How The Makers Of Mafia III Lost Their Way
In mid-2016, a few months before the release of their first game,
Mafia III, the developers at Hangar 13 in Novato, California gathered for an all-hands meeting. There, according to two people in attendance, Christoph Hartmann, president of the game’s publisher, 2K, told employees that their bonuses would be tied to the game’s aggregate review score on Metacritic.
Some people remembered the threshold being an 85; others remembered it being 80—either way, it didn’t matter.
Mafia III did not earn an 80 on Metacritic. It didn’t even get close. Today,
the open-world action game sits at 68, which is described as “mixed or average” on the aggregation website but is considered a critical disappointment among big-budget video games. Reviewers liked some aspects, but knocked the game for feeling
grindy, repetitive, and buggy. Some developers still got bonuses, but they weren’t nearly as big as they would have been if
Mafia III’s review scores had been higher.
Morale at Hangar 13 took a hit, according to people who were there, but there was also a widespread sense of optimism at the studio. Critics had praised
Mafia III’s story, which revolved around a black Vietnam veteran named Lincoln Clay, and many pundits had complimented the game’s narrative choices. Here was a game that explored structural racism in unprecedented ways, allowing the player to experience life in the 1960s from the eyes of a black man, frequently abused and pelted with racial slurs. It was designed so that if Clay stepped into a rich, white neighborhood, the cops would watch him more carefully. In poorer districts, police would take longer to show up. It was rare to see this sort of thing in a video game, let alone a multi-million-dollar action game.
In November 2016, a few weeks after
Mafia III’s launch, 2K bragged about the game’s sales, calling it the “fastest-selling game” in the publisher’s history. That was a success to Hangar 13, and the studio’s developers were excited to take the narrative and technical lessons they’d learned and apply them to a new game, whatever that might be. After
Mafia III, they felt like they could do way better.
A year and a half later, however, Hangar 13 is a far different place. This past year was a turbulent time for the studio, as it floundered while trying to figure out what its next game would be. Many of the people who helped make
Mafia III are now gone, including the game’s art director, technical art director, senior producers, design director, many design leads, and a number of other key staff. Some left voluntarily; others were let go or asked to move. Some went to other game companies; others moved to the top-secret studio next door to work on an unannounced new
BioShockgame.
What happened? How did such a promising studio hit so many roadblocks? Since February, I’ve interviewed a dozen people familiar with goings-on at Hangar 13, all of whom spoke under condition of anonymity in order to protect their careers. They’ve told stories of
Mafia III’s rocky development, of the studio’s troubled 2017, and of visions of a canceled Berlin spy game that might have made for a fascinating successor to
Mafia III, if it weren’t reworked into something else entirely.
Hangar 13 went through two different rounds of layoffs over the past year, including one in mid-2017 that went unreported. During the second layoff, in February 2018, the company moved some staff out of its office in Novato to locations in the Czech Republic and Brighton, England, offering them new jobs at lower salaries, according to two sources. (Of course, Brno, Czech Republic and Brighton, UK have lower costs of living than the San Francisco Bay area, in which Novato is based.)
When contacted for comment on this story, 2K sent over the following statement, attributed to a spokesperson:
The restructuring at Hangar 13 earlier this year was a result of 2K’s need to scale resources appropriately based on project timelines and 2K’s long-term business objectives, and were in no way a reflection of the team’s progress or creative process. These are always difficult decisions, and we made every effort to assist those affected. It is important to note that half of those individuals were offered transfers to other teams within Hangar 13 or elsewhere within 2K in order to retain top talent, and happily, many have chosen to remain within our organization.
Hangar 13 currently has more than 150 team members, and is already growing in key departments. Making games is a creative process, and Hangar 13’s work over the last year has resulted in an exciting new project. 2K remains fully committed to the talented team, the studio leadership, their creative vision, and the continued future growth of the studio.
The publisher would not elaborate further, nor would it say how many of those 150 team members were in California, as opposed to Hangar 13’s studios in Czech and the UK. When
Mafia III shipped, those on the team estimated that Hangar 13’s California studio was over 100 people, with two people telling me it was likely at least 150. Given that Hangar now comprises 150 people across three different offices, it’s safe to say that the studio’s headquarters has shrunk.
