host opinions · june 2026
The Hero-Shooter Server Graveyard: Why Most Die Before Year Two
Someone on Reddit summed up the whole genre in eleven words today: so many hero shooters failed, and yet there are more coming. They are coming because publishers keep mistaking fun for the survival metric. It never was. Server economics and a dedicated-server escape hatch were.
The pattern, in five funerals
A hero shooter does not fail the way a single-player game fails. A bad single-player game reviews badly, sells poorly, and then sits on a shelf forever, still playable, still installable, still there if you ever change your mind. A hero shooter fails by ceasing to exist. The servers go off, and the disc, the install, the launcher entry, the whole thing turns into a tombstone. You did not stop playing it. It stopped being a game.
The body count over the last two years is not subtle. Concord launched on August 23, 2024, a reported four-hundred-million-dollar Sony hero shooter, and had its servers shut down on September 6, 2024. Fourteen days. Firewalk Studios was closed within weeks. XDefiant was Ubisoft's fastest title ever to a million unique players, eleven million in its first month, and Ubisoft took the servers fully offline on June 3, 2025, leaving the game completely unplayable. MultiVersus ended online play on May 30, 2025. Anthem, never a hero shooter but a perfect specimen of the same disease, had its servers switched off on January 12, 2026, rendering a seven-year-old game unplayable in a single afternoon. And Highguard managed to launch in January 2026 and shut down in March 2026, a roughly two-month lifespan, before most people had heard of it.
Five different studios, five different budgets, five different quality tiers. One identical ending. That is not a run of bad luck. That is a structural property of the thing they all built.
Why they die the same way every time
Almost every hero shooter is three things at once, and the combination is fatal. It is live-service, so it is designed to be played continuously rather than finished. It is publisher-server-only, so the matches run on machines the publisher owns and pays for. And it ships with no dedicated-server tool, so there is no version of the game that exists without those publisher machines being switched on.
Stack those together and you get a game whose continued existence is a line item on someone's spreadsheet. As long as the population is high enough to justify the server bill, the game lives. The moment that math inverts, the rational business decision is to flip the switch. There is no archival mode, no offline build, no community fallback, because none of that was ever shipped. The product and the publisher's willingness to keep paying for it are the same object. When the willingness ends, the product ends.
This is the part that should bother anyone who cares about games as durable objects. A studio can do everything right, ship a polished, genuinely fun shooter, and still have it deleted from existence two years later because the population crossed an invisible line. Quality buys you time. It does not buy you permanence. Nothing in the publisher-server-only model can.
The matchmaking-viability floor
Here is the mechanism that does the killing, and it matters because it explains why these games die so much faster than their raw player counts suggest they should.
A matchmade shooter needs a minimum concurrent population to function: enough players online, in a given region, mode and skill bracket, to assemble a balanced lobby in a tolerable amount of time. Call it the matchmaking-viability floor. Above it, the game feels alive, queues are short, matches are fair. Below it, queues stretch, the matchmaker starts throwing lopsided or laggy games together to fill slots, and the experience curdles. And here is the trap: a worse experience drives players away, which lowers the population further, which makes the experience worse still. It is a death spiral, not a gentle slope. Once a game tips past the floor it tends to fall through it fast.
The cruelty of the floor is that it is regional and modal, not global. A game can post a healthy-looking worldwide number and still be functionally dead for a player in a smaller region, an off-peak hour, or a less-popular mode, because the floor is per-bucket, not per-game. By the time the global average looks bad, large slices of the player base have already fallen through the floor and quit weeks earlier. The published concurrent count is the last metric to flinch, which is exactly why publishers and players both get surprised by how suddenly the bottom drops out.
A dedicated-server game has no matchmaking-viability floor. Twenty people who all know each other do not need a matchmaker. They need an IP address.
The escape hatch nobody ships
The single feature that would defuse all of this is the one the genre almost never includes: a dedicated-server tool, or even a humble LAN and offline mode. A server binary the players can run themselves is an escape hatch. It means the game's survival is decoupled from the publisher's accounting. The publisher can stop paying for matchmaking and the game does not die, because the people who still want to play it can stand up their own server and keep playing with each other directly.
Hero shooters skip this for reasons that are perfectly rational from the publisher's chair and disastrous from yours. Self-hosted servers complicate monetisation, fragment the matchmaking pool the live-service model depends on, raise anti-cheat headaches, and hand players a way to keep enjoying the game without ever touching the storefront again. Every incentive points toward keeping the only servers under the publisher's roof. The result is a genre that is structurally allergic to the one feature that would let its games outlive a population dip.
So when the Reddit thread asks why so many hero shooters failed and yet more keep coming, part of the answer is that the business model that makes them so fragile is also the business model that makes them attractive to greenlight. Publishers are not building graveyards by accident. They are building them on purpose, because a walled, server-locked live-service game is a better-looking bet on a slide deck than a game players can run forever on their own hardware. Every new hero shooter announced into this crowded field is, on current evidence, a future tombstone with a launch trailer.
The games that survive a collapse
Contrast the graveyard with the games that shrug off a population collapse entirely. The reason a twenty-year-old shooter or survival game can still be played tonight is almost always the same: it shipped server files, or LAN, or both. The population can crater to a rounding error and a single community can keep a server lit indefinitely, because they were handed the means of production at purchase. The publisher walking away is an inconvenience, not an execution.
Look at the difference between the warm row and the red ones. MultiVersus is not a hero shooter, but it is the most instructive case in the list, because Warner Bros left an offline mode standing when online play ended. That is the cheapest possible gesture toward permanence, and it is still more than Concord, XDefiant, or Anthem players got. The bar for not deleting your customers' game is genuinely low. Most publishers still trip over it.
Fun was never the metric
This site keeps coming back to one filter, and the hero-shooter graveyard is the clearest demonstration of it we have. For an always-online, publisher-server-only game, fun is not the survival metric. Concord was, by several accounts, a competent shooter. XDefiant was, by its own player numbers, an enormous initial hit. Neither of those facts saved them, because neither fun nor a launch spike has any bearing on whether a game still runs after the population dips below the floor and the server bill stops penciling out.
The metrics that actually predict survival are unglamorous. Can a player run the game without the publisher's machines being on? Is there a server binary, a LAN option, an offline build, anything that exists independent of a quarterly server-cost decision? If yes, the game can survive a collapse. If no, the game's lifespan is whatever the publisher's patience and accountants allow, and it does not matter how good the gunplay is.
That is why the answer to the Reddit question is not really about whether the next hero shooter will be good. Some of them will be. Marathon, Highguard, and the rest of the incoming wave include genuinely talented studios. The question that actually matters is whether any of them ship an escape hatch. On the genre's track record, almost none will, which means most of the incoming wave is being built to join the same graveyard, however fun they are on launch day.
The honest bottom line
If you are going to play a hero shooter, play it the way you would rent a car. Enjoy it, do not get attached, and do not be shocked when you have to give it back. The genre is structurally a rental, no matter how the storefront words the purchase.
The handful worth real emotional investment are the ones that hand you the keys: server files, LAN, a real offline mode, anything that lets a community keep the lights on after the publisher leaves the building. Those games can become the kind of world you keep for a decade. The publisher-server-only ones, however polished, are borrowed time with a countdown you cannot see, ending on a date set by people who have never played them.
So yes, more hero shooters are coming. Most of them will be dead before year two, because they are being built the same way as the ones already in the ground. Judge the next one not by its trailer or its beta but by a single question: when the population dips, is there a server you can run yourself? If the answer is no, you already know how the story ends.