Thanks to Hamada for edits.
Let me be clear: this is an autopsy. By the time this article was published, the live service platform fighter MultiVersus had been removed from all storefronts. This game only exists in the sense that some people, myself included, retain a digital copy we are free to play offline “for the foreseeable future.” Until my PS4 hard drive stops working, or I delete it, as I accidentally did twice while doing research for this article, or Warner Bros. deletes it without my consent. That’s well within their legal rights. It’s otherwise a lost game, something that should be unthinkable but is normal in the preservation nightmare that is the games industry.
The story of MultiVersus is a sad one. It’s not a tragedy in the traditional sense—though it did end with a real human cost—but rather a low comedy of errors. It involves a team biting off far more than they could chew and being dragged along by their prey, a corporation at war with itself, and a marketplace that has become ever more hostile to developers. And it starts, perhaps fittingly, with another, far more successful video game.
Image: Source Gaming. Smash Bros., the game that perfected how to do this kind of brand synergy machine.
There’s never been a true answer to Super Smash Bros. Nintendo’s crossover fighting game is one of a kind, an impeccable mixture of innovative gameplay, superb game feel, and perfectly weaponized nostalgia. You can approach it as a serious fighter or a casual party game, and while that’s a rare achievement in its own right, it also shines as an interactive gaming museum. Its references are deep but not inscrutable, and they’re used to craft fascinating mechanics. Each fighter feels fully formed, authentic, but they also all radiate a look and feel that helps Smash define itself as more than an IP collection. Although not made by Nintendo’s studios, the gameplay honors their decades of mechanical artistry. And the content is just massive. Leaning on the publisher’s characters, music, and art has allowed it an unbelievable size and scale. Even in a climate that encourages companies to unify their brands with tie-ins and crossovers, it stands alone.
Unsurprisingly, that has not stopped imitators. PlayStation All-Stars Battle Royale put a Sony spin on things, but its roster was uninspired and its gameplay poor. Nickelodeon All-Star Brawl worked hard to replicate the hyper-fast pace of Smash Bros. Melee, but the feel and specificity were lacking. TMNT Smash-Up was actually made by one of the Smash Bros. Brawl teams. Ubisoft’s Brawlhalla has a largely random cast and is monetized to hell and back, leading to a game whose full cost is astronomical. Indie crossovers Brawlout and Bounty Battle both whiffed the landing, though Rivals of Aether is apparently good. Generally, these concentrate on one aspect; Nick All-Star is fixated on tournament play, Brawlhalla on A-list newcomers. They never have everything. The sheer volume of material is unreplicable for almost any studio, but the juice and cohesion shouldn’t be. This all makes the “platform fighter” genre odd: it has the one big game sucking up the room, but most of the clones struggle just to distinguish themselves.
Image: Source Gaming. LeBron James is here to honor the beyond forgettable Space Jam 2. He’s up against Arya Stark, whose face-stealing copy ability makes her this game’s version of Kirby.
Player First Games had big and small shoes to fill when making their own platform fighter, one that would draw from the century-old backlog of its parent company Warner Bros. From a player’s perspective, the most intriguing thing here was the roster potential. Warner has owned DC Comics for decades along with most of its adaptations, and superheroes are a known quantity in fighting games. Their animation division has Looney Tunes, Hanna-Barbera, and Cartoon Network, a trove of cartoon history. But then there’s also the things that exist outside typical multimedia franchising, like Turner Classic Movies and, most intriguing, HBO. This could conceivably feature the antiheroes of The Wire and Deadwood alongside family friendly animation, probably the only unique thing this project could stake a claim on. The downside is that none of this would feel coherent unless you, like me, grew up watching HBO and Cartoon Network. Like every major American entertainment corporation, WB spent years gobbling up previously independent entities and never bothered to make them part of a unified brand. It’s not artistically bad in any way that Warner Bros. lacks this, but it does mean their name has less cultural penetration, and they’ve spent the past few years fixated on rectifying this “flaw.” Games like MultiVersus and cinematic excrement like Space Jam 2 exist to prove that a list of IP is actually a natural “metaverse” deserving of your time, money, and loyalty.