It’s also safe to say that Hangar 13 is a different studio now than it was when
Mafia III came out. Throughout all of my interviews with those familiar with the company, there’s been one common thread: Everyone had hoped for more. After
Mafia III, both fans and the staff of Hangar 13 were thrilled to see what would come next. Now, there are nothing but questions, and a studio that looks far different than it did just a year ago.
On December 4, 2014, 2K made a big announcement. The publisher had formed a new game studio called Hangar 13, to be led by Haden Blackman, a veteran comic book writer and former LucasArts game director best known as the project lead of 2008’s
Star Wars: The Force Unleashed. In interviews tied to that announcement, Blackman said their first project would be a story-focused console game, and 2K promised big things. “Under Blackman’s direction, Hangar 13 will lead the charge in creating some of the most memorable triple-A experiences of this generation,” the publisher said in a statement to press. There was buzz everywhere: Blackman had recruited a number of other people who came from the ruins of LucasArts, leading developers and pundits to joke that it should’ve been called
Hangar 1313.
Hangar 13 had actually been quietly operating for months before that December announcement, working on a project that people probably didn’t expect:
Mafia III, the third game in a very strange franchise. The first
Mafia, an open-world action game revolving around the Italian mob, had been developed by a Czech studio named Illusion Softworks for the PC. When it came out in 2002, it found success in Europe but in North America was mostly overshadowed by 2001’s seminal
Grand Theft Auto III.
During development of the sequel, 2K parent company Take-Two Interactive purchased Illusion Softworks and renamed it 2K Czech.
Mafia II, out in August of 2010, received mixed reviews and was overshadowed by yet another Rockstar game,
Red Dead Redemption, which had come out three months earlier. (Rockstar and 2K are part of the same corporation, Take Two Interactive.) Analysts suggested that the second
Mafia, which had been in development for six years, was
unlikely to turn a profit. A third game might have seemed unlikely.
Yet it was happening. 2K Czech had started working on
Mafia III as soon as the second game shipped. The project was in development for roughly three years, sources said, and it went through several fits and starts over that time as 2K Czech tried to figure out where to take the series. “They had spun their wheels for a long time,” said one person familiar with the game. “2K wasn’t happy with it, so the goal was to bring the game over to Novato, bring in some new blood and a new approach, and try to reinvigorate it.”
By mid-2013, 2K had rebooted
Mafia III and moved the project from the Czech Republic to Hangar 13. Throughout the remainder of 2013 and then 2014, Blackman quietly recruited staff from all across the San Francisco Bay area, not just from the ashes of LucasArts but from several other studios, including the company formerly known as 2K Marin, which had made
BioShock 2. He also brought over a bunch of people from 2K Czech, along with the game engine that they had used for the first two games. (An engine, in video game parlance, is a collection of software and tools that drives the development of a game.)
Blackman, along with writer William Harms and his other leads, wanted to do something radical with
Mafia III. They didn’t just want to make another open-world crime game—they wanted to tell a story that drove conversations that most games wouldn’t touch. Eventually they landed on New Orleans, the 1960s, and the story of Lincoln Clay getting revenge on a crime family after they’d killed his adopted father and brother. At first, it was envisioned as a straightforward revenge tale, but 2K boss Christoph Hartmann wanted
Mafia III to compete with Rockstar, one of Hangar 13’s sister studios and the company that had been overshadowing
Mafia since its inception. Several people familiar with Hangar 13 told me that Hartmann wanted them to make a game on par with
Grand Theft Auto. He wanted districts, empire-building, and a massive open world.
As
Mafia III went into production, throughout 2014, 2015, and then 2016, that’s what it became: an open-world game that tried to say something deeper. Like
most video games,
Mafia III’s development had some serious struggles. Some of them were common difficulties. The engine, which grew out of the Illusion engine used for the first two
Mafia games, went through active development alongside the game, leading to roadblocks as designers and artists waited for their tools to work properly. One common analogy for trying to build an engine and a game at the same time is that it’s like trying to land a plane while simultaneously paving the runway.
That sort of challenge happens all the time, particularly when a game studio is working with its own technology, but Hangar 13 had unique problems. Because the Novato-based studio had taken so much technology from 2K Czech, large swaths of the code came with comments in their language. “Almost the whole codebase was written in an amalgamation of English and Czech,” said one person who worked on the game. “When you were trying to debug something, you’d go to an engineer, and if they didn’t speak Czech they’d have to first translate what they were looking at in the comments to English, check that translation, then try to see if they could fix the problem.”
