From Dead Island to Dying Light

You turn a corner and there it is: Techland. A striking black and red building with a glass front, like a race car.

Here, in this semi-industrial part of the Polish city of Wroclaw (pronounced Vrotslav, with a rolled R), the building could not be more obtrusive. There's some sort of office complex to one side, across a barren field. There are some houses here and there — small, cottage-like affairs, with well-kept yards. And then there's the place with the new cars and motorcycles. The video game company.

You hesitate to boil down a company that's been selling games for over 15 years to just one title, but they'll acknowledge it themselves. This is the place where Dead Island — the blockbuster tropical island-set, melee-happy, weapon-crafting zombie apocalypse game — was made, and the place that, in many ways, was made by Dead Island.

Walk inside and, just past a small and somewhat messy lobby, you'll see the wall art — all Dead Island. Heroes, zombies, weapons, plastered over every vertical surface, larger than life. Just as anyone who's been in love remembers their first love, anyone who makes video games remembers their first international hit. Dead Island was Techland's first international hit, and it was a whopper of one, pushing over 5 million units.

Maintaining a creative vision in the face of publisher demands and market realities is frequently impossible.

And yet underneath the pride and self-congratulation there burns a fire to do it again, but better. To erase the festering memories of all the setbacks, the heartbreaks and compromises that made Dead Island not quite perfect.

Making games is a hard business, and maintaining a creative vision in the face of publisher demands and market realities is frequently impossible. Techland will tell you this. Dead Island was supposed to be a very different kind of game, and it was released too early, some think, and too broken. In spite of its success, it was not the game Techland wanted to make.

So it's trying again.

Techland, the video

Out of a box

Wroclaw is an unusual place.

The city itself, nestled in Poland's southwest corner between Germany and the Czech Republic, is ancient, saved from the devastation of World War II by remoteness and luck. A stroll through the city's old town, situated on an island in the Odra River, is like walking backward in time, with the occasional stark reminder of the modern world.

Across the street from the Radisson hotel, in a parking lot, stands a monument built to contain the Racławice Panorama, a gigantic, centuries-old circular painting depicting the 18th-century uprising against Russia. Just down the cobblestone street, ancient buildings housing offices and restaurants jostle for space with a modern mall, complete with a KFC and a McDonald's.

It is, all of it, a reminder that for all the wonders of the modern world, people have been living, working, fighting and dying here for thousands of years. It's a span of time, and a cultural history, that makes the advent of things like video games read like a pimple on the face of time.

Into that vast chasm of history, in the early 1990s, charged eventual Techland CEO Pawel Marchewka.

"When I was younger, significantly — more than 20 years ago — I was in secondary school [in Europe, middle and high school]," Marchewka says. "I was playing games, of course. I had my C64, and then I bought my Amiga, Commodore Amiga. In the town I lived in ... you couldn't get many games. Especially for the Amiga, which was a relatively new computer at the time. So I kind of borrowed games? I copied them."

Marchewka has a fighter's build, squat and fit. His eyes never wander. He doesn't hesitate, when speaking, not even in a second tongue. He comes off as a man who will bounce back, time and again, and he is clearly the driving force of Techland.

Techland's CEO Pawel Marchewka started the company in the 1990s by carrying buckets of ripped CDs to Wroclaw's open-air marketplace.

As a kid, Marchewka would trundle his wares into Wroclaw to trade at a local market. He realized he could make some money. He visited local shops, asked if they would carry his borrowed games. They did, and just like that Marchewka became a video game distributor. Techland was born.

"At the time we had a different legal situation here, sort of like in China, so it was OK," he says. "We started delivering products to local shops, then to another town. Then I got more requests every week so my sister helped me. Then a few colleagues helped me get enough copies to be able to deliver the stuff. Quite quickly we were one of the biggest suppliers of these kinds of games, almost for all of Poland."

Then, the first setback. With Poland's growing economy came new rules and regulations. Pesky things like copyrights and trade tariffs. The burgeoning borrowed-game industry, now illegal, dried up. Techland's competitors died out. But Techland soldiered on.

Marchewka hired programmers, built a company and made games ... eventually.

