I know what a date three years out is worth to the people who will have to hit it, and the answer is nothing for two of those years and then everything at once.
Blizzard spent Saturday announcing a Diablo 5 for spring 2029 and an open-world StarCraft for 2030. Some of what was on that stage is real and close: a new Warcraft 3 campaign went live the same afternoon, the first new Warcraft RTS content since 2003, and WoW Forever arrives in November with three zones that were drawn on the vanilla map twenty years ago and never opened. Good, concrete, dated things.
The two headline announcements are three and four years away. Blizzard’s president told the press why they were made now anyway: to show confidence.
That is the sentence I cannot get past. Confidence was the deliverable. And unlike Diablo 5, it shipped on the day it was scheduled to.
Show of confidence
You can check this company’s announcements against its own receipts, which is more than you can do with most.
At BlizzCon 2023 Blizzard did something nobody had done before and announced three World of Warcraft expansions at once — the Worldsoul Saga, a trilogy, laid out in a single ceremony. This weekend the third of them, The Last Titan, got a cinematic and a promise of more details next year. Three years of runway have produced a trailer and a continent that will be rebuilt. Before Diablo 4 shipped, the plan was annual expansions; the second one turned up in April this year, a year and a half after the first, and this weekend the game’s expansion era was declared over and the next class was repackaged as a DLC pack for 2027.
Schedules slip. Scope changes. A game often gets better because somebody had the nerve to throw the third version away, and I have more patience for a slipped date than most of the people shouting about it, because I have seen what the alternative looks like on a disc.
What the record shows is what the announcement is for. A date that far out is not a schedule. Nobody at that studio is working to spring 2029 — spring 2029 is what you say out loud so that the room, the press and the parent company hear a number instead of a silence. The game will arrive when the third combat prototype stops being embarrassing, the same as every other game.
Which would be fine, if the announcement were not becoming the expensive part.
The one surface with a deadline
Level-5 ran its VISION 2026 II broadcast on 10 September, and by Friday its president had apologised for the generative AI in it. The studio wanted the presentation to be more spectacular. The side-by-side screenshots were up within hours: backgrounds with that soft wrongness people have trained themselves to spot, detail that dissolves when you look straight at it.
Here is why a machine turns up in a sizzle reel before it turns up in a game, and it has nothing to do with courage or ethics. A marketing video is the one surface in a games company with a fixed public date, no engine to run inside, no QA pass, no certification, and nobody downstream who has to open the file afterwards. A melted window frame in a trailer breaks nothing. The same window frame in the level means an artist opens it, a lead signs it off, and it ships for ten years.
So the showcase is where the cheap experiment goes. It is also, now, the only thing in the building that genuinely cannot be late. The games move. Saturday at six does not.
Put the two stories side by side and they are the same object. Blizzard needed confidence by Saturday and produced it. Level-5 needed spectacle by Thursday and produced that. In both cases the thing with the hard deadline, the thing that actually shipped on time, was the presentation.
Twenty-nine names in a Google Doc
On 11 September Polyarc stopped. Twelve years, or eleven depending on who you ask, two well-reviewed Moss games, and a flatscreen edition shipped in July that gathered two hundred thousand wishlists and did not convert them into enough money to keep the lights on. Twenty-nine people out.
The studio’s last administrative act was a public document listing every one of them — name, job title, contact, portfolio link — so somebody else could hire them. That is a closure notice written by people who were still doing their jobs on the way out of the building.
Set that against the weekend’s other news and the arithmetic is hard to look at. One company can hold a stage and be believed about 2030. Another can ship two finished games that work, that critics liked, that people who played them still talk about, and be unable to hold twenty-nine desks open past September 2026. Confidence, it turns out, is not distributed according to what anyone has built.
VR took the funding away and that is the honest cause here — the platform money stepped back about eighteen months ago and the invoices are arriving now. But the shape is not unique to headsets. A fellow developer said last week that the industry’s volatility has always been around and always will be, and he is right, and it is still worth saying what this particular round rewards.
What a 2030 does to a room
I want to be precise about the cost, because “hype is bad” is a consumer complaint and I am not making one.
A date announced four years out does not put pressure on the team. That comes later. What it does immediately is fix the shape of the thing in public, and public shapes are the hardest ones to change. A cinematic showing a soldier being turned to mush by Zerg sets an expectation about tone, scale, density and fidelity. Every one of those is an art and engineering budget. Three years from now, when somebody works out that the density in the trailer costs more memory than the platform has, the trailer will still be on YouTube with four million views and the question in the room will be why the game does not look like it.
We have all been on the wrong side of that meeting. Not the schedule side — the likeness side, where the work is measured against a picture made by people who never had to run it at sixty.
Box product
Then something worth admiring, from the same stage.
The StarCraft shooter’s director described what they are building as a box product — his phrase. A narrative open-world game with a beginning, a middle and a definite ending. Not seasonal. You buy it, you play it, it stops. In 2026, from a company the size of this one, that is a genuinely unfashionable thing to stand up and say, and it means a hundred downstream decisions about content budgets and asset reuse that artists will feel every day for four years.
And the counterweight to all of it came from a UK studio nobody put on a main stage. Bulkhead’s Wardogs sold over a million copies and hit three hundred and fifty thousand concurrent players on its launch day. No four-year runway, no trilogy announced at once, no confidence to demonstrate. It went on sale and people bought it.
The industry keeps telling us it is in a correction. Fine. Corrections are supposed to move money towards the things that work. When the announcement is the only artefact in the company with a deadline it cannot miss, the correction has found the wrong thing to protect.
BlizzCon announcements, dates and the “show confidence” remark as reported by VGC, GameSpot, Eurogamer and PC Gamer; Level-5’s statement from Eurogamer and VGC; Polyarc’s closure from Push Square, Rock Paper Shotgun and Eurogamer; Wardogs sales from Eurogamer.