Hangar 13 also had to deal with a social divide between the old
Mafiadevelopers and the new ones. People who worked at the studio say there was no animosity between the groups—“Everybody got along quite well,” said one—but they’d often spend time sequestered. The Czech speakers would eat lunch and hang out together, as one might expect from expatriates who shared a common tongue. Plus, there were cultural clashes to sort out. “They have different mannerisms,” said one person who worked there. “Czechs will tell you flat out that an idea sucks, it’s the stupidest thing they’ve ever heard. Americans will sugar-coat it, dance around it.”
One of the biggest struggles, according to several people who worked on
Mafia III, was that management finalized too many decisions too late in development. This is a common problem in games, particularly games developed by hundreds of people, where changing features can lead to a lot of wasted time. “With a big studio, the more you push in the wrong direction, the worse things get,” said one person on the project. “Because it’s hard to correct those things. And it takes a lot of time to get those things to a finished state. I think studios are not very good at failing early and failing often—they tend to take their ideas way too far.”
When
Mafia III came out, critics dinged the game’s structure for feeling repetitive, forcing players to do the same actions over and over in order to get through the story. The game had ten districts, each containing two rackets, or small criminal operations that Lincoln would have to assault and conquer. Before you could attack the boss of each racket, you’d have to deal damage to it, by killing enforcers, destroying vehicles, and blowing up contraband. At first this was a fun activity, but by the fifth or sixth district, it started to feel tedious. And there were ten of them.
For a long time, however, that damage system didn’t exist. Instead, racket bosses would be open at any time, and you could raid them well before your character was equipped to take them down. The logic behind this decision was to give the player as much freedom as possible, but those who opposed it pointed out that it would just lead people to have a bad time. “We’d say, ‘Well if players know the content is there, they’re going to go attack it and fail repeatedly,’” said a member of the team. “Sure enough in play tests, everybody was throwing themselves against the wall trying to defeat these hideouts.”
“With a big studio, the more you push in the wrong direction, the worse things get.”
Eventually,
Mafia III’s managers agreed that this had to be fixed. “We tried all sorts of different ways to solve it,” said another person on the team. “But it was so late in development, the only thing we could come up with was to lock it behind damage.”
As a result, the game felt grindier than it might have if Hangar 13’s design team had solved these problems earlier. “That represents six months of work going down a certain direction,” said a different person. “That’s a lot of time in development, when you have to reverse that decision out.”
People who worked on
Mafia III said this was a trend during the project. Designers on the game squabbled with directors over what was going to be a problem, how to identify it, and how to fix it. In conversations, nearly everyone I spoke to told a similar story: Designers and other people who played the game argued that the district system felt too repetitive and needed to be changed, while managers thought it might work and wanted to wait to see if it would all come together at the last minute, as many video games do.
“The things we got knocked for, once it came out, were always known,” said one person who worked on the game. “We just couldn’t get traction on it. Like the game being grindy and repetitive—we all knew that for years, had hundreds of conversations about it, and just couldn’t get upper management to agree with us. And then we’d go to focus testing, and they’d go, ‘The game’s too grindy.’ We’d go, ‘We’re in beta, can’t do much about it.’
Mafia could’ve been better than it was. All the issues, the floor knew about for years—we just couldn’t get the director level to agree with us.”
Thanks to these fights, technological impediments, and other roadblocks, the
Mafia III team had to crunch long and hard toward the end of the project, putting in nights and mandatory Saturdays to get their game out the door. One group of designers, engineers, and artists crunched for nearly a year to add details to
Mafia III’s world, filling the game’s virtual New Orleans with civilians, protests, and people who just loved jumping in front of your car.
Outside of the grindiness, the biggest complaint about
Mafia III was that it was full of bugs, which people who worked on the game blame on two main factors. The first was that they’d never made an open-world game like this, full of simulations and complicated mechanics. “In the previous
Mafia games, which were using the same tech, there wasn’t much going on,” said one person. “You just drove from mission to mission. It was a big leap to have an actually populated world full of content.”
By the end of development, Hangar 13 had to cut some features because they were causing too many issues. One former employee, offering an example, said that they’d implemented a mechanic that would allow Lincoln Clay to vault onto a car’s hood, then onto its roof. It would’ve been a cool little method of traversal, and something to do while you ran from district to district. But as a result of this, QA testers found that you could vault on top of a truck and use it to reach places that the developers didn’t want you to access. One morning, the team got in and found that the QA team had added hundreds of new bugs to the database. “Every single one was, ‘Took garbage truck to spot A; vaulted up to X,’” said the person. It didn’t take long for them to cut the feature.
