Piece By Piece: Assembling They Suspect Nothing
Oculus Developer Blog
Posted by Jon Davies, Narrative Designer at Coatsink
March 22, 2018

Greetings, human. I’m Jon Davies, Narrative Designer at Coatsink, and the writer of 2017’s Augmented Empire and They Suspect Nothing – our latest VR game about deception and robots, coming soon to Oculus Go. I’ve been working in games for over a decade, and playing them since the late 80s, when they were loaded from cassette for the cost of two mules and a turnip.

Coatsink is an indie developer and publisher based in Sunderland, UK, and we’ve worked with Oculus on a number of titles over the past four years. So today, for a giggle, I’m gonna tell you all about how we pieced together They Suspect Nothing; the process, pitfalls and practicalities of creating a coherent collection of mirthful mini-games. Buckle up.

They Suspect Nothing is a collection of lively VR experiences in which the player performs a series of bizarre tests to prove they’re totally a robot. As the last human in a world controlled by bungling robots, you’ve managed to elude capture by scavenging spare parts into a makeshift disguise. But now the time has come to ‘prove you’re metal’ and become a fully-fledged citizen of Arcadia.

They Suspect Nothing is comprised of twelve experiences split across three hubs: the low-tech Scrap Yard, the high-tech Human Detection Division, and the mysterious Operator’s Archives. To illustrate the process of creating each experience, I’m gonna use one of the earlier games – Open Part Surgery – to demonstrate the steps.

In Open Part Surgery, the player operates on broken robots, using different tools to extract the various components. The player’s told which tool to use and which component to extract. Sounds easy, right? Only problem is, you have no idea what the tools or components are called. So when your instructor tells you to, “Use the Jiggler on the Optical Bladder,” you’ll have to check the reactions of your fellow robots (including the patient) to see if you’re about to save them or commit accidental roboticide.

The Concept Phase

Okay, part one: Concept. First, everyone on the team ‘imaginates’ a bunch of ideas for stuff that could make a fun VR game. (And no, we don’t use that word.) Every idea has to meet several key requirements. It must be:

  • Funny, or provide lots of opportunities for humour.
  • Simple to understand, so players can dive right in.
  • Comfortable and show off the strengths of VR.
  • Re-playable and encourage players to share the experience and compete with each other.

Open Part Surgery hit these points well: it offered plenty of opportunities for daft names and robot reactions; the controls are simple (aim and click); the gameplay utilises a large interactive 3D space from a comfortable stationary position; and there’s plenty of scope for re-playability with variable names and assets.

So after the short review process and a democratic show of hands, we decide which games to take forward and the Spike Phase begins. And yes, we do use that word.

Open Part Surgery Concept Art Development

The Spike Phase

Groups of three – a designer, programmer, and an artist – pick one of the games and build a rough prototype over the course of three days. We use Unity for most of our projects – including the Esper series, Augmented Empire and Shu – and it’s great for rapidly building prototypes and getting games onto headsets to test. In a few hours we can have testable gameplay running on-device. In a few days we have a playable prototype. So after three days of people hammer-and-chiselling away, we get something that looks like this:

Initial prototype for Open Part Surgery

Yeah, yeah. Hush, you. It might not look like much, but the goal here is to nail the core gameplay and see if the idea’s worth taking further, and three days is just long enough to unearth any fundamental conceptual issues. These block-outs also provide a basis for the artists should the idea be taken forward.

With a bunch of prototypes together, everyone plays everyone else’s games and provides as much feedback as possible, ranging from the constructive (‘It’s great, but the controls feel loose,’) to the ominous (‘It’s great, but my wrist aches,’) to the completely irrelevant (‘It’s great, but that robot keeps looking at me funny.’) This feedback is then collated and we assess which games to take into full development.

Initial prototype for CataclySim

Sprint Development

Each game is developed over a four week schedule, broken down into one-week ‘sprints.’ This process uses a larger team of artists and programmers with weekly progress reviews, and each week has a specific goal to ensure development stays on-track:

Week 1: Do It Again, Only Better

During the first week, the designers and programmers recreate the prototype from scratch, while addressing the issues raised during playtests. Meanwhile, the artists concept what the proper game will look like. I sat down with our Technical Art Director, Torger Nærland, for coffee and a bagel, who took me through the art development process.

“Once we have a good idea of what a scene needs to contain, we draw up the rough concepts,” Torger explains. “From here, the team narrows-in on a few specific designs, based on readability, style and tone, and we iterate further, before the agreed-on design is passed to the 3D artists.

“The 3D artists begin by creating a block-out – a texture-less white-box version of the scene – which is then passed back to the concept team for a paint-over to inform lighting, colour palette and scene composition. The decisions made here are usually determined by narrative events and the story we want to tell. The Human Detection Division, for example, is a cold, sterile and imposing place, so the paint-overs had to reflect this. In fact, all the initial hub spaces in They Suspect Nothing are generally tall and imposing because we want the player to feel like a tiny, crushable insect next to the real robots.