"We started with some educational products," says Marchewka. "These were the easiest to do. Then some small games, thinking that it would be really nice to one day prepare something of a quality that we could sell abroad."

That one day happened, but when it was depends on your perspective.

You could say it came in the year 2000, when Techland released its first mass-market game, Crime Cities. Or, perhaps in 2003, with the release of sci-fi shooter Chrome, based on Techland's proprietary Chrome engine. You could also say it came anytime over the next several years, when Techland released one of its nearly two dozen titles from that period. But in order to say that, you'd have had to play one of those games — and most people haven't.

A more likely candidate is a day in the year 2006, when Techland released the game that would become its first successful franchise, Call of Juarez. But even Marchewka doesn't set his clock by that one. Instead, he refers to that game as simply "another step on our way to Dead Island."

Success, of a sort

In early 2011, Techland and publisher Deep Silver released a stunning trailer for what appeared to be an open-world zombie action game. The trailer featured a small girl falling through a window, and a look backward in time to what led to that moment — a family vacation gone wrong when an idyllic island resort became the center of a zombie apocalypse.

The trailer made waves, but when the game started development, it was nothing like what the trailer would suggest, or even what the game would eventually become.

"The issues we found out about through playing were creating lots of things that needed to be reworked," says Adrian Ciszewski, a producer on Dead Island. He came on in the middle of Dead Island's development, having worked previously with Techland's third-party developers on smaller games. "There were situations where we had to remove huge parts of the systems and re-create them from scratch because [the game engine] wasn't supporting some of the things that came up after two hours, five hours, 10 hours of playing. That was the problem."

When Ciszewski started with Dead Island, the game was already in trouble. He was the fixer.

"It was a different game," he says. "It was a single-player, linear, secret island, many zombies, crazy doctor — it was just a clichéd zombie shooter without any soul. ... I changed it into this clash between paradise and zombies, where you have the open world and melee combat."

When Ciszewski started with Dead Island, the game was already in trouble. He was the fixer.

Dead Islandwas released in the fall of 2011 to generally favorable reviews, but Techland (as well as some players) wasn't satisfied. Ciszewski's changes hinted at what many believe could have been the game's biggest strengths, as highlighted in the original, and haunting trailer. But the final game appeared to not live up to that promise. Conflicts with publisher Deep Silver led to the release of a game that Techland wished it could have done better. Many at the company felt they were striving for a game that would earn review scores of 9, while Deep Silver would be perfectly happy with 8s.

"Of course, there are always people who don't like Dead Island, but we have lots of fans," says Ciszewski, and sitting in a building festooned with memorabilia from the game, crowded with boxes of its merchandise, this sounds like an understatement. And yet, in spite of having sold more than 5 million units, the most important people who aren't fans of Dead Islandare the people who actually made Dead Island.

"We get lots of letters from people who are waiting for Dying Light because they're fans of Dead Islandand still playing it," he says. "There are people still playing Dead Island, a game from 2011. It's a great achievement for us, that people have played 200 hours in Dead Island. It's unbelievable, even for us, that we could create something like a game that you could play that many times. It's a great achievement for us. We're really happy.

"Of course, the biggest problem with Dead Island is it's just — it was a buggy game. A lot of glitches and bugs in the game. That was a problem — we didn't have a huge team to polish everything we wanted. It was a huge game, because of co-op and because of things that were never done in other games. We didn't have a reference for fighting with zombies using melee combat in first-person. ... We were doing it for the first time. Dying Light is different. We're smarter."

Into the Dying Light

In late 2011, flush with cash and accolades, but unsatisfied, the core Dead Island team got back to work. It wanted to build on its success with Dead Island, but this time make a game it could be proud of, as well as profit from. The new game would be a zombie game, of course, set in a open world and melee-focused. But it would need to be better polished, and also have more freedom of movement. It would need to be bigger, more open and with more potential for emergent fun. It didn't take long for the team to decide it wanted to completely break away from Dead Island to create something different.

"It wasn't called Dying Light at the time, but when we looked at the list of features we wanted to have in the game, we found that it deserved to have a new name," says Tymon Smektala, producer on Dying Light. "It should be a new IP. It's something different, something new."