The second factor was time. People who worked on the game admit that they could have used a few extra months to polish and fix bugs. But after three years at 2K Czech and then another three-and-a-half at Hangar 13,
Mafia III had already gotten plenty of time in 2K executives’ eyes, even if Hangar 13’s version was completely new.
In the days after finishing
Mafia III, the developers at Hangar 13 got some time off to recuperate from the lengthy crunches they’d gone through. Then it was on to the next project. A few people left the studio after shipping the game, but for many people who worked on
Mafia III, the thought of a new game was refreshing. They were elated by the critical conversations
Mafia III’s story had inspired, and they wanted to keep pushing the boundary with new games. Plus, their technology was in better shape now.
“I felt like going forward to the next project they’d learned some good lessons,” said one person on the team. “I have nothing but respect for Haden [Blackman]—he looked at what they’d fucked up and said, ‘Yeah we fucked this up. We need to do better.’”
As they entered 2017, the people behind
Mafia III split into two main groups. Some people moved on to the game’s downloadable content, while others started conceiving ideas for what their next project might look like.
At first, the plan was to make
Mafia IV, set in Vegas during the 1970s. It was an enthralling vision—a video game take on Martin Scorsese’s
Casino, set in the glitz and glamor of mob-controlled Sin City. It’s easy to imagine what that might have looked like: fighting enemies on the strip, managing your own casino, putting your character’s name in lights.
Plans for a fourth
Mafia didn’t last long. At one point, as two sources recalled, Blackman flew to New York City to meet with Take-Two and 2K higher-ups. He then told the staff of Hangar 13 that he’d been given a choice: They could either develop
Mafia IV or start something completely unique, a new intellectual property. He said he’d picked the latter. “The way it was pitched in the big all-hands meeting was:
Mafia IV was a great thing, it’s exciting,” said one person who was there. “But we’ve always wanted to do our own IP.”
This new IP would become
Rhapsody, a game about subterfuge in 1980s Berlin. You’d play as a Russian Jew whose parents had been murdered in a Soviet labor camp. He’d be rescued by Americans, then recruited to join a spy organization called Rhapsody. “It was hitting a lot of the beats we were good at,” said one person on the project. Like
Mafia III, this game would put you in the shoes of someone who was treated as subhuman by the people around him and thought largely in terms of vengeance. “He’s doing missions, trying to save the world and get revenge on whoever killed his parents, trying to decide between the personal good and the public good.”
One source compared it to the spy movie
Kingsman, “without the really goofy shit.” You’d equip a variety of gadgets and sneak around Berlin, spying on enemies and doing all sorts of missions for your organization. Hangar 13 prototyped mechanics like a spy car, which you’d use to cross the border checkpoint between East and West Germany. Your inventory of items for any given mission might be limited to what you could smuggle in the car.
It’s easy to get wistful about games that may never happen, and it’s far simpler to dream up ideas than it is to implement them, but by all accounts the Hangar 13 team was ecstatic about
Rhapsody. Even today, when people discuss the project, they sound fired up about how cool it might have been.
Throughout 2017, a few pivotal things happened. In the spring, Christoph Hartmann left 2K Games. Hangar 13 made some leadership changes and hired some new people to run the design department. Over the summer, the studio put two small teams to work on “incubator” projects that were designed to explore new ideas and technology. (One of those projects was later canceled.) Eventually,
Rhapsody changed.
During early production on any video game, especially a new IP, it’s common for the developers to make frequent changes as they try to figure out what their game will actually look like. What that meant for
Rhapsodywas a massive overhaul, one that ditched the subterfuge entirely.
One of
Rhapsody’s proposed mechanics had involved musical cues. Part of the main character’s backstory involved the piano that his parents had bought for him as a child, and one core idea was that the game’s “detective mode” would be triggered by listening to music. At some point, according to everyone I spoke to who was there, Hangar 13’s management decided to extract that idea and make it the focus of their game.
Suddenly,
Rhapsody would be an action game fueled by music. Maybe instead of being set in Berlin, it’d be in San Francisco. And maybe instead of playing as a spy, you’d be a superhero. “The concept was using music as a weapon,” said one person on the game. “You could use music to inspire you, channel it into weird effects.”