Concogtions, Concept Art Development

“While this whole process takes time, it establishes important information, such as the composition and colour palette, and ensures we have a concrete frame of reference before the 3D work begins. For 3D asset creation, we created initial style guides and a workflow tutorial for our internal wiki that our different artists could reference to maintain a consistent style.”

Week 2: Gameplay Refinement and Balancing

The team fine-tunes the gameplay through testing, while adding the final artwork. By the end of the second week, the core gameplay is locked. After resting entirely on their laurels for the first week – shooting nerf guns and rearranging the figurines on their desk – now it’s the designer’s turn to step up. As Designer Ella Shaw describes, getting the difficulty curve right is perhaps the trickiest part:

“Unity’s great for tweaking values and moving around objects but, after working on the same game for several weeks, you eventually get a bit too good at playing it. The hardest part is actually taking that step back and trying to imagine going in fresh, as a new player; what they’ll try to do, what they need to understand, and how you as the designer can make it an enjoyable experience in its simplest possible form, on Easy. User testing with inexperienced players was really helpful in this regard because it highlighted the problematic areas: the stuff they simply didn't get, which we figured would be obvious.”

Week 3: Sound Design, VFX and VO

By the end of the third week, the parameters for Easy, Medium and Hard will have been implemented and a first-pass of the voice-over written and recorded. I sat down with our Narrative Designer, me, to learn more about the process of writing the story and dialogue.

“I kept in mind a number of core principles,” I said. “For a start, I didn’t want the game to be a series of games with jokes thrown in… I wanted the game to be a series of jokes with games thrown in. With They Suspect Nothing, we’ve created huge and highly-detailed interactive spaces to explore, with characters to meet and tons of fun stuff to find.

“I also wanted to make the player work for the content; to hide it and make it discoverable. One of the reasons The Stanley Parable worked so well was because of that same sense of discovery – you feel like you’re being disobedient and testing the limits of the game, only to realise the game’s two steps ahead. The same applies to lore and backstory: if you plant it in the background, where it can be learned by accident, the player will see it as a reward for experimentation… rather than a lecture they’ve gotta sit through.

Promotional Artwork

“There’s also a fine line between something being funny and not funny, which usually falls between hearing a joke for the first time and hearing it ever again. I mitigated this by sticking to a few rules: First, don’t repeat good material. If you’ve got a joke you’re really proud of, roll it out once then retire it forever. Second, use sounds for frequent emotes, not dialogue. Used sparingly, short sounds survive repeated plays far more than hearing the same (long) joke. Third, if you have to pull funny dialogue from a finite well, make sure it’s deep and drawn from sparingly.”

Week 4: Testing, Polish and Bug Fixing

Although there’s a full week dedicated to solid testing, we’ll continue to polish and tweak all the completed games until the project’s end. Meanwhile, all the hubs and characters are built by the artist and animators.

Open Part Surgery, finished game

I grabbed Torger and Ella one final time to get some tips for other developers working on Oculus Go.

“Use all the space available and guide players so that they look around them fully,” Ella says. “It's easy for players to focus on what’s right in front of them and forget that literally every other direction’s available. It sounds simple, but easy to overlook. One of our early prototypes focused all the action in a single point dead ahead of the player, so their head was locked rigid while they stared at a single point. It was a great game, but we nearly didn’t develop the idea further simply because the initial viewpoint didn’t capitalise on VR’s strengths. Bring the player into the middle of the action, and fill that space with cool stuff.”

Debugging, finished game. The initial prototype viewed the action from much further back. The final game puts the player at the centre of the action.

“The Oculus Go is designed for portability and simplicity, so just remember to work to the system’s strengths,” says Torger. “Know what’s achievable, select an appropriate style, and get your basics spot-on – colour, lighting and composition – as these are what will truly make your game visually distinctive and shine.”

What’s next for Coatsink?

After helping to bring Boneloaf’s Gang Beasts to PS4 last December and releasing our cartoon platformer Shu on Nintendo SwitchTM in January, we’re looking forward to assisting PHL Collective bring their frenetic Clusterpuck 99 to the Nintendo SwitchTM soon. We’ve also partnered with Fierce Kaiju (makers of the Viral Quarantine series) for an ambitious upcoming VR venture along with continuing our partnership with Oculus, so keep an eye on our Twitter and Facebook for all the latest news.

As for me, I’m heading to GDC 2018 (in accordance with the prophecy) to talk about assembling They Suspect Nothing in more depth. For a country fop who never finished Fantasy World Dizzy, it’ll be a surreal honour loitering around the Oculus booth in Silicon Valley, so come say hello. I might even let you play. For a turnip.

They Suspect Nothing is currently available to demo in the Oculus booth at GDC 2018 (#401, South Hall) and is coming soon to Oculus Go.