GamesBeat: You were also coming up with the notion of an open world action-RPG at the same time?
Smets: I think we heavily underestimated those aspects.
GamesBeat: Just how big a world it would be?
Smets: Yeah. The first Killzone was a very linear shooter. We were used to that. We hadn’t had to deal with a gaming environment where the player could go anywhere. That was an aspect we underestimated. We did have the ambition. I remember the first version of the map we came up with. It was about five times the size of GTA. It was very big. I think it showed what the ambition was, though. We wanted the world to feel really big. We wanted you to be able to explore, to finish the story and still be at 35 percent completion.
It took us quite a long time to realize that it wasn’t so much about the size of the map. It was figuring out what we wanted the content density to be. We didn’t want a huge map that felt empty, and you just had to walk everywhere and got bored. At the same time, we didn’t want everything to be too close together. We needed you to feel this was an open world that you could explore.
GamesBeat: We’re talking about a creative process that happened eight years ago. All these ideas seemed very new at the time. I think everybody was coming up with post-apocalypse worlds, but then you had this idea of a post-post-apocalypse world, a place that’s green and recovering into life.
Smets: Yeah, a lot of post-apocalypse worlds were coming out, but they all painted this dark and somber picture of what the future could be. I think because we came out of the dark and somber world of Killzone — 10 years of Killzone is a lot of Killzone. We wanted to take a more hopeful look at the future. The world can also be beautiful in the future. Why not paint that picture? That’s how we came up with the world of the post-post-apocalypse. The first humans mess everything up. Nature takes over. Then we start again. That was one of the things that inspired the whole team, this idea that we could start from scratch, with a blank canvas.
GamesBeat: You had this worry that there might be another game out there that was too similar. Ninja Theory had Enslaved?
Smets: This was really early on. We’d just had the round of pitches done and everyone was excited about the Horizon pitch. I remember getting a text message from JB [Jan-Bart van Beek], the art director, full of cursing. He had a link to some pieces of concept art that they’d used to announce their game. It wasn’t a lot, but the few pieces that they’d released showed an overgrown city, with beautiful nature. It had a woman lead character, looking very cool. And it also had a little flying drone robot. All of these were key ingredients we had in Horizon. We were thinking, “Oh, we thought were on to something super new, but look what they have.”
We weren’t happy about that, but at the same time Killzone 3 was just about to ship, so we did have a lot of time to think about our next steps. Basically, the rest of the studio jumped in on shipping Killzone 3. I think we got really lucky there. By the time Killzone 3 shipped, which was five or six months later, Ninja Theory had released Enslaved, and it took a really different direction than the one we were planning. At least that meant we could still explore Horizon.
One thing that did come out of that is we decided to develop a second idea in parallel. We thought this situation was a bit scary. We shouldn’t put all our eggs in one basket, so let’s do a second prototype that’s a little bit less ambitious, a little closer to our existing skill set and technology. That was called Dark Science. It was a brawler game about a mad scientist. It had humor in it. Completely different tone, very different game. It was much closer to what we could already do.
GamesBeat: It sounds like it didn’t have a female lead character.
Smets: No, it didn’t. It had zombies. Those were our zombie characters.
GamesBeat: Did you pitch both of these ideas to Sony?
Smets: We shipped Killzone 3. We started a small team, around 20 people, and they started developing both Horizon and this Dark Science concept for about five months. That’s when E3 came up. After E3 we have this great meeting at Worldwide Studios, this post-E3 summit, where the studios get together and they share their knowledge, their ideas. I really love what Shawn was talking about around Worldwide Studios yesterday. As developers it’s great to be part of this sort of family of studios. We are all very different in how we work and the games we make, but you feel you’re in this group, and you can learn a lot, particularly from teams who have a slightly different way of working. You can get some really good ideas.
We figured, well, we’re part of this bigger family of super experienced developers — Naughty Dog, Media Molecule, Santa Monica — so let’s ask them what they think about our two concepts. We made two presentations and showed them at the meeting there. It was really great to be able to talk to them. When you work for a long time on a concept, it’s hard to see everything very clearly.
I think the feedback was pretty clear. All the love went to the Horizon concept, but they did point out that it was pretty crazy in terms of ambition. “It’s an open world action-RPG and you guys have zero experience with that. It’s super huge in scope.” On the other one, Dark Science, to make that game would be much easier. It was the safer choice.
Then we went back home to Amsterdam and we asked the whole team that same question. We gathered everybody in the canteen downstairs and we showed them the two pitches, the two concept presentations. Then we asked them to rate all of the concepts. We gave them five questions. Which one fits the best with our technology? Which one has the most risk for character development and storytelling? Everybody said that Dark Science was the safest to go to. But then the last question asked which project you wanted to make. Everybody picked Horizon.
