Welcome to 2016! For our first blog of the year, we’re putting on our Designer hat.
One of the questions we get asked the most is what makes designing virtual reality games different to flatscreen games. So, we’ll walk you through some the design challenges and opportunities we’ve faced over the course of the development ofThe Assembly– our first-person interactive story in VR.
From first steps to growing pains
After we’d finished movingThe Assemblyover to its new graphics enginearound this time last year, we invited some expert gamers to our studio to test the game and provide some valuable player feedback. We then used the tester’s observations from these playthroughs to help shape our design decisions for the game.
Our goal for The Assemblyhas been the same right from the very start: to create an immersive, atmospheric game that doesn’t overwhelm newcomers to VR. What we found from these tests was that player behaviour – even among experienced gamers – is different in VR than for flatscreen games.
For example, in flatscreen first-person games, it’s not uncommon to walk right up to the end of an in-game area strafe across it while mashing the action button, in the hope of finding a secret area. This is known as the ‘needle in a haystack’ or ‘Wolfensteinwall’ approach. In VR, however, that never happened – not even once. This is because VR mimics the real world so comprehensively, with your sense of presence and spatial awareness carried across to the virtual environment intact. As a result, players respond as they would in real life, rather than the gamified behaviours we see when playing flatscreen games.
We gave our testers an early version of theE3 demo buildto play, comprising ofThe Assembly’s first two levels, which introduce our protagonists Cal and Madeleine, plus the murder-mystery puzzle level set around a banquet table. This allowed us to see how players would perceive the environments we’d built, as well as serving as a good testbed for our narrative delivery systems, interaction system, and first-person movement controls.
We were pleased to find that in Madeleine’s intro level – wherefree movementis limited – all these systems meshed together and worked well. However, we soon discovered that to make these systems work as successfully in other levels, we’d need to take a second look at some of our designs.
Follow the trail, eat the path
In Cal’s intro level, set in a laboratory, players need to follow a series of steps to identify a particular virus found in a specimen sample. Players immediately had fun simply exploring the virtual environment, however we found that reminding them of the sequence of actions they needed to perform to progress the game’s narrative (known as “breadcrumbing”) could feel intrusive.
So, how did we solve this issue? Well, naturally we did what anybody would do in our situation and applied a bit of feng shui.
Essentially, we re-modelled the laboratory by employing real-world architectural rules that assist flow of traffic and pathfinding… which is the very essence of feng shui. This turned out to be an essential way to encourage players to feel comfortable in the space we’d built. Players were then better able orientate themselves and thus make a path to the game’s objectives more easily.
Conforming to expectations can be a good thing
While re-designing our systems, we had to be aware of not over-complicating the player experience. As VR reflects real life so closely, we wantedThe Assemblyto remain quite pure and true to life, rather than overly gamified. It therefore became important for us to strike a balance. We needed to give players plenty to interact with, while also ensuring that key items – ones that drove the narrative forwards – would be easy to identify.
We soon realised that creating a consistent language for our visual prompts – to help players recognise whether an in-game object is incidental or important – is much more critical in VR than for flatscreen games. While this kind of consistency is always important for a videogame, players seem to be much more forgiving when it comes to flatscreen games than in VR. Breaking expectations feels much more contrived in VR, as it becomes very frustrating seeing something that you can't interact with in VR as you would do in real life.
By focusing on creating a very consistent visual language, we’ve been able to design much more engaging, dynamic and rewarding levels.
More show, less tell
One of the major issues we had with our visual language boiled down to timing. We wanted to make it very clear to players exactly when they needed to interact with an object in order to further their progress, without making the game overly linear or removing player choice.
Our first solution was to implement a traffic light system. If an object glowed red, that meant it might be interactive at some point and the player should come back to it later. If an object glowed amber, that meant the object was incidental – players could interact with it, but doing so wouldn’t drive the story forwards. If it glowed green, that meant it was the current objective and that players should be interacting with it straight away.
This system was not particularly convincing and we quickly discovered that it coddled the player to the point that it restricted their freedom of choice. We decided to ditch it, instead implementing an in-world labelling system.
Now, if a player focuses their gaze on an object, a hovering interaction icon is displayed. The icon itself is context-sensitive and based around real-world actions – that is to say, it shows what your hands would be doing in real life.
