Solving the "unsolvable" issue of mobility in VR
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.
To hear many people talk, virtual reality has an unsolvable problem. To some, it's an issue that cripples the entire VR effort and leaves all the effort and investment dead in the water. I'm speaking of course about handling player movement (and rotation) in VR.
Some people have a natural unease when wearing a VR headset if their character or camera(s) are not stationary in an environment. Early demos on the very first Oculus Rift Dev Kits (commonly referred to as a "DK1") often resulted in users being queasy or dizzy... we all know about this or have experienced this by now. It is known.
Fast forward to Oculus Connect a week ago, and the unveiling of the newest prototype, dubbed Crescent Bay. It's very hard to find someone who tried the CB demo and wasn't blown away by the quality, and for all intents and purposes it crossed the threshold where basically everyone is comfortable and natural in those worlds. People who were immediately queasy with earlier VR devices are right at home when the frame rate is so high, precision is so excellent, latency is so low, etc. Users were able to literally walk around the demo room, freely exploring a bit of virtual space, poking their heads against objects and truly immersing themselves in the world. It's absolutely magical, and that's not hyperbole.
(Again I applaud the Oculus team for striving to make the first consumer version such a highly tuned experience. If you see that demo, you get it. You only get that first impression once)
There's a catch. Of the ~10 demos on the Crescent bay units, only two scenes had the camera moving within the world at all. Epic's excellent "Shootout" demo had a slow steady linear camera path through a paused Matrix-like firefight. It was like standing on a moving walkway in an airport, moving through a 3D mural of powered armor, debris, and mechs (and it was beyond gorgeous). The other movement demo was a similar straight track through an abstract Tron type world, but yeah, only two demos with the most basic and uncontrolled movement possible.
Clearly player movement and camera controls within VR worlds is a huge issue to basically everyone with a stake or opinion on VR.
To discuss this issue I prefer to think there are actually two problems that are often rolled into one. They're absolutely related, but in my mind they're very much two distinct problems. (I am sure there are more established terms for these, but I'm no academic, so I'm going with these)
Problem One - "Comfort"
How players instinctively feel when the world around them moves or rotates.
Problem Two - "Controls"
The control scheme for how the players control that movement.
When people simply say "movement in VR makes me ill", it can be any number of facets of either of those problems, or both. Comfort is very much affected by technology improving, latency and such, but it's not entirely about tech. Some people are naturally more sensitive than others in this regard, and there's less that we can do to impact that. As a designer I'm extremely interested in Problem Two, Controls, the actual input scheme for making a character or camera move through an area. Controls are what we need to nail in order to escort as many people as possible across the threshold into VR. It's the ultimate test of making something "feel" right, and it's what I'd like to talk more about.
*Allow me a slight tangent before I continue. Personally, I don't believe that we will see in-home treadmills, or hamster balls, or slick sock trackpads, etc as a solution for control devices. To me it's a mass market non-starter, a great way to get VR skeptics thinking you're completely out of touch with reality, and an excellent way to have casuals happily dismiss all subsequent opinions from you. Some demo could change my mind, I'm just stating my current opinion on this. I'm not trying to create a Holodeck; It's ambitious enough just wanting to navigate an avatar around a 3D world instead of just set pieces.
On a topic with opinions flying fast and loose it's nice to know there are at least a few absolute truths. Here's one such truth: people have extremely individualized issues with VR.
A month before Oculus Connect, I set out to make a demo that focused on mobility. This particular prototype was about zero gravity grappling hooking around an asteroid cluster, using head tracking and only two buttons. I thought it felt excellent very early in development. I would say around 70% of people who tried my demo felt comfortable with it (not awful, but not a slam dunk by any stretch).
At one point I showed it to someone and it didn't go well, they didn't care for the method by which I controlled rotation of the character, and they gave me some good feedback on my deadzone settings and sensitivity. As he walked away I fired up my editor and created a different build on the spot with several aspects modified.
Two of my old Epic buddies, Nick Donaldson and Nick Whiting swung by to check it out next. Nick D tried my original version and had nearly identical feedback to the prior player. I fired up my newer "low responsiveness" build, and he too found it much easier to use. I considered for a minute that maybe I needed to swap all to my settings to that mode and rebalance for it.