And most importantly, something Techland could create without publisher interference.

When Dying Light is released in 2015, it will be published by WB. The 100-year-old company, publisher of the Batman: Arkham games, has given Techland full creative control over Dying Light. It will be released when Techland decides to release it (the game's release date was officially pushed, by Techland, from late 2014 into 2015), and will be the game Techland wants it to be.

"[WB was] very open," says Marchewka. "They understood that we would like to deliver a game with our vision. They were happy with that on many levels. ... Maybe that's just how they operate. But it is a different approach [from Deep Silver]. ... Generally it's just a good company. They understand that quality comes first. We have the same values."

Although Techland is circumspect about what happened between it and Deep Silver with Dead Island, it's a matter of public record that Techland does not own the IP. What Techland will say on the subject is that Techland's ambition for Dead Island did not always mesh with Deep Silver's intentions, and that eventually Deep Silver won that battle.

"Generally there's always a slight difference between publisher and developer, the way they see that the product should be finished, how much time we need to polish it, versus what's a good spot to release it," says Marchewka. "Of course, as a developer, you have certain obligations to deliver it on time and contractually. So ... we would have released the game slightly later, but we were not fully controlling the process."

As publisher, Deep Silver had invested heavily in Dead Island's development, and therefore determined how the game was produced and when it was released. While none of this is uncommon in game development, Marchewka remains deeply disappointed that his company did not have the time to make the changes it felt were necessary. He believes that the Dead Island as envisioned by Techland could have been a top-rated game, something known for quality as well as fun. Deep Silver's vision, he believes, was a little lower.

"I think we missed in the quality observation, how good the game should be," he says. "We knew about areas that should be more polished, and our publisher was quite happy for a few months before we released, so there was no real pressure — 'Can you correct that, or can you correct here?' It was only, 'Guys, just get it as it is. We need it on time. That's all that matters. The game is good enough.'

"Some players maybe wouldn't agree with that."

Techland completed work on a 2013 Dead Island pseudo-sequel, Dead Island Riptide, but would not commit to future installments. Instead, Deep Silver carried the franchise forward with Spec Ops: The Line developer Yager. Dead Island 2 was announced at the 2014 E3 expo.

Techland also moved on, hoping to go backward in order to go forward.

"[We] wanted to evolve and deliver something better [than Dead Island]," Marchewka says. "We saw areas for improvement, for changing the formula of the game. Dead Island was an IP. It had certain elements already set up and explained the way they were. We couldn't change it the way we wanted.

"We felt it would make sense to actually change and create our own IP and make it 100 percent our vision. That's what we did. I think it was a very good decision."

"I don't want this game to suck"

"I remember sitting with [Ciszewski]. We would playtest the game, check out the new movement system, and as we played it, it still felt kind of like Dead Island," says Maciej Binkowski, lead game designer. "We already knew that there was going to be a lot of new stuff. ... We looked at each other and said, 'It seems like the stuff that we're doing already feels like it's something more. ... It grew from there."

00048.mts.still001 Lead game designer Maciej Binkowski is a former stuntman. His experiences with almost dying have contributed to development.

Techland's goal for Dying Light is to create a more fully immersive feeling of surviving in its world. Of being able to climb walls, surmount obstacles and actually grapple with zombies — and this would involve inventing multiple new technologies — having distinctly different styles of play in the daytime versus at night.

"The basic concept we had was that the night should be the time where you're really scared for your life, as opposed to the day, where you're more hunter than prey," says game programmer Radoslaw Malicki. "We figured out that the zombies can be changing [at night]. They should be actively and aggressively following the player during pursuit sequences."

For Malicki, the process of inventing the Dying Lightday-night cycle is representative of how Techland operates as a whole. A lot of the game's core design concepts were actually invented by line programmers, rather than the designers above them on the org chart, as at other studios. While it's not uncommon to hear that a studio seriously considers ideas from its line programmers, very few actually hand full creative control over to them, to the same extent as Techland. Here the designers are more shepherd than auteur, helping to keep the best ideas moving forward, but everyone is free to try anything. Even when they're told not to.