That’s not unique, and in reality, every big player in the entertainment industry is laser focused on this. Since Disney bought Marvel Comics, Star Wars, and 20th Century Fox, its brand has largely shifted from its animation past to these new live action blockbusters. Microsoft became the largest and most incoherent game publisher in the world through buying Bethesda and Activision, oversized conglomerates whose own intellectual properties already had no connective tissue. Before Ubisoft’s ongoing corporate hemorrhaging, it was lifting heaven and earth to make shared worlds out of its properties. Plus, you have Universal’s cross-corporate theme parks, Sony’s adorable PlayStation history showcase Astro Bot, and every shared universe across film and TV. Even the studio logos that play before any superhero or video game movie are big IP collages now. “Look at what we own!” There’s little about Warner Bros.’ attempts that is less legitimate, but theirs do have the rankest whiff of being forced.
Image: Source Gaming. The official announcement of MultiVersus‘ death on the main page. It gives you a good look at some of the character redesigns.
If we take MultiVersus on that metric, as an attempt to turn the shield on the Water Tower into something real, it does put in work. Like in Smash, everyone can fight, so Scooby-Doo’s Norville “Shaggy” Rogers can be a Street Fighter Shotaclone who throws a sandwich instead of a fireball. Also like Smash, everyone shares a cartoony look so as to not visually clash, something that’s useful when your starting roster includes noseless postapocalyptic LARPer Finn the Human, 1940s Oscar-winning mammalian rivals Tom & Jerry, a Justice League vaguely defined by the DC Animated Universe, and Arya Stark from the decidedly live action Game of Thrones. This direction applies to everyone, from Bugs Bunny to Steven Universe to the many fighters added across the game’s life, including Marvin the Martian, Gizmo from Gremlins, and Rick and Morty from, uh, Rick & Morty. Expressions are accentuated; heads are stretched and squashed. The final result feels reminiscent of the anime-inspired American cartoons from the 2000s. Less Avatar, more… Xiaolin Showdown. My beautiful boys Bugs and Steven come out of the grinder largely unscathed, but poor Batman looks like his own terrible 2000’s anime-inspired self from The Batman. At the same time, you do need to do something to make these characters feel like they weren’t poorly Photoshopped in, and where MultiVersus’ look shines is in its bouncy cartoon aesthetic. All soft curves and deep hues. Bringing everyone into this style was probably the right call, especially since it points to WB’s stellar animation division.
Of course, what matters most in a Smash clone is the gameplay, and on that front this one suffered on two ends. On the basic combat level, it was slow and oddly floaty. Characters took long to throw a punch, but those punches carried little force. When I played it for the first time, during its public Beta in 2022, I started with the intention of being a Superman main, only to find the heavyweight oddly stiff and weightless. Superman! Weightless! But Supes was at least an accessible fighter with a standard moveset, which at times felt like a rarity. Taking another cue from Nintendo’s series but in a way that misses the forest for the trees, a large number of characters were rather overcomplicated. Tom & Jerry were animated to look like they’re fighting each other in a way that’s conceptually neat but not additive, Finn used an inscrutable currency system that didn’t relate to Adventure Time in any real way, and the Iron Giant got to eclipse everyone else in size without being fun to pilot. If this is Smash, it’s a Smash where every other castmate is Steve. Many of the movesets were imaginative, but they were also needlessly complex and almost never felt “true to the roots” of the character. Instead, it was gimmickry for its own sake, and the Icons ended up far less organic, accessible, or fun than their Nintendo alternatives. Some of this can be explained by an emphasis on cooperative play—the game was built around 2v2, so a character could be overpowered in one way but still need a partner—but only some. It reminded me of fan movesets for Smash characters, the kinds that are nothing but references.
Image: Source Gaming. Gifted with flight, shields, bubbles, and more, Steven Universe triumphantly… summons a “Watermelon Steven” from a semi-obscure episode. He’s got a lot of unrelated status effects and abilities, so it feels randomly grafted on.
