Promising Three Years Out — Who Actually Delivers? A Deep Dive Into Blizzard’s Long-Game Survival Play at BlizzCon 2026#

Chapter I — Closing Day: A “Vision” With No Gameplay, and a 2 a.m. Newsroom Argument#

When the lights went down at the Anaheim Convention Center, BlizzCon 2026’s closing ceremony was already forty minutes overtime. On the big screen, the last frame of Diablo V’s CG concept film froze on the hooded silhouette of the Plague Knight; the applause from the floor was sparse, punctuated by a few whistles. Immediately after, Dan Hay appeared at the tail of the StarCraft open-world trailer — a wasteland corroded by the swarm, a lone marine, the camera pulling back, and nothing more. No gameplay, no engine demo, no interactive UI, not even a single legible screenshot of how the game plays.

I sat in the last row of the media section, notebook open on my knees, with one line typed on my screen: “Two promises, zero gameplay, release windows pointing at 2029 and 2030.” Beside me, Long Zhuiri turned his phone screen around — the group chats were already on fire; someone had posted a screenshot of Blizzard’s president being interviewed after the ceremony, under the headline “delivering confidence.”

“Do you believe it?” Long Zhuiri asked me.

“Believe what? That they can finish it, or that they mean what they’re saying right now?”

He was silent for a few seconds, then said: “I’m asking whether you believe they still care what players think.”

That is not a question that can be answered with a “believe” or “don’t believe.” On the walk back to the hotel, we started an argument that ran until three in the morning. He took the veteran player’s position, I took the industry-analysis position, and neither of us convinced the other — but we agreed on one thing: Blizzard is selling “the future” itself as a product to today’s players, in a way the industry has never quite seen before.

Chapter II — Dissecting the Two Promises: What Diablo V and the New StarCraft Game Actually Committed To#

The first one: Diablo V, release window spring 2029. The official selling-point list looks plump — procedurally generated dungeons, the return of “set items,” the most launch classes in series history, and a brand-new class, the Plague Knight, making its debut.

Taken apart, those four selling points carry wildly unequal weight.

“Procedural dungeons” is an old acquaintance of the Diablo series. The mechanic was applied heavily in Diablo III’s Adventure Mode, and appears to varying degrees in Diablo IV’s dungeon system. The problem is that in the Diablo series’ historical reputation, procedural generation has always been a double-edged sword: it provides replay freshness, but it easily produces that “random yet meaningless” stitched-cave feel. In Diablo IV, Blizzard already tried “hand-tuning” procedural generation — inserting preset events and puzzles at key nodes. So is Diablo V pushing the mechanic to a deeper level, or just giving Diablo IV’s dungeon generator a new skin? The studio didn’t say.

The return of “set items” is an even subtler signal. Diablo III’s set system propped up the entire seasonal endgame at its peak, but in Diablo IV the status of sets was sharply reduced, replaced by unique gear and the aspect system. When Blizzard now says “return,” does that mean back to Diablo III’s six-piece era, or back to the classical Diablo II mode of Tal Rasha and Immortal King — few pieces, powerful effects? With just the word “return” and no details whatsoever, the selling point itself is a blank check.

“The most launch classes in series history” — what does “most in series history” mean, concretely? Diablo II launched with five classes, Diablo III with five, Diablo IV with five. If Diablo V wants “most,” it starts at a minimum of six. The studio gave no number — only “most.” That wording itself leaves room to maneuver: if it ultimately ships with six, that is indeed more than five; if it ships with five, “most in series history” becomes an out-and-out lie.

The Plague Knight is the only new class with a concrete image. In the CG concept film, he wears gray-black chainmail, rotted skin faintly visible under the visor, a long spear in hand, green-black plague mist spreading beneath his feet. The design visibly fuses the Necromancer’s “disease” theme with the Paladin’s heavy-armor-melee position. And once again: no skill showcase, no talent-tree screenshot, no numerical framework — just one concept render and a name.