Nick W picked up the headset, still on the "low responsiveness" build and again I heard the comment of, "I don't care for this turning scheme". For the hell of it I fired up my original build for him. The result was night and day. He was zipping around, interrupting grapples in mid move, behaving instinctively how I also was with my very early builds.
The line between, "Ew, I don't care for this", and "Holy hell, don't EVER touch these controls again, they're awesome!" is a very fine, and very individual one.
People's opinions on VR Comfort are very polarized because we are acting on deeply subconscious levels. What we believe, we believe strongly. You can't debate with someone that they're not actually uncomfortable, if they're uneasy, they're uneasy damn it. How people feel in VR is not really opinion, it's a personal fact, it's how they're wired.
In VR some people don't like open spaces, some people don't like lateral movement, some people can't go down stairs, some people don't like being close to large objects, some people can't handle yaw rotation, some people can't handle HUD elements, etc. I've done a lot of caving in real life, but found my personal VR Kryptonite is moving through tight corridors for now (alas, no Space Hulk from me).
Some people quickly get their "VR legs" and adapt, some simply don't.
"OK, Lee, we get it! People are snowflakes, move on!" Yes, yes, ok! Chill out, I have a point...
Time and again the conversation comes up that people are waiting for "The solution" to VR movement control schemes. Why would we assume there is "a" solution to a problem with so many personal variables?
Must there be a magic bullet that flips some binary switch where suddenly every living person can leap about in Minecraft or TF2?
While that's a noble goal, I think many are holding VR up to a far higher bar than any other platform when it comes to expectations of a single universally accepted control scheme.
It's possible that "The" solution looks more like an array of customizable options and control schemes that will naturally evolved as industry standards. Look at nearly any random first person game and you'll find options for sensitivity, auto assist, dead zones, reversing both or either axis, auto sprinting, etc. I'm not giddy and clapping at the idea of option screens (they're a pain in the ass and the majority of players don't even open them) but for the players who need them to flip a critical switch that makes the entire game playable to them, it's a huge deal. It's not perfect, no, but I don't see why it's any worse for VR to have option screens or alternate control schemes than all the other games we play on all the other platforms.
IMO, a "one size fits all" control scheme prerequisite may be unrealistic, unnecessary for the platform's success, and it might be a distraction keeping us from finding an array of suitable individual solutions. I would love nothing more than to see an uber-scheme emerge, don't get me wrong, but we can't wait for every light down the street to turn green before we hit the gas.
Not every game, every genre, and every control scheme has to serve as an ideal first experience for completely inexperienced casual VR users. It's possible that some people are going to have Comfort problems even under the best of cases, it sucks to admit that. Do we not create things because of that factor? Designers obviously don't want to create something that is incompatible with a chunk of people, yet we readily accept that practically every modern game is meant to appeal to a specific subset of customers. Ideally we make what we want to play, and there's no crime in that.
Racing games, fighters, RTS, shooters, adventure games, puzzlers, etc all have dramatically different expectations of control systems; in five year's time I will be very surprised if we don't see a variety of genres within VR that have very different expected control schemes as well. Consider that high character-mobility games could be a genre within VR that some people love, and some people can't tolerate.
One guy at Oculus Connect was showing this swanky Descent-style rogue-like horror indie game. It flies in the face of quite a few assumptions about what can work in VR, and I have to say it was extremely cool (check it out! www.nulloperator.com). I thought, "you go, dude! I want to play your game, I handled the motion really well!" I can imagine some wouldn't handle it so well, but why shouldn't he make a game like that if he's following his passion?
I'm reminded of a moment around 1999 when everyone was passing around videos of this ridiculously slick looking PC FPS shooter named Halo. The announcement that Microsoft had scooped them up and it would be an exclusive for their first console was... let's say "dramatic". People following Halo absolutely flipped. the. hell. out. Would they ship a mouse and keyboard with the Xbox?! They'd have to, because the idea of a FPS with a controller was nothing short of repulsive!
Obviously that turned out well.
