Punched out: EA's ill-fated Marvel Comics fighter

How boredom and overconfidence led EA's Chicago studio to ditch its cash cow and gamble its future on an ambitious — but directionless — superhero game.

Jump to

One word tends to pop up more often than others when speaking with former EA Chicago employees: innovation. While a common proclamation and goal for development teams large and small, at Electronic Arts' former Midwest studio, employees say it was a mantra and a driving force behind all projects.

It worked brilliantly with 2006's Fight Night Round 3, which debuted shortly after the Xbox 360 and impressed many with best in class visuals and right analog stick punching controls. But that was the third iteration of a series that had already established its fundamentals.

Def Jam: Icon proved a different story. Casting aside the elements of previous, well-regarded entries developed elsewhere, the studio's 2007 take offered a bold aesthetic and environments that reacted to the beats of its hip-hop soundtrack — but faltered with sluggish, awkward fighting mechanics.

EA Chicago's unfinished Marvel Comics fighter embodied that mantra through and through. Attempting to recast iconic heroes, create fully destructible environments and deliver a constant variable in the form of civilian crowds, its incredible scope had production crawling at a turgid pace; former employees claim that a lack of leadership and defined direction kept the team from matching the ambitious vision. When the studio shuttered in November 2007, the game went with it, despite more than a year and a half of development.

Marvel (as it was simply called within the studio) may not have singlehandedly sunk EA Chicago; former employees note that lavish spending on a new downtown facility, rapid growth and Icon's poor performance were also likely factors. But the project's languid development didn't help the studio's chances when Electronic Arts began polishing its axe to appease shareholders. And it all began with one significant trade.

"Marvel" teaser trailer

The Canadian swap

Known as NuFX before Electronic Arts acquired it in early 2004, EA Chicago had put its stamp on the popular NBA Street franchise before revitalizing EA's tired Knockout Kings brand with Fight Night 2004 and Round 2. Fight Night Round 3 was an immediate sensation, thanks to its pairing of fluid analog punching controls and incredible slow-motion punch animations, which shuddered virtual cheeks in the game and likewise dropped jaws in real life.

It made a lot of business sense for EA Chicago to continue with the franchise, considering the immense critical and commercial performance of Round 3, but after releasing three iterations in as many years, the team was tired of simulated pugilism. When development wrapped at the end of 2005 following an intense crunch, the studio started discussing other options with EA corporate.

Fight Night Round 3
Fight Night Round 3

Marvel Nemesis: Rise of the Imperfects from Nihilistic Software and EA Canada had recently launched to a middling response, and EA Chicago President Kudo Tsunoda and team made a push to take over the license and make a fighter for Xbox 360 and PlayStation 3. However, EA Chicago couldn't accommodate three AAA projects with the personnel it had, and Def Jam: Icon was already well along in production. As such, Fight Night was sent away in favor of embarking on something new. "It was essentially a trade, and we gave Fight Night to EA Canada to do," says Producer Alan Martin.

"I don't know if I would have shipped Fight Night for Marvel," says Michael Mendheim, who was the executive producer for Def Jam: Icon before briefly doing design work on Marvel. "Fight Night is kind of the golden goose, and Kudo and the dev team turned a so-so, bland product into a powerhouse sports title. As a studio that is a satellite of EA, it's very risky to give away the goose that's laying the golden eggs to do something that may be more inspiring."

While much of the studio worked to complete Icon, EA Chicago put a team of about eight to work on conceptualizing Marvel's direction. For Martin and Mendheim, it was another chance at tackling a Marvel Comics game. A decade prior, both had worked on Ashes of the Apocalypse, a PC action RPG that was supposed to serve as a sequel to the "Age of Apocalypse" comic arc in the X-Men universe. Their team at New Wave Entertainment worked with Marvel on the game for months of pre-production, and it got far enough to yield a two-page, promise-filled advertorial in Marvel's comics at the time — but then was canceled.

"I don't know if I would have shipped Fight Night for Marvel."

Pre-production on EA Chicago's Marvel fighter, on the other hand, initially focused on creation of an "X Video" — a common practice in which a team generates a cinematic to show the overall aesthetic and planned concept for a game. "It almost acts like a commercial internally for the entire team to get behind," says Josh Tsui, who was EA Chicago's art director and now heads up developer Robomodo. As the concept clip shows, the fighter was meant to spotlight iconic characters battling it out in large, highly destructible settings, with panicked citizens nearby to interact with.

