the only comparable operator, and it applies the same pressure to film.

The announcement slate is the operator's version of the same point. Coco 2, a Coco land at California Adventure, Coco turning up inside Kingdom Hearts, and Frozen 3 are not four separate announcements. They are a content calendar, sequenced the way a live game sequences its seasons.

Insomniac read its X feed as its player base

Insomniac's developers said publicly that the internet was being cruel about Wolverine. The complaints were familiar: too much like Spider-Man, not open world, not innovative enough. The response on the show was the sharpest operator argument of the hour.

By his own tracking, Wolverine is pacing ahead of Miles Morales and sitting in Ghost of Tsushima's range, on a path to 10 million lifetime copies. That figure is his firm's, stated on air rather than published, so treat it as his read and not a public number. The direction is what matters. A studio was measuring its reception against a feed tuned algorithmically to its own team's interests, and the feed disagreed with the purchase data.

The MJ missions taught the wrong lesson

The more useful diagnosis followed. The backlash Insomniac actually earned was never about the studio.

Players did not resent Spider-Man 2 because they resent Insomniac. They resented being handed a play style they had not bought. Someone who picks up a Spider-Man game wants to play Spider-Man or Miles. Dropping them into a third character with an entirely different loop reads as a bait and switch, and it reads that way no matter how well the boss fight at the end of it lands.

That distinction matters for anyone reading sentiment for a living. "They hate us" is not actionable. "They hate the mechanic we swapped in at hour twelve" is.

Netflix is fast fashion, and games take four years

Justin Ruiss asked whether Netflix needs to adopt Disney's playbook, hand its IP to people who make games for a living, and stop building in-house. The answer from the host chair was structural rather than strategic.

Netflix cancels aggressively, and games take three to four years to build. No studio can responsibly commit that window to IP that may not survive its second season. It is why the games attached to Netflix skew mobile and disposable. The build time has to fit inside the shelf life.

Colan Neese went at the numbers instead. Netflix says its games reached a billion people, he noted, so why is it cutting everything rather than doubling down on that? A company that believed its own reach figure would be expanding against it. The behavior is the disclosure, and it points at a lifetime value that does not justify the spend.

His prescription was to stop charging at the door. Netflix competes with YouTube, TikTok and Meta, all of which are free and ad-loaded, while insisting the audience subscribe before seeing anything. On creator payouts the gap is not closeable at all. Google owns the ad stack it bought with DoubleClick and turned into Google Ad Manager, which is why it can pay creators at rates nobody else can match.

The GTA 6 leak is a preview, not an incident

The week's loudest story was a leaker pushing GTA 6 footage, up to and including the ending, across every feed in games. The material reportedly came out of Rockstar's India office off a PC build, though that was flagged on the show as speculation rather than established fact.

Two things about it should worry operators more than the spoilers do. The first is the direction of travel. Colan Neese called the game the pyramid of Giza of games, roughly a decade of human work and possibly the last title at that scale built end to end without AI anywhere in the pipeline. His concern was not this breach but the next one. Models are shortening the time it takes to get through existing software, and most security postures still assume an attacker who moves at human speed.

The second is that there is no retreat available. The genuine fix is an air gap, and an air gap is incompatible with a development pipeline distributed across continents. You cannot ship hard drives to an outsourced team and expect them to work on the game at the same time.

The amplification is the third problem. Outlets condemning the leaker ran beat-by-beat live trackers of the leak, and the manifesto he attached to it (no digital preorders, no non-single-player DLC) turned out to be packaging for a crypto coin. The attention was the point, and the industry press supplied it.

Check the tracking, not the timeline

Four companies, four readings of the same instrument. Disney is the one that looked at a report card instead of a discourse, concluded that a decade of buying studios had failed while licensing had not, and acted on it. The other three are still reading their feeds.

For an operator the discipline is dull and cheap. When sentiment and the numbers disagree, write down which one you are about to act on, and why.

The full conversation is on the replay, including the D23 breakdown, the Netflix debate in full, and a Disneyland versus Disney World argument that got genuinely heated.

Get the week's argument in your inbox with The Operator's Briefing.

Or bring your own read to the room in Discord.