One person brought up an example that was bounced around at the studio: You’d be listening to, say, Survivor’s “Eye of the Tiger.” Press buttons on the controller at the right time, and you’d summon an actual tiger to help you in a fight. Another idea, described to me by a different person on the team, involved taking riffs and song stems from a variety of tracks and combining them all to create different abilities. Hangar 13’s leadership had loved the audio style of
Mafia III and wanted a game that was all about music, from Metallica to Miley Cyrus.
Morale sunk. As part of this transition, Hangar 13 laid off around a dozen people, and those who remained weren’t sure what they were even doing.“A year into it, most of us still didn’t get what the concept was,” said one person who worked on this new version of
Rhapsody. “We made a prototype of this for a year, and nobody still knows what the game is.”
In the midst of all this confusion, some employees of Hangar 13 couldn’t help but look with envy at their neighbors. Next door, a small group of people were working on a project code-named
Parkside, quietly recruiting from across the video game industry for a game so secret, they wouldn’t even tell their colleagues at Hangar 13 about it. Word got out, though, that it was in fact a new game in one of the most interesting shooter franchises of the past decade:
BioShock.
“If anyone from Hangar 13 tried to strike up a conversation with them, they’d say, ‘Oh, we can’t really talk about it,’” said one person familiar with the studio. “It was all very tight-lipped.”
This caused some tension, as some developers at Hangar 13 looked over and wondered,
Why are we not there? The
Parkside team was working with the familiar Unreal engine rather than volatile proprietary technology, and
BioShock has a ton of cachet among developers and fans. Between
Mafia III’s mediocre reviews and
Rhapsody’s constant shifts, some members on the Hangar 13 team were desperately hoping to move next door. A few got to do it, but not everyone.
“The issue is that
Parkside’s not ready for lots of people,” said one person involved. “They’re trying to be really smart about figuring out what the core thing is. They’re careful about not falling into the same problem every studio has, where they have too many people and nothing for them to do.”
In fact, having too many people and nothing to do was starting to become a problem at Hangar 13. By the end of 2017, according to one source, the studio’s management had all but given up on the music idea. It was never practical to begin with—the licensing costs for all those songs would have escalated into the millions. Instead, the project started morphing into more of a traditional superhero action game. “They never came out and said, ‘Hey, we’re not doing the music thing anymore,’” said the source. “It just kind of silently went away.”
On February 15, 2018, Hangar 13 laid off a number of staff. The company would not say how many, but three people who were there told me that
it was significant. Some of those staff were invited to take new positions at 2K Czech, which was rebranded to become part of Hangar 13, or at the company’s studio in Brighton, UK, which has not yet been officially announced. Today, 2K PR says that there are 150 people at Hangar 13, but that means something far different than it did just a year ago.
“I know a large part of the reason for these layoffs was financial,” said one person involved. “It’s a lot cheaper to hire and maintain engineering [and other staff] in the UK and Europe. The cost of living in San Francisco in exorbitant.”
There’s one clear piece of public evidence that supports this theory: Hangar 13’s jobs page currently lists six openings, all of them in the Czech Republic:
It’s not clear what Hangar 13’s next game will be. Two people close to the company told me that right before the layoffs, the company had discussed adding drop-in co-op to the newest iteration of their superhero game. Rumors floating around Hangar 13 circles have suggested that the company is looking to chase the “games as a service” dragon and make something that can support ongoing revenue streams.
Even if the studio does pivot to multiplayer, it’s not impossible to imagine a Hangar 13 that continues to take interesting risks with video game storytelling. The main writers of
Mafia III are still there, as is Blackman, who has spoken frequently about his love for narrative. One truism among game developers is that a studio’s culture is driven by the discipline of whoever is in charge, and the guy who runs Hangar 13 is a writer. Two different people familiar with 2K’s operations told me that the publisher’s executives believe strongly in auteur theory, and that the company sees Blackman as a strong creative leader, much like they see original
BioShockdirector Ken Levine.
A good new game could still emerge from Hangar 13. But to those who saw
Mafia III as a flawed-but-important work and couldn’t wait to see what the studio would do next, Hangar 13’s rocky 2017 is not encouraging. This was a studio full of people who came close to greatness, and if they achieve it with their next project, it will be the result of an unexpectedly arduous journey—and an unexpectedly different set of faces.