Back to the presentation, though. This game’s actual best idea was to exploit the TV and film history of its characters and give them dialogue—like, a lot of dialogue, and for specific allies or opponents. Tons of pre-, mid-, and post-battle quips, often by the character’s original actor. This might end up being Mark Hamill’s last time as the Joker. Interactions also aren’t limited to people from their series or who are tonally similar, so Batman’s just as likely to call Shaggy a hippie as he is to chat with Superman. I think this is the one thing it had over Smash, not because the dialogue is amazing and certainly not because it adds to the gameplay, but because it’s leveraging the thing Nintendo lacks, which is decades of movies and shows instead of plot-light games with largely silent protagonists. This also incentivized trying battles in different combinations, which gels with the cooperative focus. The enormous amount of writing, alongside pretty graphics and a surprisingly strong soundtrack (whose standouts included a remix of Steven Universe’s Crystal Gems theme and a jazzy Scooby-Doo-inspired number), made the project feel bigger, larger, and more compelling than other Smash clones.
However, the dialogue was central to another issue, that MultiVersus was plagued by leaks from the outset. Rumors about the game circulated before its reveal, exposing characters who wouldn’t be acknowledged for months. Shockingly uncut assets pulled from the code suggested that Player First had started work on content from The Avengers, Naruto, and The Lego Movie, presumably before asking and being rebuffed by those properties’ owners. The most extreme of this was when hours of those character barks were leaked after one private playtest. Alongside jokes and taunts, these lines included implicit references to dozens and dozens of potential fighters, including ones from surprising Warner Bros. assets (or “then-Warner Bros. assets” like Lord of the Rings). When Garnet from Steven Universe hails the Milestone-cum-DC teen hero Virgil Hawkins or when Velma Dinkley cracks about “Defying Gravity” in a victory line, it’s clear that Static and the Wicked Witch of the West were intended for some point down the line. It solidified the game’s breadth.
Image: Source Gaming. A few lines in the Beta referenced Raven, and she did eventually appear in the full release. Most of the newcomers from that point didn’t, however, so you wouldn’t have heard taunts mocking, uh, Jason here.
While very fun at the time, these leaks hurt the game in retrospect. They revealed ambitious plans that never fully materialized. They also meant that Player First never had control of the narrative for the game’s long public Beta, as pretty much every newcomer had been exposed. The opposite was true for the formal release; a scant few of the newcomers from the “1.0” relaunch onward had been in the leaks, and I suspect the staff abandoned their plans rather than stick by them. And though this is less important, these rumors revealed the limits to those plans. The breadth of Warner Bros.’ holdings extend wildly far, but the final roster was primarily Looney Tunes, Cartoon Network, and DC, and the leaks bore that out. HBO wasn’t represented beyond one Game of Thrones character, so no Tony Soprano or Omar Little (the leaked audio suggested two others, but they were also Game of Thrones characters). Movies fared a bit better. Beetlejuese got in to promote Beetlejuice 2. Assets for a Barbie stage were found. Discussions of this game dined out on the prospect of it being not just big, but demented, and that ebbed over time as the contours of its imagination crystallized. While the roster choices are individually solid—the only actively bad pick is Black Adam, an exhausting DC antihero added solely to promote Dwayne Johnson’s cinematic flop of an ego trip—and a few are crazy, it feels oddly safe in retrospect.
Honestly, though, once you started playing MultiVersus, the driving theme was not fun characters or corporate branding or even the gameplay. It was the monetization. Even by the standards of free-to-play live services, the economy was ugly. There were multiple currencies to confuse players, one of which was gated behind real world money and the other of which was so infrequent as to not be much better. Daily bonuses were regular and largely insubstantial. Almost all of the cosmetics weren’t just gated behind “Gleamium” but at absurd and inconsistent prices. One Batman: The Animated Series skin went for $25 dollars, probably because they felt it could. Although the fighters themselves could be unlocked without paying, you’d have to spend hours and hours for even a few. And the game inundated you in it the second you logged on, even if you played in the offline mode that unlocked everyone. The menu chugged along, almost weighed down by every outfit for the Tasmanian Devil or Harley Quinn. It was awash in notifications and absurd EXP meters for how many times you played as a character.