Typically, the X Video for a project took a few months to envision and execute, with the goal of pushing right into the tech side of things immediately thereafter. As a former employee who asked not to be named tells Polygon, the X Video process for Marvel spanned more than six months, in part because Def Jam: Icon was delayed and still required plenty of attention, so engineers came onto the superhero project much later than expected.

By the time the team ramped up to about 70 people, with Def Jam: Icon winding down, the Marvel pre-production team had been spinning its wheels for a bit, and the project was already behind schedule with little to show for the time spent.

Marvel, modified

One early hurdle had been EA Chicago's attempts to put a fresh spin on classic comic characters. Producer Mike Kennedy recalls the artwork being "progressive," but according at least one former employee speaking anonymously felt the team was losing the heroes' iconic qualities in a bid to make them seem more grounded in reality.

"When you see Spider-Man in the comics, he's pretty damn muscular. So the idea was that Peter Parker is an awkward teenage kid — let's make him a really skinny teenage kid. Early on, what ended up happening was that we had a Spider-Man who almost looked anorexic," the source says. "As a 3D model, you [had] a Hulk that ends up looking like a really big, fat guy."

Wolverine concept art shows EA Chicago's somewhat "progressive" approach to the Marvel universe.

Doctor Doom's interpretation was way more out there; our source says he "quite literally looked like a tank," and adds, "When you see something that is that completely different without context, it just breaks down." Team members were urged to go for broke, and not fear potential blowback from the comic publisher. "We were being told, 'Hey, don't worry about Marvel right now — the idea will be so cool that they'll totally use it,'" says the source.

"You get diamonds from friction."

But Marvel Comics pushed back; the first wave of designs deviated so much from the well-known comic appearances that the publisher demanded changes. What we see in the screens and concept art are the revised costumes, which skew closer to the source material while still featuring a distinctive edge in parts. As Mendheim sees it, though, the back-and-forth was essential for nudging the aesthetic forward and doing something unique that still respected the property. "You get diamonds from friction," he says.

One area where Marvel didn't seem to mind a fresh slant was the voices of the comic characters. "In some cases, it's not who you would expect to hear on a Marvel Saturday morning cartoon," says Kennedy, who notes that all of the fighter voice acting was recorded by the time the studio closed. "We were trying to make the dialogue a bit more anachronistic."

Filling the role of Spider-Man was David Faustino, best known for the role of Bud Bundy on Married... with Children. "He did a great Peter Parker. He had that same kind of cocky, cracking jokes kind of attitude, but it wasn't the Spider-Man you were expecting to hear," says Kennedy. Meanwhile, Steve Blum (Spike Spiegel from Cowboy Bebop) played Captain America like a "hard-nosed drill sergeant," says the producer, while Wolverine — voiced by Gregg Henry (Payback) — embodied a "Louisiana suave attitude." As Kennedy says, "We weren't shying away from redefining them a little bit, but nothing so far as to make them unrecognizable."

City strife

As shown in the X Video footage, one of the Marvel fighter's biggest planned selling points was the ability for players to demolish their surroundings. From the start, the team's goal was to have every building be procedurally destructible — with opponents thrown through walls and individual floors of structures — plus elevated train tracks that could be trashed and much more. Despite the lack of a formal announced title or codename, project UI Lead Todd Omotani says the rumor was that the final release would be named "Marvel: Destruction."

Creating heavily destructible environments was a time-consuming process, however. It's telling that all of the screenshots and leaked video clips are set on the seemingly Chicago-inspired, urban Midtown stage: It was the only working environment that was completed by the time the studio closed its doors. Additional planned stages were grey-boxed out using placeholder public domain models, but only Midtown was ready to show.

Marvel destruction CG screengrab
This screengrab from the Marvel team's X Video shows the game's intended environmental destruction.

"It was tough to define our environments because we had characters with varying abilities ... and how they were going to interact wasn't completely pinned down. Trying to design environments around that, with the size limitations and the destruction built into it, was very tricky," explains Martin, who was the environmental producer for the game. "It was a fight against performance all the time. It was how big the environment could be, how much destruction we could do, plus the effects hit on performance really limited us."

Coming off of Fight Night, the team remained sold on analog-centric fighting, with movements of the sticks generating punches for the Hulk, or specialized attacks for other characters. However, the sheer variety of abilities and fighting techniques threw the team for a loop and made it difficult to customize the movements for each distinct style. "When it worked, it worked really well," claims Kennedy. "When the Hulk did an uppercut and sent Captain America flying into the air, it felt pretty good."

