We talked to two of the creators of Virtual VR, about the process of developing the game for a mobile headset and how many drafts it took to get the story straight.
Take it from me, and the rest of the brains behind the site you're currently perusing: writing a story is hard work, and writing one for other people to experience and enjoy is even more laborious. A good story requires more than relatable characters and a plot to hook the readers; it also requires iteration.
It's the same process writing the plot for a game set in virtual reality. I talked to Samantha Gorman and Danny Cannizzaro, two of the creators behind the award-winning Virtual VR, about how designing and writing a narrative-driven game is kind of like putting together a play, or a movie. The Daydream VR-exclusive tells the story of a virtual world where robots are officially using humans for fun. The final product is an outstanding example of how effective VR storytelling can be, even if the device is small and the story is short.
How did you two decide on virtual reality as the platform for this particular story?
Cannizzaro: Samantha did VR back in the early 2000s, which is kind of what started it.
Gorman: I worked from 2002 to 2010 at Brown University and led classes for storytelling and art. I also studied the medium as an academic, so that led to the inspiration behind V-VR. I started to see the next wave of the industry and kind of wanted to respond to that. We started to do some prototypes, and there was a hack-a-thon that we entered where we did a broad concept.
Cannizzaro: There was all this hype and language around it that was a little overblown. All these promises that we wanted to respond to. Sam wrote an article, and we made some videos, and ultimately the response we put the most effort into was to create this game about virtual reality in virtual reality.
Gorman: For this project, it needed to be in VR to make the points we were talking about regarding design for VR and thinking about narrative in VR. There's a point where I'd been at a festival panel and one of the curators said they didn't think long-form comedy could work in VR. So, I wanted to try that.
Cannizzaro: We like working with new media — VR was pretty wide open about what could be done.
"I'd been at a festival panel and one of the curators said they didn't think long-form comedy could work in VR. So, I wanted to try that."
What came first: the narrative or the art?
Gorman: It's a symbiotic process. For this particular project, I did some pre-writing at the same time Danny was developing concept art.
Cannizzaro: In all our projects, we try to tie the work into the device we're making it for. Once we knew this was going to be on Daydream, a mobile VR platform that has specific constraints, we were able to make it for its constraints. We knew the controller would have orientation, but wouldn't necessarily have position tracking. Then, that starts to influence how [the player] will grab objects, which influences the moments you can create.
How did you settle on Daydream VR over Gear VR?
Cannizzaro: When we had started development, Gear VR didn't have a controller — or any announced controller. We were looking at it and trying to find that balance between accessibility where there can be a lot of headsets, but there's still that immersive experience. The controller was one of the hard limitations; We needed that kind of interaction over tapping on the side or a game pad.
Daydream was hitting that sweet spot with the possibility of going broad audience and having the controller and immersive interactivity.
Gorman: We're updating it, too, for the new headsets coming down the line.
Cannizzaro: We'll have a mini expansion that's optimized for inside-out tracking.
How long was the development process for a three-hour title like Virtual Virtual Reality?
Gorman: We got a three-hour game done in roughly eight months. The timeline for development was probably not a standard time for development.
Cannizzaro: At the hack-a-thon we did, we ended up producing a video, and that just sat for eight months or something. We thought it would be an interactive video project. The Indiecade that year is probably when we first started talking to Google.
Gorman: We started the conversation with Google about the roject in about 2015.
Cannizzaro: 2016 is when we kicked off.
Gorman: Again, our timeline isn't necessarily standard.
Cannizzaro: If you look at something like Oculus Story Studio, they have many people working for a couple of years on about 15 minutes of content. We looked at the constraints of what we could do and realized we were trying to do three hours of content with a six person team in about eight months.
One of the design decisions we made early on is that the writing and voice acting would be high quality, but we probably wouldn't have the resources to do a lot of custom animation. The game ended up featuring a lot of animation that's all physics driven.
The main robot, Chaz, doesn't have anything hand-animated. He's just always trying to balance himself like those Boston Dynamics robot dogs because of that. It made him feel like you could always grab him and interact with him. Most of the other characters also have different mouths and things that move in sync with the audio.
What kind of tools did you use to create the project? Did the characters come first, or the narrative?
Cannizzaro: We did a little concept art, but one of our teammates is really good at Blender and would whip out models as we needed them. Between him and I, we did all the art assets. He did much of the modeling, and I did much of the level and art design.
The game started off flat, but we realized early on there is just not much performance. When you have to do virtual reality, there's so much rendering. We decided we weren't going to attempt to be photo realistic; we're going to push it towards a cell-shaded, stylized art direction.
We decided we weren't going to attempt to be photo realistic, that we're going to push it towards a cell-shaded, stylized art direction.
If you look back at Game Cube games, they look dated except for Legend of Zelda: Wind Waker, which looks beautiful because it doesn't go for realism.
Gorman: The main thing that helped performance is that we don't have any textures in the game.
Cannizzaro: When you try to upload a texture to the GPU, it can be a bottleneck on mobile phones. But instead, we painted everything onto the vertices of the model and used very flat colors. Our levels are also minimal, which meant we could hit the 60 frames-per-second frame rate much easier.
We also tried to get lighting working well. Lighting was a way we could set an emotional tone more effectively, and having control over that was important to us.
We also did a lot of vertex distortion and displacement of things like the vacuum when it pulls everything. We found that it works well, performance wise, on the mobile phone. It's not a solution to every game, but it worked well for us.
How long did it take to settle on the narrative of the game?
Gorman: Sometimes I get information on what is possible and what is doable, and I'll do a combination of free-writing and do a diagram on a dry erase board. But inevitably, as we work with the tech to figure out what's possible and what's satisfying, I think I had roughly the month of September 2016 to plan things out and finalize the narrative. We had to write the scripts and do the voice recording in November.
Cannizzaro: How it would work is: Samantha would write a script, we would record it internally and then put it in the game. We'd do some play tests, and then all of us would sit down and look at the timing and what needed to be cut. We tried to do that before sitting down with the voice actors.
Gorman: We had seven or eight drafts of the whole intro sequence. There's one sequence we re-wrote at least three times because we had to get across a lot of information, but it was too expository and too long. We had to figure out how to break it down and remove it.