AS THE ELEVATOR doors open onto Behaviour Interactive’s studio, screams may ring out across the capacious office.

“We have some sound booths over there, and there's quite a lot of screaming going on for Dead by Daylight,” explains Matt Jackson, design director. “It's a bit, uh, interesting. Once in a while, you're like, ‘Oh, is somebody hurt?’ No. That's okay. That's normal.”

Ash Pannell, a creative director at the studio, has supplied some of the screams players hear as they stalk through the horror-inspired locales of Behaviour’s hit multiplayer game. “It's really quite awful work,” says Pannell. “The one I hate is when you have to do the breathing. Like, ‘Huff, huff, huff.’ First of all, it's weird. But second, after like five minutes, you're dizzy and hyperventilating in a sound booth.”

Those screams are part of Behaviour’s history. Not so long ago, they may have been real instead of improvised.

Behaviour Interactive occupies the fifth floor of a brick edifice in Montreal, Quebec. Over 100 years ago, before the advent of video games, that same building was a sweatshop. In the late 1800s, the garment industry was booming, and Montreal was at its epicenter. Factories dotted city streets. Inside each site, overseers presided over 10,000 workers who toiled away for 15 hours a day. They made everything from sack suits (leisurewear for middle-class men) and dresses to bespoke orders. Their wages: 50 cents for every two dresses.

Workers reached their breaking point in the summer of 1912 when nearly half the working class went on strike for 45 days. Conditions gradually improved until many of the sweatshops were abandoned. Now Montreal is home to hundreds of tech companies, many of them indie and triple-A developers. As of 2017, over 11,000 individuals worked in Quebec’s gaming industry. Montreal is the fifth-largest center for game development. As of 2019, nearly 500 of those developers work at Behaviour Interactive.

Behaviour was co-founded by Rémi Racine in 1992 as Multimedia Interactive. It started as a development house for games on CD-ROM and grew until Malofilm Communications scooped up Multimedia Interactive and merged it with another studio, Megatoon. The resultant entity was called Behaviour Interactive.

“We're celebrating 25 years, a little more than that,” says Mathieu Côté, a game director at the company as well as its head of partnerships and resident historian. “And mostly it's been making games for other people: licensed products, ports of existing games, things like that. Behaviour was very well known among developers and licensors. These people knew Behaviour stood for something: That you would have a quality game; that you would have it on time; that it will cost you exactly what we said you would pay for. We had a great working relationship with these people.”

But game development is an industry full of passionate and creative people, and Behaviour’s creators were bound to conjure pixels and gameplay from their imaginations instead of someone else’s design documents and licensed cartoons. “There were a few instances where Behaviour created its own IP,” says Côté. “Way back when, it used to be in collaboration with a publisher, because you had to. You had to reserve a space on shelves, and other sorts of things that demanded contacts and expertise we didn't have.”

One of Behaviour’s first original IPs was Scaler, an action-platform game where players channel special abilities such as transforming into a lizard able to climb walls and lob bombs. Another was Jersey Devil, a platformer rife with secret levels and collectibles such as tokens. Behaviour has made over 250 games, many immortalized by murals that depict scenes from its games and adorn the walls of conference rooms.

Collectively, those murals represent one way Behaviour’s development teams have left their mark on the building. The walls have been freshly painted, plumbing and electrical wiring torn out and upgraded. Conference rooms big and small are used for meetings and presentations. The floors are rough and unfinished, a powerful reminder of the line of work that occurred there decades ago, and a testament to how Montreal’s history of labor has changed.

There are no individual offices at Behaviour. Every square inch of floor is occupied by desks and equipment. Even Rémi Racine, still president of the company he started in the early ‘90s, shares workspace with his developers. The office layout speaks to Behaviour’s flat, egalitarian makeup. Ideas can come from anywhere. Developers are even allowed to moonlight on their own projects so long as their work doesn’t conflict with anything Behaviour has coming down the pipe.

Today, Behaviour is most well-known for Dead by Daylight, an asymmetric multiplayer shooter that pits a group of players against one significantly stronger player. “We were kicking around this prototype that was essentially hide-and-seek,” Côté recalls. “One absolutely powerful person running after someone with no power at all, who's trying to accomplish a thing. It just needed to be a symbolic thing, and it was super-fun.”

Two developers put together a prototype that mashed hide-and-seek mechanics with the notion of playing the villain in slasher movies from the 1970s and ‘80s, like Jason Vorhees stalking the horny teen counselors through the grounds of Camp Crystal Lake. “When we were trying to get that off the ground, I worked with a very talented guy to make a unique prototype of that game,” says Ash Pannell.