One of the biggest issues with the fighting action was finding ways to keep players close enough for things to stay interesting. "When you are in a big play environment, and say you have the Thing over here and Beast over there, and they're two blocks away — they're just kind of running towards each other on a street," says Mendheim, who departed the studio early into full production. "It wasn't very compelling."

"That was the part we didn't quite figure out the mechanic of yet, or hadn't gotten quite down to where it needed to be," says Martin, regarding the state of the action near the project's cancelation. "In a boxing ring, fighters are forced together because of the proximity of the ring. But in [Marvel], if you're in a big-sized chunk of the city, it's much harder to keep that together."

Gregory Allen, audio director for EA Chicago, says that the scale of the experience kept shrinking over time. Stages initially conceived to have an open-world feel were condensed to about five city blocks, and only certain aspects of the settings could be completely destroyed. "The size of the thing was just enormous," says Martin. "We had to scope it down, but some of the big ideas didn't make it, and some lost their big-ness."

Crowd control

Among those big ideas, one of the most trying, pie-in-the-sky concepts was that of the "living city" — meaning you weren't commanding Wolverine or Juggernaut in some vacant metropolis, but rather in a setting full of citizens whose daily routines could be dramatically affected by your presence and actions.

At the concept level, civilians were affected by your direct or indirect actions, whether it was attacking a bystander or failing to save one in imminent danger, plus traffic would bunch up and trains would fly off the rails should track segments be destroyed. Moreover, players could sway nearby civilians to your cause to turn them against the opponent. "It was very, very crazy ambitious design," asserts Mendheim. "You were throwing ripples into the system."

For example, if you saw a citizen in harm's way, saving him might encourage close bystanders to start rooting for you, and even to attack your opponent, potentially affecting the fight. However, reckless destruction or intentional attacks towards citizens would turn the city's population against you. Cops might start shooting at you as a result, and there were plans to have heavy-duty military come in to assist, but the concept was cut by the team as the project continued.


One in-progress screenshot obtained by Polygon shows two Hulks fighting, one of whom is standing on a crane hanging over the city. A developer caption written on the screen says, "Crane operator trying to shake the Hulk off would rule!" Clearly, the team had big ideas for how the public could sway the tide of battle, but it was one of those plans that sounded great on paper and just didn't work as well in the game. Martin concedes that the team couldn't fit in as many bystanders as planned, and that their detail was lacking — plus the concept required a complex artificial intelligence system.

"The concept by itself is a great idea, but trying to lump it in with some of the other ideas, it just became mush," says Tsui. "The problem is that the AI for something like that is so complicated. Fighting games are so fast that any distraction that happens between you and your opponent can end up being a nuisance."

And the time-consuming concept proved divisive among the staff. "In the beginning, when we're doing all this research on crowd stuff," says Allen, "it's like, 'Are we making a crowd game, or are we making a fighting game?'"

Lacking leadership

Despite the sarcastic tone of that comment, some team members felt the project lacked direction. In particular, former members of the team note a strong absence of leadership, and of people willing to make decisions to limit the scope of the project. "After Fight Night, people were overconfident on things, and we weren't going through proper checks and balances to design a game," says one source.

Part of the issue, we're told, had to do with an inexperienced hand running the ship. EA Chicago named Darren Bennett to the roles of executive producer and creative director on the title, though he had previously only been an art director on the studio's games. "You need to put a producer who's actually produced products before, that knows about budgets and can organize this stuff," says Allen, who opines that more senior employees were passed up for the role. "It's not just about sheer will to be able to do things. There is a reality check." (Polygon contacted Bennett to comment for this story, but did not hear back.)

Producers were encouraged to bring ideas to the table even well into development, and concepts were rarely shot down, with nobody putting their foot down as needed, according to team members.

"Comic books are all about story, so I was a little surprised [we didn't have much of one]."

"The scope hadn't been narrowed early enough in the project," says Martin. "Producers at EA Chicago were both designers and project managers, so it was our job to keep it in check, as well, definitely. Some of the people that really needed to be on top of that and make those decisions let them go on too long." He adds, "At some point, somebody's got to go, 'OK, we've got our ideas and we need to make decisions, and we need to lock things down.' Otherwise, you're in this eternal waterfall."

Allen cites what he calls "a very loose plan in place" for completing the game, and says EA corporate was breathing down the necks of EA Chicago heads as production dragged. Echoing that theme of overconfidence expressed by one source, Allen says, "Some people, because they've pulled off miracles before, always think that they can pull off miracles, and we'll just figure it out as we go."