Now the second: the StarCraft open-world shooter, release window 2030. It is helmed by Dan Hay, former head of the Far Cry series, and emphasizes “building a world with a sense of authenticity.”

The project’s key information consists of exactly two things: Dan Hay’s name, and the phrase “authentic world.”

Dan Hay’s track record at Ubisoft is genuinely impressive — Far Cry 5 and Far Cry New Dawn were both his projects, and open-world structural design and regional pacing are his strengths. But the phrase “authentic world” has been severely devalued in this industry — nearly every open-world game wraps itself in “authentic,” “believable,” “organic” at reveal, only to deliver, in the end, a map bristling with question marks.

Doing an open-world shooter with the StarCraft IP is, in itself, a massive gamble. StarCraft’s setting is a highly militarized, interstellar-scale war narrative, while the core experience of open-world shooters tends to be “free exploration + zone liberation + outpost assault.” Whether natural narrative tension exists between the two is unproven — Blizzard offered no gameplay footage to demonstrate that the combination works.

I organized the two projects’ key information into a table:

The two promises share one defining trait: CG concepts only, no gameplay demo of any kind. That means players cannot judge whether the core gameplay holds, whether the art direction is coherent, whether the technical architecture is feasible. Everything sits at the “concept” level — and concepts are the cheapest part of game development: they need no engine, no code, no art assets, only a storyboard good enough and a storyteller skilled enough.

Chapter III — The “Delivery Confidence” Theory: What the Blizzard President Said, and What He Didn’t#

Interviewed by media after the closing ceremony, Blizzard’s president, facing the challenge of “why announce so early, with no gameplay footage,” gave this explanation: this is “signaling delivery confidence” — giving players and capital markets reason to believe these projects genuinely exist and the company is executing to plan.

The sentence deserves to be taken apart word by word.

“Signaling to the market” — who is the market? Players are the market; investors are also the market. But the wording implies a priority ordering: if the primary aim were to reassure players, you would say “signaling to players”; if the primary aim were to reassure capital markets, you would say “signaling to investors.” “The market” is a vague collective noun that covers players and capital at once — and that vagueness is itself a strategy: it lets Blizzard flexibly re-aim the interpretation of the sentence depending on the occasion and the audience.

“Delivery confidence” — delivering what? Games, or “confidence” itself? This is a rhetorical slippage. If Blizzard said “we will deliver a complete Diablo V in 2029,” that would be a concrete, verifiable commitment. But “we signal delivery confidence to the market” shifts the object of the promise to “confidence” itself — as long as the market believes, the promise is fulfilled, even if the game ultimately slips to 2031.

That is the real commercial meaning of “delivery confidence”: it swaps the object of the promise from “product” to “trust.” Once trust becomes the deliverable, then as long as trust is kept alive, delivery is permanently “complete.” It is a closed loop — a closed loop that never needs a product to verify it.

What Blizzard did not say: why does transmitting confidence require “announcing three years early”? If the projects are genuinely on plan, why not wait until there is gameplay footage? The answer is actually very simple — gameplay footage requires development progress, and development progress requires time. If Blizzard has no presentable gameplay content today, these projects are at a very early stage. At such a stage, the only “confidence” that can be transmitted is “we do have these projects greenlit” — and that sentence, on its own, could be proven by an internal memo; it did not need a closing ceremony.

There is a subtler detail here worth savoring. On closing day, Blizzard placed the two heaviest “vision-tier” titles in the finale, rather than in the main announcement block. That scheduling itself manufactures a “closing-day afterglow” — it keeps media and player discussion fixed on “Diablo V and the StarCraft game” into the second and third day, instead of drifting to competitors’ showcases. In other words, these two gameplay-less titles serve not a “product showcase” function but a “monopolize the conversation” function. The more they resist being seen through, the more they get discussed — and the discussion itself becomes the best marketing.