We’ll be sharing more information about the in-game UI we’ve developed forThe Assemblyas part of a future blog, but for now it’s safe to say that our current visual language is as immediately understandable as we always intended it to be.
Actions speak louder than words
By improving the way in which we direct players towards their next objective, we were able to reduce the reliance on V/O or dialogue for progression. For example, previously you would learn about the results of a certain experiment that the Assembly had conducted by listening to audio recordings and written reports. Now, we enable the player to find the results themselves via maps, charts and specimens.
Speaking of maps, in earlier versions of the game we provided players with a HUD they could call up to display the layout of the Assembly’s bunker – a very traditional way to help the player navigate an environment, in videogame terms. While that works well in flatscreen games, we found that in VR it was a bit clunky.
We went on to replace the HUD map with ones that they find on walls at various points in the bunker, which is much more similar to how we’d find area maps in the real world. It turns out that players respond much better to a wall-mounted map, as it reinforces the believability of the game’s world. This shouldn’t come as a surprise – as Is often the case in VR, a solution found in the real-world is often superior to a traditional gaming method.
We’re very pleased with how we now emphasise realistic interactions and physical actions that change the environment as a way to drive the player forwards through the game. This makes the player feel much more in control of the experience, rather than being a passive observer.
We would love to give more detail about all the work we’ve done in this area, but we don’t want to spoil the game. I mean, we could tell you… but we all knowhow that one ends.
Question everything, including yourself
Throughout the development of the game, we’ve taken aniterative approachtoThe Assembly’s gameplay. This means that new features are designed, implemented, improved and refined in sequence after extensive play testing by both the team and gaming experts we invite to the studio (lovingly referred to as “outsiders”).
One of the early design choices we made was to introduce phobias as a gameplay element. We gave both Cal and Madeleine their own psychological Achilles’ heel that we thought would translate into interesting and exciting player experiences in VR. For example, we gave Madeleine a fear of heights and designed areas that forced her to confront her phobia.
Naturally, we wanted these areas to heighten (geddit?) the terrifying feeling of acrophobia (that, not vertigo, is the correct term -look it up, kids) without causing extreme discomfort for the player. After all, as we said in the very beginning we wantThe Assemblyto be as accessible a game as possible.
After multiple iterations of this concept, we found we couldn't strike the balance between ensuring players who genuinely suffer from acrophobia felt comfortable in-game, and for the feature to have an impact for those without any fear of heights in real life. We definitely didn't want to leave players feeling like they needed to rip off their headset, nor did we want the feature to come across like a shallow gimmick.
While we feel that this type of experience definitely has a place in VR games, it’s a feature that would benefit from being at the core of a game - perhaps even havea whole game built around it. As part of a larger game, however, it could detract from the overall experience or even as a cut-off point that certain players wouldn’t be able to pass.
In the end, we developed a much more nuanced alternative that really compliments the game’s story. Now Madeleine's fears are played out on a more psychological rather than physical level. No spoilers! We just can't wait for you to see exactly what we mean when you come to play the game.
See you to be me
One of the key design decisions that we discussed as a studio was whether or not to give the player a physical presence in the world (aka an avatar), so that if a player looked down then they would see their torso and legs, as well as have arms and hands that were more-or-less permanently visible.
Far from amplifying the feeling of actually being in the virtual environment, we found that having an avatar actually breaks immersion sensationally, because its movements in-game wouldn't match up 1-to-1 with your movements in real life. However, we expect that this is only a temporary situation and while it’s not possible to get the avatar to replicate your exact movements at the moment, as technology improves this will not doubt be realised in short order.
Additionally, we originally designed Cal and Madeleine to be different heights, as early tests showed that it conveyed the uniqueness of VR quite convincingly. However, after further investigation we learned that it messed with player's depth perception and thus broke immersion.
While switching sizes between a mouse and a man would work very well as the shift is dramatic, small human-sized shifts comes across as uncomfortable and strange. We now assume that VR headset makers have come to same conclusions too, as they currently recommend a 1-to-1 scale and consistent camera height throughout the course of a single game.
The end’s in sight
We've now reached a point where we're going back out into the wild with our current version of The Assembly for more player feedback, so as to prepare us for our final push.
We’re very close to the stage where all that’s left in terms of the game’s development is some final fixes and a round of polish to bring the game to quality bar we’ve set for it. Don’t forget that you’ll be able to play it for yourself atEGX Rezzed in April.