Designing for Virtual Reality

A few months ago, during an interview with this very website, I said, "Nobody knows anything" - a comment aimed at current VR development and design. Of course, that was a slight exaggeration. It's been two years since developers got their hands on the Oculus DK1, and some people - like Michael Abrash - have been working on it even longer than that. So of course some people (quite a few people, really) know something. But I stand by the general thought behind my statement - VR is brand new and most of us working with it are experimenting more than we're applying existing knowledge.

When it comes to designing for VR there are, as far as I can see right now, two major issues, and they keep getting mentioned in almost every article: Consumer comfort and control input. Working at nDreams means working full time on VR, and over the last year or so we have looked at different projects and different solutions for both issues. Below are some of my thoughts on designing for VR.

The Elephant in the Room

"Does it make you feel sick?" is probably the question VR developers get asked the most. The topic is often skirted around, and a lot of people try to avoid it altogether, but it's one that absolutely needs to be addressed. It varies from person to person and even from experience to experience, but it exists, and if VR really aims to conquer a larger part of the market, any element that introduces simulator sickness or a sense of unease has to be removed.

"We only really have one shot to convince consumers that VR is as amazing as those of us working with it know it to be"

I believe that consumers are so used to experiencing digital media in other forms (through TV, console and mobile) and to such high levels of quality and comfort that a bad VR experience can actually put them off entirely. I still think that simulator sickness and unease are discussed far too little and only acknowledged in passing by many developers.

"It might be a bit odd to start with, but you get used to it!" is something I have heard many times over the last few years when trying out new demos. But I think we only really have one shot to convince consumers that VR is as amazing as those of us working with it know it to be. The first wave of games and experiences heading for the open market must be comfortable for the user. Rather than push all the way for realism and immersion, it might be better to focus on user comfort.

In my opinion, the sky's the limit when it comes to what's possible with VR in the long term. The potential application possibilities go far beyond gaming and entertainment. The first prototype nDreams released was SkyDIEving. Since then we've announced The Assembly, a first-person adventure game, and have created two different products for the Gear VR. At the same time we're looking into other application possibilities, and it's actually quite difficult not to get carried away with all of the things we could be doing. But the one thing all of our efforts have in common is to create a comfortable and immersive experience for the user.

Watch on YouTube

As expected, player movement and control are the big ticket items. We found that if the player is able to rotate the character while simultaneously being able to look around with the headset, a disconnect (and with it unease and even nausea) can occur. So we made a conscious decision on Gunner and Perfect Beach to not give the player any control at all over character movement. The player character is in a fixed position in both, but is still able to use head movement to look around the VR environment. We kept the designs simple, we focused on comfort and user experience, and throughout extensive focus and play testing we never once had anyone feel uneasy or unwell. Some of the things we played around with on other prototypes, and in particular on SkyDIEving, showed us that sudden acceleration, rapid movement close to objects and sudden changes in direction could all lead to a feeling of unease. These are all things that can be overcome, partially through better tech (the step from DK2 to Crescent Bay clearly shows this), and partially through control and presentation.

But I strongly believe that this process needs to be gradual and deliberate. VR is in a unique position to try and convince our brains that what's visually presented to us is reality. What VR is not able to do is to completely shut off the user's senses, but what we've found so far is that, if the user is presented with a believable context, the user's mind can ignore certain discrepancies and just enjoy the ride. For example, in Gunner the user is strapped into a stationary gun turret. This sets the scene and the user doesn't question the fact that they can't move. Instead, looking around in a VR environment becomes the point of focus and enjoyment.

"There is a temptation at the moment to go all out, to push VR as fast and as far as possible. I feel this could be the wrong way to go"

There is a temptation at the moment to go all out, to push VR as fast and as far as possible. I feel this could be the wrong way to go. I believe keeping game design and gameplay reasonably simple, and allowing for higher player comfort, is much more important. This is not about playing it safe, but about understanding that the vast majority of our consumers will be new to virtual reality, while many of us working with it have been doing so for a while. VR itself is a jaw-dropping experience for every person trying it out for the first time. Is it possible to create a Call of Duty style game for VR headsets? I believe it is, but at the moment I'm not sure it would be an experience worth having. Running, spinning around and fast target acquisition, not to mention the potentially overwhelming combat experience, might be too much - at least for now.

Control

Another big aspect with VR is how the user controls the experience. Gamers will be used to standard controllers for consoles and keyboard, as well as controllers for the PC. Both have issues. Wearing a VR headset makes the use of keyboard and mouse slightly more awkward than it would be for traditional gaming. Additionally, any gamer will instinctively use the mouse to look around/move the camera. Give the player a controller in a first-person VR game and muscle memory will make sure they use the controller to look around and walk around. One of the biggest causes of discomfort in VR (or at least it is for me) is the user looking around the environment with the headset and at the same time trying to look around using traditional controller input. Decades worth of gaming habits are actually working against VR in this case.

So controlling the experience is the second big challenge when designing for VR. As mentioned above, removing all external control and simply allowing the player to control an entire game by looking is one solution. It works well for Gunner, but it obviously places restrictions on complexity of gameplay, so that even simple things like bringing up a pause menu can become a design challenge.

Watch on YouTube

We felt that it was a good solution for a first wave VR experience. Eliminating the need for a controller, as well as eliminating potential disconnect between player input and head movement, means that player comfort levels are extremely high and the VR experience itself is at the forefront. Other games like Ikarus on the Samsung Gear VR have gone a similar route, allowing players to look at objects and places in the world and acknowledging choice by tapping on a touch pad. Other games don't allow the use of a thumbstick, opting instead for a move-where-you-look approach. All these ideas will take getting used to, particularly for traditional gamers, but as long as the experience is comfortable and enjoyable I don't think anyone will mind.

"Eliminating the need for a controller means that player comfort levels are high and the VR experience itself is at the forefront"

If you watch people experience VR, they almost inevitably reach out to point at elements only they can see, or to try and touch things in the virtual world. This, to me, highlights the Holy Grail of control in VR: hand-tracking and hand control. There are a number of solutions being worked on right now, with Oculus getting involved by purchasing Nimble as well as companies like Leap Motion. In the long run these will make interaction in VR environments more natural, and open up opportunities for game design that we can only dream of right now.

But we're already seeing a range of different approaches to VR design. There might well be certain themes that seem to be popular in VR (space and horror come to mind), but the variety of possible gameplay is immense. I've noticed a return to much simpler game design in many cases, a focus on the essential things, stripping away a lot of the fluff that surrounds traditional development these days. With VR's potential to astound people, however, there is no need to create blockbuster movie moments just yet.

The number of innovators in the VR space is incredible, and the speed at which ideas are developed is something I haven't seen in the games industry in a very long time. It might still be a few months before VR truly spreads far and wide and reaches critical consumer mass, but I for one firmly believe that this will happen. The number of people not only supporting the hardware, but creating an immense variety of games, experiences and other applications points to a healthy ecosystem on the content side, ready to support the entire spectrum of hardware still to come.

Andreas Gschwari is the Lead Game Designer at the VR-focused development studio nDreams. Prior to joining nDreams, he worked as a designer for Avalanche Studios, Starbreeze, Codemasters and Funcom.