Pannell and his co-collaborator were excited by the idea of a multiplayer game that put players on uneven sides: one a gaggle of relatively normal folks, the other a ruthless killer. The trouble was framing that idea as a game that could be properly specced out and fine-tuned so that one side couldn’t exploit its advantages—strength in numbers versus one mega-powered brute. Pannell and his colleague decided to let their progress speak for them.

“It was just an idea at the time, but it was so successful, that prototype, in being able to explain the core concept of the game, that the powers that be here decided they should create a prototyping department,” Pannell says. “It would be separated off and mandated to do nothing except build prototypes for future [projects].”

Released in 2016 on PC before coming to consoles a year later, Dead by Daylight casts four players in the role of survivors while a fifth hunts them with murderous intent. The killer wins when all the survivors have been reduced to corpses. The survivors win if they manage to avoid capture and escape. Starbreeze, an independent developer and publisher based in Sweden, offered Behaviour a sweet deal: Starbreeze would handle publishing, and Behaviour would share royalties and remain in total control of the Dead by Daylight IP. “That's why we got with them in the first place,” Côté says. “They were one of the only [publishers] who offered to let us keep the IP. Obviously that's the most valuable part of the whole deal.”

As the game’s success grew, Behaviour found itself in a position to buy back the publishing rights. Management got along swimmingly with Starbreeze, but publishing the game themselves would enable them to pocket 100 percent of Dead by Daylight’s profits. The deal was finalized in January 2018. “We'd never have been able to do this without them,” Côté says. “They were great. They had a lot of support and expertise. They knew what they were doing. They helped us grow into the company we are now, and the face we are presenting to our players and to our business partners. But now, we have an expertise that we created through that, and we're able to do this on our own.”

IN THE SPRING of 2016, Behaviour’s leads and directors convened to discuss next steps. Dead by Daylight was picking up steam in early access—a path for developers to release an unfinished game and flesh it out using feedback from players—and they wanted to ride its momentum into another asymmetrical, competitive title.

Before that snowball got too big to stop rolling down its hill, however, they took stock. What did they like about the asymmetric genre, nascent as it was? What could be done to push it further? To plant Behaviour’s flag more firmly in its rich, loamy soil? “Asymmetric games are not a huge genre,” says Matt Jackson. “There are some, but Dead by Daylight is arguably one of the most successful ones. So we felt like, and we feel like, asymmetric games haven't necessarily reached their peak. There can be more of them.”

Over the course of several meetings, an experiment bubbled up from the studio’s newly minted prototype department. The pitch was for two lopsided teams to duke it out in arenas called Deathgarden, which became the game’s title. “Deathgarden was, I think, number six or seven,” says Mathieu Côté, of the prototype developers went through before hitting on another with potential.

Asymmetrical games boil down to numbers: How many on one side, and how many on the other? In the prototype, one side featured a single player. The other started out with four players, same as Dead by Daylight. After its gameplay caught on with developers, the two engineers who staffed the department rolled out another early build that upped the player count to six against one. It might even be possible, directors speculated, to structure the game like an MMO, where instances supported dozens of players. But, no, that would be too chaotic. The player count scaled back slightly to five.

The important thing Pannell and Jackson, the lead directors on Deathgarden, wanted to dial in was the opposites-attract feeling that had defined Dead by Daylight. Deathgarden’s gameplay was still in flux, but they knew that however it shook out, they knew what they wanted to avoid. Most competitive shooters assigned players the same objective: Riddle each other with bullet holes regardless of which side they landed on. That wasn’t Behaviour’s style. “The asymmetric games we make are literally completely the opposite,” Pannell says. “Completely different things happen on both sides: I am hiding with my third-person camera; I'm in first-person trying to put a dot on your head. Every single mechanic is fighting each other, and complementing each other.”

“Deathgarden was a competitive...” Jackson pauses. “I don't want to say sport, even though we do use that term internally. It was more of a competitive sports-like game.”

The premise of Deathgarden would be familiar to players who’d played Unreal Tournament and other competitive shooters of yore. In the far future, society has raised the bar for athletic entertainment from basketball and baseball to gladiatorial arenas. Five Runners carrying bows and arrows leap, scrabble, and shoot through an arena using parkour-style movement in their goal to open the exits and escape from a single Hunter armed to the teeth with firepower. Runners, who play from a third-person perspective, can incapacitate the Hunter but can never kill it. From first-person perspective, A Hunter can knock down each Runner three times before performing an execution that eliminates them from competition.

Runners open exits by capturing control points. Each time a Runner claims a control point, the Hunter is alerted to their presence. It’s a game of cat-and-mice that the development team believed in when Deathgarden went live in early access on Steam in April 2018.