Image: Source Gaming. The storefront. It’s all just dross. It won’t make you better at playing the DC heroine Nubia, it won’t help you fight Gizmo or Agent Smith; it just… seems to kill a bit of the life these characters have.
Replaying the game’s final 5.0 version for this article, it’s those menus that stuck out the most. They were painfully slow and unoptimized, and it made any accidental push of the left or right bumper feel agonizing. I wanted to try more characters, see how Superman worked in this final version, but finding where to do that was a struggle. Cosmetics, now those the game put front and center, but not the basics of play. Even the oft-touted offline mode was hidden in the options. It might seem strange to focus on the UI, but it is the core of the experience. Maybe it’s unintuitive because the only real goal of the menu is to get you to spend money.
The monetization, updates, and huge backlog of potential heroes were, of course, all symptoms of live service design. This was not a premium game, instead a content farm that would endlessly produce content before its plug was eventually pulled. I’m not a fan. It’s predatory, mendacious, detrimental to a healthy artistic vision, and self-destructive. A live service can only last as long as the publisher wants to fund it, after which point it gets taken down. Maybe the game will remain available to those who downloaded it, or maybe the publisher deletes it, like what Sony did to every copy of Concord. This fate was always over MultiVersus’ head, but it was also inevitable. Live services exist in one of two states: dying—usually through delisting or a “complete” mode sold after the fact, like Animal Crossing on mobile—or staving off death. I often remember GameStop as a graveyard of promotional crap for games that never took off. Thanks to live services, the entire gaming landscape feels like that now. MultiVersus had art assets and code and some of Kevin Conroy’s last voice work, and to what extent that still exists is only via YouTube clips, fan Wikis, and a few hard drives.
Image: Source Gaming. Two strange fighter picks. Banana Guard is a classic joke character, but joke characters don’t really work when the game is struggling. Reindog, on the other hand, is an original character, and it’s hard to justify those in a corporate crossover. No one’s gonna remember him.
That being said, the game was actually quite popular when it launched in early access in mid-2022. Twenty million players, all unpaid playtesters, downloaded an explicitly unfinished version that was still hocking microtransactions. Perhaps it was the roster, or the FTP model, or maybe it boils down to releasing months after Super Smash Bros. Ultimate ended its three years of post-launch support. Either way, it got enough to even build a small competitive community. It was lauded for its dialogue, the quality of its cosmetics, and a momentum other Smash clones lacked. Where most live services try to drip feed you along, this just didn’t stop and released newcomers at a fine clip. Of course, the reason for the drip feed is because even slow is an unsustainable pace, and that caught up to Player First Games. Adding everything they had in a six month period left them with nothing, and by that point the negatives—the economics, the balancing issues, the bad game feel—were in focus. With mounting issues, a cratering player count, and months without new content, Player First took a bold step by announcing that they’d be putting the game on ice in June of 2023 before it was ready to formally release in “1.0” form. That ended up being May of 2024. No refunds for buyers with second thoughts.
Obviously this is a gamble, to turn off the tap for a year. I’m not sure who could pull it off, but it’d have to be a studio with absolute control over their product and a strong relationship with their audience. Player First had neither and went one step further by deciding, mid-production, to switch engines. They went from Unreal 4 to the newer, harder, and quite different Unreal 5. This would have entailed effectively remaking the game, in a challenging engine the team had far less experience using. While the Beta was sluggish and unsatisfying, the competitive fans with the most investment found this version even worse. Sure, 1.0 added the Joker and Agent Smith from The Matrix, but popular characters are less important when you’re dealing with a game that plays worse than it used to. The microtransactions were also pricier, the options more limited, and you could no longer play as everyone offline (that one was fixed later, if only somewhat. More on that later). The game’s standing never recovered. You’d see a news story about it adding Marceline the Vampire Queen, Raven, or Samurai Jack, all excellent choices, and it… didn’t matter. Even the only real curveball in the official release—Jason Voorhees, bizarrely, because Warner Bros. owns New Line and New Line owns Friday the 13th—fell flat. And though it was easy to miss, I did notice that the co-op format, the one the game was built around, was no longer the default. It was 1v1, the standard for fighting games. Was that a reaction to negative feedback, to jettison just a bit of their own identity?