We decided, then, to go with Horizon, knowing it wouldn’t be the easiest journey for us. It was the one that the whole team was behind, that we were most passionate about. So off we went.
GamesBeat: Women as heroes in triple-A games are becoming more common in the last couple of years. You guys were talking about this seven years ago, though. At the time, there wasn’t a lot else to look to in that area. How did you decide to do that, and what kind of discussions did you have about it?
Smets: Honestly, we never considered the gender of the lead character. In the very first pitch we had, she was already there. She was a girl. Nobody even questioned it internally. It wasn’t a topic. We were really concerned with getting a good lead character. In the Killzone series we always tried to have these cool hero characters, but in the Killzone games it was always the enemies, the Helghast, who were the stars of the show. We were very focused on building a good lead character. And she was a girl, but honestly that was never a point of discussion. It was only later, once we starting the concepts to more people….
GamesBeat: Outside the company, did you get more questions or reactions?
Smets: I think from the publishing side, they definitely questioned if she had to be a woman. We said, “Well, we don’t think it’s an issue.” What we always do with our concepts is focus testing, concept testing. We figured it would be an issue with the testing population. But the opposite actually happened. The tests confirmed that people were very excited about the gender of the lead character.
It also showed that there were still some problems with Aloy as a lead character, though. Initially she was much younger than she ended up being. She had a bit of a Disney princess, miss-perfect quality to her. That’s one of the reasons why it’s great to do focus testing with people, get a fresh set of eyes. We worked on Aloy for two years, on everything — hair, outfit. I remember watching a video from one of the focus tests. Someone pointed out, “I think it’s great. She looks like a strong character. But how believable is it that such a young girl would be so perfect and taking on these robot dinosaurs?”
It was a good point. We made her a bit older, a bit tougher, a bit louder. We changed her voice. We got Ashly Burch, a great V/O actress. We also got some freelance concept artists to do art for us. Our concept team had been working on her for so long that we wanted to get some fresh perspectives. One of the concept artists really nailed her in terms of style.
GamesBeat: You’re a woman in the lead of a development team. What did having a woman in the lead of the game mean to you?
Smets: Honestly, for a long time, not that much. I thought it was just normal. Maybe that says a lot about who we are as a developer. We’re very international. Half the team comes from other countries, including five people from Spain. In Dutch society, gender equality is pretty normal, and I think that’s reflected in the studio. I never felt I was treated differently in my career because of my gender.
It was more once I became more senior and began speaking to other people in development that I became more aware of the struggles that still exist. I feel very lucky to never have had to face that. As the project progressed, particularly after we announced the game, seeing the love for Aloy really touched me. Of course we were hoping that the response to the announcement would be good, but people really rallied behind Aloy. I think it was only two days after that we saw the first cosplayers putting outfits together. It was really heartwarming. It still is, actually. It never goes away, I guess.
GamesBeat: It seemed like an innovation in the game industry, to have this. She was actually fully clothed.
Smets: How silly is that, right? If you back to our Killzone games, though, I think that ties into our design philosophy. We don’t objectify characters, male or female. We have a philosophy where we want to create realistic worlds. We’re really doing science fiction in that sense, not so much fantasy. Everything needs to look as if it really could have been built. You see that in the outfits of the tribes in Horizon.
We researched, first—we talked to anthropologists a lot at first, to learn more about what defines a tribe. It’s all about resources. They make clothing out of the resources they have access to. For Aloy’s character, there’s a certain type of animals that live around her, and you see all those materials coming back in her clothing. They’re not very advanced, so they don’t have advanced ways of producing textile. Very basic knitting is something they can do. Then other tribes are more advanced, making more refined textiles. There’s one tribe that knows how to work metals, so you have metal objects coming back in their costumes.
That’s something that I think is just a tradition in our studio. Even with the robots, we tried to make them look like they could really exist. There are a couple of engineers on our concept art team who think about, if you wanted a robot dinosaur, how could you really build one?
GamesBeat: Yesterday we heard Todd Howard talk about how it took them 10 years to get to this point on Starfield. For you, the whole game took seven years. Some of this is starting to look a bit intimidating to developers, I think, that when you’re working on an IP from scratch — why do you think it took this long to get it done? It was five years before you got to your reveal.
Smets: Five years, yeah. But it was a lot of fun. I get that it’s intimidating, but it’s also a lot of fun to get to create a whole new world. There are two sides to it, I think. Sometimes you think, “How will we ever resolve this?” but there are also many moments where you say, “Wow, we’ve got something really cool here.” There’s a lot of love and passion happening at the same time.
The first year was a really small group. It’s not as if you put the whole company behind a new project. Sometimes these ideas — you just need to think about them a little bit longer. You can’t just speed everything up with more people. You need that iteration time, time to think things through.