I pointed this out to Long Zhuiri. He was silent a moment, then said: “So I call it pie-in-the-sky, and it tells me the pie is being discussed quite enthusiastically, thank you very much.” The remark was half self-mocking, but it cuts to another face of the “delivery confidence” theory: when a title’s actual content cannot sustain discussion, vagueness itself becomes the tool for staying hot.

Chapter IV — Industry View: Distant Promises, Slate Economics, and Expectations Management#

Seen through industry mechanics, Blizzard’s move is not an outlier. The whole game industry is undergoing a “slate-ification” transition — large publishers increasingly resemble streaming platforms, sustaining market heat with “three years of future product lines” rather than proving their value with “what’s on sale today.”

The economic logic of this model is not hard to grasp. Development cycles keep stretching; a AAA title now commonly takes four to six years from greenlight to launch. If a company has only one product on sale, its revenue and stock price face huge swings across the cycle. But if a company holds three to five “announced, unreleased” projects, capital markets can use those projects to construct a “future growth curve” and sustain valuation at the current stage.

That is the core logic of slate economics: expectation itself is an asset.

In financial markets, expectations management is a mature discipline. Listed companies transmit future-earnings expectations through guidance, executive statements and periodic reports, thereby moving the stock. Game companies do exactly the same thing — only the carriers of their expectations are trailers and release windows instead of earnings reports.

By announcing 2029 and 2030 projects in 2026, Blizzard is essentially issuing a “long-dated guidance.” It promises no specific product quality — only “these projects exist.” And “existence,” for capital markets, is already enough to support valuations for a stretch of time.

The weakness, however, is fatal: slate economics requires continuous delivery to keep expectations credible. If a company announces a 2029 project in 2026, then announces in 2028 that it slipped to 2031, capital markets will not keep paying a valuation premium for “the project exists” — they will instead impose a punitive repricing for “expectation failure.”

Blizzard’s historical record here is not encouraging. We will dissect it in detail later, but hold onto one fact first: over the past decade, almost no major Blizzard game shipped on the release window it was first given.

To sharpen the industry logic, we can contrast “slate economics” with the older “trade-show reveal” tradition. Ten years ago, announcing a new game looked more like a product launch: “I’ve finished it, or nearly finished it — here it is now.” Today’s announcements look more like a fundraising roadshow: “I’m making it — please believe I’m making it.” The difference shows up in the psychology of watching a showcase: the former makes you expect “playable soon”; the latter makes you read “this is a performance staged for capital markets.”

Long Zhuiri pushed the critique a step further: “If expectation itself is an asset, who holds its pricing power? A player who sees a trailer in the off-season can choose to believe, or choose to doubt. Capital markets are different — it has concrete valuation models, analysts, quarterly reporting dates; it examines ’expectation’ inside a framework that can be quantified. In other words, the same promise is precisely priced in the capital market, while remaining a blurry emotion in a player’s heart. When players are full of hope and capital markets have already begun pricing that hope, a gap opens between the two — and the ones who ultimately pay for that gap are, more often than not, the players.”

That passage exposes a problem many people overlook: an overpromise is not the same object to capital as it is to players. Capital values the maintenance of “certainty”; players value the delivery of “gain.” When Blizzard says “signaling delivery confidence to the market,” what it promises capital markets is a continuously refreshable “projects still on track” signal; what it promises players is a distant, gameplay-less “future title.” The same act carries, by nature, different standards of delivery for the two audiences — and that is exactly where the controversy over “cashing in anticipation early” is rooted.

Chapter V — The Veteran’s View: Spent Nostalgia and Trust Inflation#

Long Zhuiri sent me a long message at two in the morning. I am reprinting it verbatim:

“I’ve played Blizzard games for twenty years. From Tristram in Diablo II, to Warcraft III’s Frozen Throne, to StarCraft II’s Heart of the Swarm. I lived through Blizzard’s most glorious years, and its worst. But this time is different.