The thing to consider about that example is that MS and Bungie did a truly mind boggling amount of work to get the original Halo controls to feel as awesome as they did. Their acceleration curves, adhesion, reticules, movement rates... all of it pounded through usability tests and forged until they ultimately laid the groundwork for literally nations of gamers to enjoy shooters on their TV. It was a huge investment, and it paid off (options screen and inverted mouse testing and all).
Oculus is in the position of trying to launch this entire VR movement. Their reluctance to push out demos that might alienate a segment of players is understandable. They need to appeal to as broad an audience as humanly possible. They have a ridiculous amount of pressure on them to make that first experience for random users be an awesome one that feels completely comfortable. Given the Herculean tasks ahead of them with launching Gear VR, a platform, the Consumer Version of the Rift, etc. it seems highly unlikely that they're going to put the kind of effort into tackling "mobility in VR" that MS and Bungie did with Halo.
This isn't 1999. Back then it used to be impossible for small developers to create and distribute content for consoles, the job had to fall on the shoulders of MS and Bungie. This is not the case now. Dev kits from Oculus are pretty easy to score, and Unity and Unreal Engine 4 both make it incredibly straightforward for devs to experiment with these devices. The unsolvable "Problem Two - Controls" is now a challenge distributed among many tens of thousands of people with dev kits and know-how. It's a hell of a thing to vote against that amount of intuition and passion.
There's another interesting truth about all things VR. The solutions are often incredibly unintuitive. Things that feel like they should be smooth are actually jarring, and vice-versa. Things that seem like they should help immersion like camera shakes or walking bobs actually break immersion. Some really great tips sound kind of horrible on paper. Really until you just try something, you can't know what the outcome of an experiment in VR will be.
Allow me to list everyone who is an absolute authority on Virtual Reality:
How cool is that? How about the list of people who can say conclusively that your idea won't work? Pretty much the same list. There is not a single person alive right now who can say conclusively what can't work in VR. Sure, we know a good deal about things that do work, but the amount we don't know is a vast wilderness comparatively.
If you're a designer, what you can not afford to do right now is listen to everybody who believes their name should be on that list above here. Too many people have decided "VR is not good for X". Don't buy into it. They are coming from a different set of preferences, and you should counter those preconceptions with your personal instincts.
Oculus has an amazing document listing their best practices. I encourage you to really study their points... then intelligently and consciously push against them. They are first people to tell you (even within that document) that the points are only well founded suggestions and helpful hints. Know the rules before you break them.
Even if someone has tried something themselves and tells you it didn't work, it's worth considering that they might not have tried it the same way you would. Perhaps they're in the 20% of people who didn't like what they created, but you really might have loved it? The industry is littered with people who tried and discarded something that was incredibly successful for someone else.
I doubt that it will, but I hope that VR does not become just another format for the exact same games we've seen on all the other platforms. The point of better VR controls is not simply to play Skyrim as-is in VR, it's about making something with even more immersive potential than Skyrim.
I have no doubt that there are some extremely cool game concepts waiting to happen in VR while still coloring within the lines of the Best Practices document. Some great games could happen just expecting the players to stand up and walk freely around in the sensor area, like the Crescent Bay demos. That said, I still can't imagine that in 5 years everything in VR will be chair simulators and "experiences". At some point players are inevitably going to want what we think of casually as "games" now.
Designing within parameters is a critical skill for a designer, but so is pushing the boundaries.
Imagine what VR can be like in 5 years, hell, even skeptics often say "VR needs more time". The problem is that time isn't what solves problems. That future version of VR doesn't just manifest itself because the calendar says "it's time". Experience and developers experimenting will be what solves these problems. People playing early games, getting used to them, getting their VR legs, that's how we slowly erode at these "unsolvable" issues.
To summarize! VR is a far more personal game experience than anything we've ever seen, so don't be afraid of in-game control options. Solutions are often crazy unintuitive, so instead of listening to why things theoretically won't work, try things! If something seems magical to you, run with it.
Thanks as always for reading, I know that was a long one!
P.S. If you're doing things in VR that involve interesting approaches to character or camera controls, I'd love to check it out and compare findings, please comment. I plan on putting a build of my grappling demo on the 'share' site as soon as I'm done with GDC China and GDC prep.