Image: Source Gaming. The final version of the home page. There’s something sad about it, like it’s resigned to its own demise. Maybe that’s just me.
The final “Season 5” version of the game had some things the Beta lacked. It had a more talky tutorial. The suite of store pages was more voluminous. The cast was far, far larger than the initial roster. There were a few references to an overarching plot that had been teased in the Beta, a story weaving in the antagonists from The Neverending Story and the game’s one original character. But it in no way felt like a stronger or more singular game, just a bigger one. What it failed to learn from the “more is more” avenue of game design is that what that philosophy works, it’s because the glut of content and the strong core at the center empower each other. Nothing in this work managed that; its virtues were independent and never synergized.
The impending death also gave it a strange poignancy. Once Warner Bros. announced that they would be delisting the game, they removed the ability for players to spend real world money, but it was still a game defined by the drive to get players to spend real world money. The menus, the grift, the structure; these were the bones and organs of this project. Each screen was jazzed for me to buy something, even though I couldn’t with the pittance of company scrip you’re given in the Battle Pass. I described this article as an autopsy, and the experience of playing before the end felt like seeing an autopsy on a body that hasn’t yet died. In my first session, I accidentally started (and lost) an online match because, again, I think the menus really were designed to push you into engagement rather than help you. The second was just spent looking through the menus themselves, trying to see how to access the fully free part of this free-to-play game. In the third, I did find the offline Training mode and noodled around with different characters. Superman felt a bit better, but only because the hitstop was more noticeable on what few attacks I landed. Tom & Jerry really were too visually gimmicky to be fun. Everyone still felt somehow sticky and waterlogged. And while the fighters were all available in Training, it was the only offline mode that seemed to provide that, and the cosmetics didn’t come with them. These things were supposed to generate the revenue, and they’re somehow now even more gone than the rest of the game.
Image: Source Gaming. After dozens of gimmicky fighters, Samurai Jack was probably meant as a breath of fresh air. But even he is surprisingly slow.
MultiVersus’ failings must be laid at Player First’s feet. Their responses to the leaks and feedback were bad. Switching to Unreal 5 was self-sabotage of the highest order. The roster and gameplay choices were theirs, and they implemented the monetization. But they had the heinous luck to be at Warner Bros., especially as the company fell under the auspices of current CEO David Zaslav. Zaslav’s slash-and-burn tactics have disappeared decades of art, turned unfinished and finished films alike into tax write-offs, and threatened the careers of those who work there, all in the name of a growth that hasn’t happened. None of the big studios are great places to work (especially as all of them have adopted these tactics), but Warner does its best to seem the most antagonistic to its own staff. And WB culled much of its gaming division in 2025 in response to the underperformance of this and other high-profile releases. After adding Aquaman and Space Jam’s Lola Bunny, MultiVersus stopped receiving updates and was delisted at the end of May, and Player First Games, WB San Diego—who were allegedly making another WB crossover, this one a Mario Kart clone—and Shadow of Mordor’s Monolith Productions were all shuttered in February. All of these exist only in the past tense. In a way, that’s especially true of MultiVersus, as Warner Bros. is divesting from Cartoon Network as part of its split from the disastrous merger with Discovery that put Zaslav in charge in the first place. He’s never been into the cartoon side of his business. Or things that aren’t franchises. Or the name “HBO.” Or his own creative staff. You have to wonder why this reality TV mogul made the acquisition when he’s so disdainful of their products, and why he still gets to be in charge.
I was excited seeing the first leaks. Heck, I headlined a listicle about it, although I suspect it would be sad to read in retrospect and feel uncomfortable linking to it directly for that reason. Warner Bros. has published several of my favorite works of modern American art, from Steven Universe to All Star Superman to The Wire. Mixing these probably wouldn’t work in practice, and we’re already awash in crossovers, but there was real potential for something deeply unhinged in a way that can tickle my fancy. I won’t deny it’s a goal fraught with artistic pitfalls; perhaps I was simply half-remembering an old crossover ad for HBO and felt wistful. But the game itself also seemed to understand the value of high concept craziness, if only fleetingly. In this I see a sadder failure than most. A lot of talented people worked on a doomed product, and for all my love of gonzo aesthetics, their prioritizing them was never gonna be a path to success. And yet even then, that craziness was always on the fringe, in the form of characters who never appeared or potential that went ignored. The far more important pillars of game design, like game feel and balance, were somehow treated even worse. Player First pursued monetization and scale over every other aspect, never openly acknowledging where such a road leads.