Blizzard announcements used to come with something to look at. World of Warcraft had a gameplay demo at announcement. Diablo III had one. StarCraft II had one too. You can argue those demos were never fully honored in the end — but at the moment of announcement, Blizzard had something in hand.

And now? Diablo V is one CG. The StarCraft game is one CG. Blizzard is telling us: ‘We’re making these things, but you don’t get to see them.’ It’s like someone promising to repay you in three years who won’t even take out his wallet for you to glance at.

I understand the business logic. I understand Blizzard needs to hold up its stock price and convince investors it still has a future. But players are not investors. What players put in is trust, expectation, time. When Blizzard sells ’the future’ to capital markets, it is simultaneously overdrawing player trust.

Trust works like inflation — the more you print, the less each note buys. Every ‘vision-tier’ project Blizzard announces now dilutes player trust in Blizzard’s promises. By the day Blizzard actually ships something worth anticipating, players may already have stopped believing it.

This is ‘crying wolf.’ But worse than crying wolf: Blizzard isn’t crying once — it cries every year or two, and moves the date each time it cries. Players are being trained into numbness, into taking Blizzard announcements less and less seriously. And Blizzard itself is dismantling, with its own hands, the most valuable asset it ever had — player trust in it.”

I read the message and did not reply right away. I knew every sentence Long Zhuiri wrote had merit — and I also knew that from the industry seat, Blizzard may have had no other choice.

Chapter VI — Diablo V’s “Procedural Generation” and “Set Return”: an IP at War With Itself#

In Diablo V’s list of selling points, “procedural dungeons” and “set item return” are the two phrases most worth excavating, because they point straight at the deepest controversies in the Diablo IP’s history.

Procedural dungeons in the Diablo series trace back to Diablo III’s Adventure Mode. The “bounty hunts + Nephalem Rifts” system propped up Diablo III’s replay value with random maps and random events. But Diablo III’s generation carried a long-criticized flaw: the patchwork feel was too strong — players could plainly feel “this is a randomly assembled cave, not a designed dungeon.” Diablo IV tried a hybrid of “hand design + procedural fill” to correct the problem, with mixed results — some dungeons genuinely gained better pacing, while others still felt hollow.

For Diablo V to make a real breakthrough in procedural generation, it must solve a fundamental design problem: how do you give randomly generated maps the narrative density that only hand-crafted maps possess? No game has yet produced a perfect answer. Path of Exile’s map system substitutes modifier combinations for map design — but that is not really “generation,” it is “permutation.” Diablo IV’s “dungeon key” system likewise strikes a balance between “preset dungeons” and “random modifiers,” rather than true procedural generation.

If Diablo V really intends a fully upgraded version of “procedural dungeons,” the engineering and design investment it requires is enormous. It needs a generation framework that understands “narrative pacing,” “difficulty curves” and “reward density,” not one that simply randomly places rooms and monsters. If that technology can be achieved by 2029, it is genuinely revolutionary; if it cannot, Diablo V’s dungeon experience could regress to Diablo III’s early “random cave” level.

“Set item return” equally demands careful reading. Diablo III’s set system became the core of build-crafting in later versions, but on a framework of “six-piece sets + specific skill bonuses.” Diablo IV deliberately weakened sets after launch, elevating unique gear and the aspect system instead — which many players read as a correction of Diablo III’s “set monoculture.”

Now Diablo V says sets “return.” Which direction is that return? Back to Diablo III’s six-piece model, and Diablo IV’s aspect system was wasted work; back to Diablo II’s “few pieces, strong effects” model, and set-driven build depth may be too thin to carry long-term seasonal play. Blizzard has offered no details — and that is precisely the problem: “return” is a fuzzy promise. It lets each player imagine the version they prefer, and at delivery, some share of those players is bound to be disappointed.