According to some, not only did the game lack direction in some respects, but also purpose. Opting to focus on the one-on-one combat and destruction, the game offered no story mode or surface-level narrative posturing. "Comic books are all about story, so I was a little surprised," says Kennedy, who has written numerous comics and now is the publisher at graphic novel company Archaia.

He says the team discussed ways to loop the game in with the "Planet Hulk" comic arc, or a Skrulls invasion, but ultimately opted not to. "It was just decided and agreed all the way at the Marvel Comics level that we'd just do our own thing, aesthetically," he says. Instead, the game took inspiration from classic arcade ladders of enemies, but aimed to offer a more dynamic approach using a branching hub of potential fights.

In the last months of development, though, Kennedy says the "fighting was starting to get fun," but that ultimately, "there still needed to be a motivation for players to keep going from fight to fight, and that wasn't there yet."

Marvelous destruction

On November 6, 2007, Electronic Arts closed EA Chicago, and canceled the game along with it.

At the time, EA Games President Frank Gibeau sent a memo to all employees calling it the "toughest decision I've made in my career," and not reflective of the studio's talent. He claimed a willingness to invest in people and ideas between releases, but maintained that studios must maintain profitability. "Unfortunately, EA Chicago hasn't been able to meet that standard," read the memo. "As it stands, EA Chicago has no expectation of hitting our profitability targets until [fiscal year] 2011 or later."

EA Chicago had been bleeding money. Early in the year, the studio relocated from suburban Hoffman Estates to a posh downtown Chicago location, with the publisher signing a long-term lease for the studio and retrofitting it with plenty of modern flair. As Gibeau's memo noted, the staff tripled in size over three years, and all told, the overhead to run the studio was tremendous. One source speaks of outlandish office furniture and a rooftop deck — unnecessary perks that pushed the outlook deeper into the red.

According to Allen, who was part of EA Chicago's executive staff, Electronic Arts corporate had asked for a three-year plan for the studio. EA reportedly came back and said that EA Chicago would also have to develop a Wii version of Marvel — in addition to the planned Xbox 360 and PlayStation 3 versions — and have all three ready by Christmas 2008. The game was still far from alpha status, and another year of development seemed like a very conservative estimate of what was needed to complete the project. Team members were flabbergasted by the Wii request. "We were going to kill ourselves to even get these [existing SKUs] released on time," says Allen of their reaction.

Marvel UI
The above mockup shows what the team was planning for Marvel's menu screen.

Def Jam: Icon's poor performance had been a drain on the studio's internal image at EA as well, particularly since the publisher bolstered its sales goals based on flashy demonstrations. "The expectations internally within EA were much greater, because it was showing so well," says Martin. "Kudo was an awesome salesman, so he was getting everybody pumped."

Part of EA Chicago was working on a Def Jam: Icon sequel — which our sources say was playing significantly faster than the original, addressing players' primary complaint — alongside Marvel, but Allen claims that Electronic Arts wanted out of licensed properties. "They wanted to basically cut as many IPs that they did not own, besides the sports IPs," he says. "They dropped Def Jam. That was the first thing we heard about before the closing of the studio."

Additionally, there had been a shake-up in the senior leadership of EA just months before EA Chicago closed, with Executive Vice President John Schappert — who oversaw the studio's efforts — departing in August 2007. "When that happens, nobody's really safe. Kudo was like a rockstar; he was a brand almost on his own," says Martin, who says Tsunoda's perceived persona didn't reflect his care for the employees. Still, he notes, "When your big corporate umbrella protector isn't there anymore, that's trouble."

Over budget, behind schedule and reliant on the kind of expensive license that EA wanted to shed, Marvel — and its unprofitable developer — became expendable. "EA basically did not believe that we could pull off an 80-plus [Metacritic] title with all three SKUs in the amount of time that we had," says Allen, who suggests that further limiting the scope and not doing a Wii version could have yielded a successful release.

Following the closure, many members of the team stuck together in Chicago and founded Robomodo, while others — like Tsunoda and Bennett — regrouped at Microsoft and helped bring the Kinect and launch titles to market. Others moved all around the industry, ending up at places like Bungie, Bioware and 343 Industries. And while their last work for EA Chicago never came together as intended, it hasn't dampened some of the memories of that time and that team.

"I worked at a number of EA studios, and EA Chicago was very different ... when people daydream about what it is like to work in games, they are imagining EA Chicago," says Omotani. "It was an amazing experience. It was one of those places where legends come from."

But unfortunately for the team, not the place from which a legendary Marvel fighter — or anything else after Def Jam: Icon — would emerge. Babykayak

Images: Electronic Arts