Things went well over the first 24 hours. More players joined up on day two. On day three, the game’s population dropped off a virtual cliff. Players trying to join in selected a character—Runner or Hunter—before drumming their fingers for up to 30 minutes while Deathgarden tried fruitlessly to match them with other players. There was no one playing.

Developers monitoring the game knew player retention was to blame. But retention was a symptom. They scrambled to diagnose the cause. As they browsed forums and message boards, one reason for the abandoned servers became apparent. Beyond the thrill of the gameplay itself—assuming players could find a match—there was no reason for them to keep playing. “We knew at the time that we hadn't launched with progression as a mechanic, because it wasn't at that point in our road map what we thought we needed,” Pannell says. “But it became apparent that that might be a very critical factor in what was happening.”

As hours ticked by, the situation grew more dire. Multiplayer games tend to live or die based on self-fulfilling prophecies: If new players research a game and find an active community, they’ll join in. If no one playing appears to be playing, they’ll assume the game is dead and move on.

Deathgarden’s leads sent out surveys to their players. “People who were enjoying the game were enjoying the game, but one part of overwhelming feedback was, ‘There are other games I can play to give me a reason to keep going,’” Pannell remembers reading.

Balance emerged as another ongoing issue, one typical in asymmetrical games. While the Hunter was unkillable, it was easy—too easy—for two or three Runners to gang up on their foe and incapacitate them repeatedly while their teammates captured control points and opened exits. “More than just frustrating people, it damaged the fantasy of the game. That was one of the problems we and the community identified,” Jackson says. Hunter players countered by executing players as soon as possible, which in turn annoyed Runners who found the game too aggressive.

There were other problems common in any community-driven game. One was the steep learning curve. “It was very focused on team play,” Jackson continues, “and a lot of people really gravitated to that, but a lot of other people felt it was overwhelming: to start a match, to learn these team-play rules with a lot of pressure to perform.”

Pannell had spent the first two days of Deathgarden’s early access launch watching a particular ebb and flow. One of four things tended to happen: a team of Scavengers would click and utterly dominate; a team of Scavengers would fail to harmonize, each player getting caught up in his or her own goals and making everyone easy pickings for a Hunter; a Hunter could be skilled and able to isolate Runners; or a Hunter might get overwhelmed and grow frustrated at what some in the community perceived as Hunter bullying. “We had significant balance issues with our team-based game,” he admits. “They were solvable, but they would have taken significant work and time to solve. That had caused us a bit of a problem on retention as well. It wasn't encouraging people to stick around.”

Then there were the Dead by Daylight faithful, the diehards who wanted an experience like their survivors-clash-with-killers sensation, yet different enough that they had a reason to play both games. In particular, surveys revealed that players preferred the grimmer tone and aesthetics of Dead by Daylight compared to the tamer milieus of Deathgarden. Sure, lore for the latter described it as the blood sport of the future, but there was sports, and then there was running from a deranged killer in scenes ripped straight out of a blood-soaked horror flick.

“We want to embrace that darker tone and appeal to some of the same types of people who would play Dead by Daylight,” says Jackson. “Hopefully some of the same fans, and maybe some fans who understand the horror genre. Deathgarden's not a horror game, but it definitely has horror themes. That was part of the reason we wanted to shift closer to a genre that people who played Dead by Daylight understood, as well as more broadly in the marketplace.”

Over the following weeks, Jackson, Pannell, and other leads met in conference rooms to ask hard questions that drudged up harder answers. “Being a designer, it's always difficult when the thing you've poured blood, sweat, and tears into isn't perceived the way you wish it had been received,” Jackson says.

Pannell asked the others, as he had asked himself, what they had set out to do with Deathgarden, and whether they had achieved those goals. Though they were proud of what they had created, they had to face facts. “We wanted to communicate to competitive, team-based players,” he says, “and we didn't have a mouthpiece loud enough to break through the wall of Fortnite and other games out there—all of them, to be honest—to make a difference and say, ‘Hey, here's our game. Check it out.’”

Jackson and Pannell had kept the game’s development team in the loop. The directors talked with them again and stated that they had come to a crossroads. They either had to address the mountains of feedback accumulated from surveys and forums, or consider the possibility of letting their Deathgarden go fallow. “It became clear at that point that this wasn't going to be a minor tweak,” Pannell says of those meetings. “It was going to be, we need to ask some hard questions about what we're asking players to do, and how they're expected to interact. We still don't know if this is going to work because we've not [relaunched] yet, but we hope that this time, through what we've added to our game, that they will be appealing to people who wanted to play in the first place.”

“That period for me was very challenging, because it's very personal. It's not meant to be personal, but I took it personally because it's design, and I did this work with my team,” Jackson says.