Add the Plague Knight, and if Diablo V’s launch roster really is “most in series history” as officially claimed, that means at least six classes. That means Blizzard must deliver six complete skill trees, talent systems, itemization and balance tuning at launch — a colossal workload, especially after Diablo IV’s balance patches exposed the problem of “excessive gaps between classes.” If one or two of Diablo V’s six launch classes turn out conspicuously weak, “most classes in series history” becomes the joke “most unbalanced classes in series history.”

Chapter VII — Timeline and Risk Ledger: Blizzard’s Historical Delay Account#

Now we put the two projects’ release windows alongside Blizzard’s historical delay record, for a risk comparison.

What deserves attention in this table is not the ubiquity of delay, but the distribution of delay magnitude. Blizzard’s large projects average roughly two years of slippage from announcement to launch, and both Diablo III and Diablo IV ran about four years.

Now, Diablo V’s announced window is spring 2029. Extrapolated by Blizzard’s historical average delay, the actual release could fall between 2030 and 2031. Extrapolated by the “four-year delay” extreme cases of Diablo III and Diablo IV, Diablo V’s actual release could land between 2032 and 2033.

The StarCraft open-world shooter’s window is 2030. Given the genre’s development cycle — open-world shooters typically need five to seven years of development — the 2030 window is already extremely tight on its own. Add Blizzard’s historical delay rate, and the actual release could fall in 2032 or later.

Risk comparison table:

And one more risk that cannot be ignored: cancellation. Over the past decade, multiple announced Blizzard projects were ultimately canceled — the most famous being the Warcraft mobile game (a Project Titan spin-off) and StarCraft: Ghost. If Diablo V or the StarCraft game hits a major direction change or a technical wall, Blizzard could well announce “the project is canceled” at some point — after all, one of the biggest perks of announcing three years early is that you can cancel three years early, without ever facing the after-sales pressure of a shipped product.

Chapter VIII — Two Views in Collision: Commercial Logic Meets Trust Logic#

Back to the scene of the 3 a.m. argument. Long Zhuiri and I sat facing each other, two note-covered sheets and a cup of cold coffee on the table.

“You’ve laid out all this industry logic,” Long Zhuiri opened, “but you haven’t answered one question: what Blizzard is doing — is it good or bad for players?”

“From the industry view, it’s a neutral act,” I said. “It’s a rational choice made under current market conditions. If Blizzard didn’t announce these projects early, capital markets would question its future growth, the stock would come under pressure, and that would feed back into its R&D spending. Announcing early is a defensive strategy.”

“Defensive?” Long Zhuiri snorted. “You talk as if Blizzard had no choice. It could choose not to go public, could choose not to chase quarterly earnings, could choose — the way it once did — to announce only when the game is done. It made this choice because it cares more about capital markets’ reaction than about players’ feelings.”

“But player feelings don’t put food on the table. Blizzard needs funding to sustain R&D, needs a stock price to sustain employee confidence and paychecks. If a stock collapse forced Blizzard into layoffs and project cuts, that would be the real disaster for players. Which do you want: Blizzard canceling Diablo V under stock pressure, or Blizzard announcing Diablo V early to hold up the stock and keep the project alive?”

“You’re justifying the means with a hypothetical end,” Long Zhuiri said. “If Blizzard ultimately does deliver Diablo V, the early announcement can indeed be read as ‘a compromise made to save the project.’ But if Blizzard announces a slip in 2029, or a cancellation in 2031, the early announcement becomes an outright con. You cannot know the outcome now — so you cannot defend today’s bad behavior with tomorrow’s possible good outcome.”

“So you’re saying Blizzard should announce nothing at all until development is complete?”

“I’m saying Blizzard should announce when there is something to show. It doesn’t need to wait for completion — but it at least needs one piece of gameplay footage to prove ’this project is real.’ In Diablo V’s CG concept film, the Plague Knight’s hood ripples in the wind, and it looks beautiful — and it proves nothing. It proves neither that procedural dungeons work, nor that sets are balanced, nor that the Plague Knight’s skills are fun.”