"We had a big argument about — I wanted to do a swimming mechanic for Dying Light," says game programmer Bartosz Kulon. "[The designers] said, 'No, we don't have the time for it.' I said, 'OK, I'll come in on Saturday and I'll do it.' They said, 'OK.'" So we have swimming."

Kulon was frustrated by the experience of playing Dead Island, having a game world that was literally an island, and not being able to swim in the game.

"Sometimes I get stubborn," Kulon says. "We want to do something really good, and I think we really need to have all the basic mechanics and movements. Swimming, to me, is basic. It was a lot of pain for me in Dead Islandthat you couldn't swim. Our publisher told us that we couldn't do it because we had time restrictions. They told me the water could only go to your knees. If you got to the deep water, you were going out of the playable area. It sucked. I don't want this game to suck."

Dying Light's day-night cycle was iterated upon and enhanced by this same programmer-focused technique, but the concept came directly from game director Ciszewski.

"We always wanted to achieve that dual gameplay, with day and night," says Ciszewski. "The general idea of being the hunter during the day and prey during the night wasn't actually done in Dead Island, but the idea was in my head before Dead Island. ... It creates a polarized game. You have two games in one box. You can sneak around and play really differently.

"You have different feelings. In the day, you can do some upgrades to your skills. You need to learn more about fighting, more about running. But very soon you can be a really great hunter during the day. During the night, you can be a hunter, but it's really difficult. It'll probably take you lots of hours to improve yourself enough to be a hunter during the night.

"At the very beginning, it's just scary."

Free fall

Ciszewski's day-night cycle didn't make it into Dead Island for various reasons, mostly having to do with team size. Following the success of that game, team size is less of a problem for Techland.

The studio is full to bursting with talent. Finding room for interviews or conversations is a challenge. People are everywhere, constantly in motion. It is a studio alive with activity.

Yet this is only partly because of overcrowding. An entire top floor of the studio has been closed off for construction while the company adds an extension to the building. When construction is complete, the studio will be roughly twice its current size, housing the many vagrants currently taking up too much space in the lower floors, as well as perhaps some of the developers from the Techland Warsaw team and the team currently working on medieval dungeon crawler Hellraid, at a satellite building a few minutes' walk away.

Meanwhile, in spite of the mess and overcrowding, a game must be made.

"Having a bigger space will definitely make it a lot easier just to breathe," says Maciej Binkowski, lead game designer on Dying Light. "At the same time, being close together, it's a huge benefit. Nobody takes stuff personally. Sometimes we're harsh to each other, but not on a personal level. We're just doing the work that we do. Pretty much anyone can say anything about anything. If you don't like it, you can come and say, 'Hey, I think that's stupid. This is how it should work.'"

Binkowski comes to Techland via stints as the following: a computer science student, physics student, college dropout, martial artist, computer science school graduate, game developer, professional stuntman and, again, game designer.

"I got a job as an associate producer on Dead Island," he says. "In the process, I started designing little things. I really enjoyed that a lot, so I asked the game director — like, do you think that with the next project, I could do full-time game design? 'Yeah, sure.' So here I am."

One particular area of expertise Binkowski brings to Dying Light is what happens to a body when it falls from a great height — and lands. He's done that.

Like the time he almost died falling off of a giant construction crane:

"It's funny now. It's a life-changer, really.

"After I graduated stunt school, after about a year of working in the business, [my team and I] went to a show where we would perform live. One of the stunts that I did was getting on a crane, about 15 meters up, and performing a backflip down to an airbag. We would do it hundreds of times. You get on the crane, you wave to the audience, then you turn yourself away and push back.

"I pushed myself away from the crane, and as I started falling, I can already see my friends sitting in the crane with their faces like this [makes a shocked face]. So at this point, I know I'm in trouble.

"I'm going down, I see their faces, I know I'm in trouble, and there's really nothing you can do in midair, except for controlling your body. What happened was, I'd pushed myself a little bit too far. Thankfully, we had this new airbag. A regular airbag that's used by firemen — if you land on the edge, it's just going to bounce you off. That's not good. We had this new airbag constructed by our old stunt coordinator, and the airbag just pulls you inside. But still, I landed on the edge, falling with my head down to the ground. Again, thankfully, there's a whole team securing the jump.