Image: Source Gaming. The costumes side of the storefront. Some of them are fun references, some of them are bad references, but they were ultimately the only reason this game existed.
And doomed it was, because that is what live services are: growth until death. The model does not lend itself to strong visions or healthy work cultures. It lends itself to content, crunch, and turning as many people as possible into suckers. This was always the game’s fate. The best case scenario for MultiVersus would have been for it to get a “complete” paid version after the fact, and that scenario doesn’t exist when your publisher is axing almost everything that lacks the Batman, Harry Potter, or Game of Thrones logos. What it needed wasn’t big gets, but a production model and economy that could simply let it be a game. MultiVersus, its underperforming Warner Bros. pal Suicide Squad, Concord, Anthem, The Crew, Marvel’s Avengers, Skull and Bones, Babylon’s Fall, The Day Before, Chocobo GP, Redfall, Star Wars Hunters, Foamstars, perhaps the upcoming Marathon and just-released MindsEye; these are a plague upon our industry’s studios and programmers, pushed onto them by leadership desperate for the next industry-defining epoch. To be the next Fortnite, something that’s impossible for every game bar one and something a developer shouldn’t want to achieve. Fortnite is not kind to those who make it. That’s not a secret.
Why do games follow this at all? Well, the investor class hungers for exponential growth. Profit in the conventional sense is immaterial unless it’s both huge and constant, from year to year. Success that’s merely stable is a failure, and it will be punished by shareholders pulling out. So because Double- and Triple-A games are typically made as one-off purchases with increasingly high budgets, the profit they generate is less useful and harder to justify funding. This is why Arkane is making a Blade game, not because Blade fits immersive sims but because a licensed game or a sequel is easier to pitch to the bean counters who already see prestige games as a dead end. Why make a $60 or $70 project for dedicated machines when you drop a free game in the ocean and overcharge the whales after the fact, and on any computer or mobile device? That’s what Fortnite does, it and Apex Legends and League of Legends. Unseating the ten or so “black hole games” is the goal of almost every major publisher, because the pursuit of unsustainable growth is the only way to keep shareholders happy. And that’s the other half of why I don’t think this game had another future, because I don’t believe Warner Bros. would ever green light a paid version of MultiVersus. Zaslav and most executives seem to resent funding any art, but especially anything humble and mid-sized. It has to be either cheap and disposable or a white elephant. This was the only life it could be allowed, one that was also an instant death sentence.
Image: Source Gaming. The very first thing I saw after loading the game, a notification for people who partook in the Beta. That’s nice, but it came across as a game fully ignorant of its situation.
Will MultiVersus be an object lesson on the dangers of games as a service? Probably not, and not in any way that couldn’t be taught by Anthem or Concord. Even a Warner Bros. executive reporting on the failings of their live services to investors literally used those failures to justify making more live services. It also added little to its own genre. And while it shows the limits of a corporate crossover, that won’t be heeded at a time in which every corporation seems hellbent on building its own metaverse. Everyone’s gotta be plugged in to every damn brand and trend and franchise. In such a world, this game’s failure and death is… pedestrian. Yet another failed forever game, yet another obligatory crossover. But if it is pedestrian, it is still in a sea of madness. Alongside cryptocurrency and AI and, live service obsession is breaking the industry, and the hunt for brand singularity ain’t far behind. We are not starved for evidence of this. MultiVersus failed in almost every test. But even if it aced them all, it would have still failed. And so for god’s sake, let this one teach at least someone, anyone that this is not a path to take.
- Indie World March 3, 2026: Information and Reactions - March 3, 2026
- On Phil - February 26, 2026
- Escape from Ever After | Review - February 17, 2026