“But this is the industry standard. Nearly every AAA game opens with CG-only reveals. Sony’s first The Last of Us was CG-only at reveal; Nintendo’s The Legend of Zelda: Breath of the Wild was concept footage at reveal. You can’t single out Blizzard.”

“You’re right, other companies do it too. But other companies don’t carry Blizzard’s delay record. When other companies post CG, players think ’that game looks great.’ When Blizzard posts CG, players think ‘when does this one get cut?’ That is the price of trust inflation — Blizzard overdrew its own credit, and now everything it says gets discounted.”

I looked at Long Zhuiri. His expression was calm, but I knew the emotion underneath was not. He was right: Blizzard’s historical record has forfeited its claim to the trust dividend of “industry standard practice.” When a company has repeatedly delayed, canceled projects, and delivered below expectations, its “standard practice” no longer reads as “standard practice” — it reads as “here we go again.”

Chapter IX — Closing: Can the Long Game Hold, and What Should Players Do#

Back to the original question: can the model of “promising three years early and surviving on the long game” actually work?

From industry mechanics, it works in the short term. By announcing distant projects early, Blizzard successfully signaled “this company has a future” to capital markets, holding up the stock and investor confidence. As long as capital markets keep believing “these projects exist,” Blizzard gets stable valuation support across the development stretch. As expectations management goes, it is a successful strategy.

From a long-term view, however, the model carries a fundamental contradiction: it needs continuous delivery to stay credible, and Blizzard’s historical record shows its delivery capacity is very limited. If Diablo V slips from 2029 to 2031, and the StarCraft game slips from 2030 to 2033, capital markets will discount Blizzard’s “long-dated guidance” — and the next time Blizzard announces a 2032 project, the market will no longer pay the same valuation premium. It is a spiral of trust decay, and Blizzard is accelerating down that spiral.

For players, my advice is simple: treat every release window Blizzard announces as “the earliest it could possibly ship,” not “the date it will ship.” Don’t invest any emotional expectation in a 2029 Diablo V; don’t plan any game time around a 2030 StarCraft. Wait until Blizzard actually shows gameplay footage and announces a believable release window — then start anticipating.

At the end of our argument, Long Zhuiri said one thing, and I have made it this article’s closing line:

“I still hope Blizzard can make great games. I hope Diablo V’s procedural dungeons genuinely surprise me, I hope the Plague Knight genuinely becomes the Diablo series’ most interesting new class, I hope Dan Hay’s StarCraft open world genuinely makes me feel that ‘authenticity.’ But I’m no longer paying for those hopes in advance. Blizzard wants me to believe it three years early; I’ve decided to believe it the day it shows evidence.”

I agree with him. Blizzard can promise three years out, but players don’t have to pay three years early.

Editorial Afterparty#

Having finished this piece, I owe a confession: I did not fully convince Long Zhuiri, and he did not fully convince me. We each laid out our position clearly, then agreed to return to this table when the first gameplay footage of Diablo V or the StarCraft game drops — and see who was right.

I love Long Zhuiri’s last line. He said: “I’ve seen Blizzard at its most glorious, so I can’t bear to watch it survive on promises — but I can’t bear to watch it starve either. I hope it can make one more game that would make me willing to believe in advance.” He stubbed his cigarette out on the rim of the cup, then added: “Just don’t make me sit through another four of these four-year waits.”

As an industry observer, I accept that overpromising is commercially necessary — but I believe more firmly in that other line: expectation is an asset, but when it cannot be delivered, it becomes a liability. Blizzard’s ledger gets settled by what actually ships in 2029 and 2030. And until then, the only thing we can do is stay clear-eyed: neither cheering early, nor sentencing early.

— Chasun and Long Zhuiri, NTGame Editorial Desk