"As I was going right down with my head to the ground, the stunt coordinator put his arm out to my arm, and I pivoted around his arm, landed on my feet, and went, 'Ta-da!' Everyone in the audience, they thought it was supposed to be like that. It wasn't.

"I was that [holds his hands a few centimeters apart] close to just smashing my head into the ground. The funny thing is, it's not like you have a moment to stop. You're in the middle of a show, so I just went, 'Ta-da!,' and went for my next trick.

"As I was running I turned back and I could see my stunt coordinator going just like [shakes his head] ... And I'm like, 'Yeah, thank you. You just saved my life!'"

With Binkowski leading the way, Techland has been able to integrate climbing, jumping, falling and athletic parkour moves into Dying Light, with the hope that players will never be backed into a corner trapped by the game's technical limitations.

In real life, when you're cornered by zombies (let's suspend disbelief), you might have a way out, by climbing on boxes, over fences or even on top of the zombies themselves. But in zombie games the technology just hasn't previously existed to allow for freedom of movement on this scale.

Traditionally, game designers create the possibility of a character moving through a game world by assigning certain objects to be usable by the player in certain ways. A game map, rendered in wireframe polygons, will have certain points, called hooks, at which a player might grab a ledge or be able to climb on a building. These hooks are assigned by level designers, and then artists must build a world around them, to create something that would make sense for a character to be able to climb on or over.

And that's exactly how Techland had designed Dying Light, at first.

No more hooks

"When we started to create exploration in Dying Light, we were thinking the same as we did in the previous project," says level design team lead Piotr Pawlaczyk. By "previous project," he means Dead Island. "'OK, we have horizontal ways, we have one path here, we can do it. We can script it. We know where players will go.' Then we started work on the natural movement."

That's when everything changed.

For a game like Dead Island, since the designers knew what the player could and could not do, scripting the game in a linear fashion was easy. They knew, at any given area on a map, where the player was likely to be and could plan zombies and events around those locations. And, as in theater, make sure that what the player could get to or see looked good and worked, but not have to worry about any of the rest, like what's behind a bush or on top of a building the player can't climb.

Once you allow for total freedom of movement anywhere on the map, however, things get tricky fast.

"Then we run into problems," says Pawlaczyk, "because we must think in 3D — really 3D. ... We must do many fun things for players, give them many opportunities to have fun in two ways — in horizontal and vertical. ... In our previous project, when we created quests, we could design them along a path — A, B, C, D — and we could script an action where some special enemies would come through, or some human. We'd know where players would go. Here, we can't. We create A, the start of the quest ... then we have some awesome exploration, and finally point B, the end of the quest, where we have some final action. We can try to script that, but not always. ... The endpoint is something you can approach from different directions and different angles and different heights. We always have to consider that."

The old way involves creating a hook for every movement and every play where a character in a game could put a hand. You see it in games with certain colored ledges, indicating where you can climb up. That's a spot where a designer has placed a hook.

The problem with hooks is that every change to the design of a level requires a change to the hooks. And artists could create an embellishment that would render hooks inoperable. For example, a box placed high on top of a building could be in the way of a hook placed there by the level designer to allow the player to climb up. In order to allow the player to climb on any item in the game involves hand-coding (and continually adjusting) the specific hooks that instructed the game how to handle that operation. It is a painstaking and brutal process.

"When we started to create the [movement] system, we were talking about controlling the player," says Kulon. "We thought that we would manually select the ledges where the player could climb. We'd have the control, and they wouldn't go where we didn't want them to go. But it actually sucked.

"The first thing, in the slums, the first prototype of that slums map, we had like 50,000 hooks manually set. They broke all the time. [The designers] moved something and they needed to reset the hooks. It was a pain in the ass."

In the spring of 2012, Kulon decided to make a prototype to test a theory about how the game might work without hooks. It took him all of two days. The result was a technology that allowed the game to scan the environment around the player for objects the player could climb and then, effectively, "assume" those objects had hooks on them. The result: traversable terrain, minus hooks.

"It opened our eyes, because we gave a vast amount of places for the player to go," says Kulon. "We opened up the maps vertically."

Every building, every wall, every object in Dying Light can be traversed, and the game will continually detect where the player is and what type of movement animation it will render to describe the action the player is performing. The game engine recognizes physical objects, runs some math about the distance between the object and the player, how high the object is and so forth, and then calculates how and where the player might be able to climb on it. And it's performing these calculations nearly instantaneously.

"In Assassin's Creed, they did it in post-processing," Kulon says. "They generated the maps and then added the movement. In our case, we have all these things on the fly. It makes the work easier, because we don't have to compute anything. The level designers just put something [the game] and it works."

The engine will also recognize that if there's a barrier to the player's movement, the player should not be able to climb. If a wall has a broken ledge, for example, or if there's a vehicle parked in front of it, the actions performed by the game character will change, or climbing will not be allowed, realistically. In one playthrough we witnessed, the player approached a concrete wall that had been broken at the top, with rebar sticking out of it. The game allowed the player to climb the wall, but the climbing movement adapted to the contour of the broken wall, and the character actually reached out and put a hand on the jutting rebar, as if using it to pull himself up. An animation created by the game designers, to be sure, but accessed on the fly by the game engine according to its own internal logic.

"At the beginning, the level designers were scared," Kulon says. They felt that allowing the player to literally traverse everything would add an infinite amount of work to ensure that when players did traverse these things, they [would find] something interesting to interact with. But those fears eventually subsided. When comparing the work required to make all of the game world feel interesting with the increasing demands of fine-tuning the hook system, the designers embraced the challenge. Dying Light became fully traversable.

"It was so cool," says Kulon. "When the first prototype was running, everyone was laughing their asses off. It was so much fun."

00022.mts.still001 The Dying Light art team.

Another zombie blockbuster

The excitement at Techland is palpable. Every person working on each part of the game seems to feel as if they're on the verge of something great.

The walls are lined with Dead Island memorabilia. The studio building, with its still-under-construction new addition, was built on the back of that game's success. But for Techland, there's more to do.

"[Dying Light] is the biggest game we've ever done," says Binkowski. "Being able to pull all that stuff that we want to pull, that's impossible without so many people working on it. We're dealing with it. We're improving our processes. I'm really happy about how we learned to improve communications with big teams. There's still a lot of things to do, but it's definitely better than it was a year ago."

For Techland, the process has been iteration and improvement, and that continues. What started out of a small box in a bazaar in Wroclaw, has now, over a decade later, come close to the brass ring. Each step along the way provided more knowledge, ability and technology. Now, with Dying Light, Techland feels that its star is ready to go on the map alongside the industry's biggest AAA studios.

Now, with Dying Light, Techland feels that its star is ready to go on the map alongside the industry's biggest AAA studios.

"The way we produced Dying Light and the relationship we have with our publisher is allowing us full control of how the product is put together, polished, and how we execute the vision," says Marchewka. "Which is what we wanted with Dead Island. Now we can express our vision, make sure that it's going to be the best game we can create."

After that, Marchewka says he thinks the sky is the limit.

"I think what's always driven us was kind of a simulation of reality. I'd like to see our games be very realistic. ... In 20 years maybe we won't even need [VR] glasses anymore. It'll be old stuff. Games will be something you can actually feel, taste, live. All the experiences and the emotions we can put into our games, to feel like a real person, but in a fantasy world with the greatest adventures of your life for a couple of hours.

"I think that would be our way to go, to give players a couple of hours — maybe 20, maybe 30 — to spend in a completely different reality, where they feel really powerful, where they feel like they're saving the world, where they feel every bone and every part of their body. That would probably be the way to go. I don't know what's next, but it will be quite an achievement."

Images: Tom Connors, Jimmy Shelton / Vox Media
Video: Jimmy Shelton, Tom Connors, Caleb Green, Phil Pasternak

Polygon goes to Poland

Polygon's features team traveled to Poland in the first half of 2014. Over two weeks we covered more than 300 miles. We visited some of the largest and oldest cities in Poland. We met with nearly two dozen teams, and spent time in the homes and workplaces of the individuals making games in the heartland of Central Europe. Check out all our coverage here.