Five Years of VR: Lessons Developers Learned from Oculus Rift to Quest 2

Oculus Developer Blog
|
Posted by Hayden Dingman
|
March 29, 2021
|
Share
Quest
Rift
Accessiblity
Locomotion

This month we’re celebrating five years since the release of Oculus Rift. Many of you have been developing VR experiences for longer, of course—since DK1 and DK2, and the Oculus Share days. But we wanted to take this chance to look back on how far we’ve come, and to reminisce.

You can find a lengthier oral history on Tech@, where we talked with people behind both the hardware and software about the last five (or so) years—getting fabric onto Rift CV1, how Beat Saber saved Quest’s tracking, Lone Echo’s beginnings on a gamepad, the prototype version of Oculus Touch with a touchpad, and much, much more.

But in the course of interviews for that oral history, we also asked developers what lessons they’ve learned over the last five years. Useful, silly, and everything in between—their responses are a great reminder of how far VR development has come, and how much we still don’t know. As our Director of Content Ecosystems, Chris Pruett, said, “Every day that you work with VR you’re inventing something new. This is still a very early phase of VR as a medium, and the people working on it right now are writing the grammar for future interactions. That was true five years ago and it’s still true.”

Read on for advice (and a few funny stories) from developers at Cloudhead Games, Polyarc, Cyan, Twisted Pixel, Insomniac, ILMxLAB, and more—and thank you all so much for helping turn VR into a reality.


On Presence

Denny Unger - CEO & Creative Director, Cloudhead Games: After 8 years of development, we've come to recognize just how important raw physicality is in VR. VR can be a passive medium but we're starting to see a shift in attitude towards what VR does so well, which is physically suspending disbelief.

Games come at it from different angles—the elegant simplicity of Beat Saber and SUPERHOT VR, or the blown-out complexity of BONEWORKS and Half-Life: Alyx—but what makes those titles great is their ability to immerse the user wholly and completely in the experience. The more developers can do to engage those "Lizard Brain" reactions, the more engrossed people become.

Just like putting on your shoes to go for a walk, strapping on a VR headset is just something you do to prepare for something physical and players are embracing that mental shift.

Vicki Dobbs Beck - Executive in Charge, ILMxLAB: We've learned that the most impactful content experiences in VR lean into its two greatest strengths—the power of presence and the power of connection. In addition, with the introduction of the Quest, the ability to design in 360 degrees has been an absolute game changer—especially when considering the possibilities with lightsabers!

Shaun McCabe - Head of Technology, Insomniac Games: Two words: spatial awareness. A responsive HMD with exceptional spatial audio makes it so much easier for players to understand the space around them. There are a lot of rules in screen-based games with controllers: keep everything important in front of the player or never, ever, force players to look up. VR allows you to really play with space.

Hilmar V. Pétursson - CEO, CCP Games: Less is more, the immersion and sense of mastery in VR is so strong that you need to factor that in a lot when starting the journey for the player.

Then once the initial awe normalizes, which can take quite a while, you need return hooks like any game—and the return hooks for VR need to be very strong as the activation energy is so high for VR experiences.

Ian Fitz - Creator, Sealost Interactive LLC: Even more so than traditional games, it’s important to keep what’s happening in the game in line with what the player is expecting to happen. In a traditional game, having something not work how you expect can be mildly frustrating, but in VR it can be extremely off putting, anger inducing, and even physically and mentally disorienting.

Grace Morales Lingad - Creative Director, Sanzaru: I was listening to Greg Kasavin talk about Hades on a podcast the other day, and he talked about how game development isn’t just the big stuff. It’s also the millions of little moments, and we have to get as many of those in the game as possible because you never know which is really going to resonate. It’s going to be different for every individual.

Ruan Rothmann - Lead Designer, GORN: In traditional games, designers often rely on convention to set up player expectations. In VR, although those conventions are still forming, players still have a very strong intuitive expectation of how things should work (due to VR mimicking the real world so well) and that is something that a VR designer should exploit. You can get away with so little tutorial and instructional text if you can effectively anticipate player intuition and expectation surrounding interactions. This means VR games have a potential to play so naturally, without the typical hand-holding and instruction we see in 2d games.

A good VR game is a continual conversation between the player and designer. The player is continually asking, “Can I…?” and it’s the job of the designer to keep answering “YES!” as long as possible. This means that you really have to curate what you put in the environment as the player will expect any small object to be interactable, and it is often disappointing when something is not.

Even better is when an interaction is better or more surprising and detailed than the player anticipated.

Ru Weerasuriya - Creative Director, Lone Echo: In some ways, we all had to go back to school when designing for VR. A lot of old notions about gaming don’t really apply. If players see a drawer and expect, “Well, it’s a game, I’m not going to be able to interact with it,” and suddenly the drawer pulls out, they’re like, “Oh my god, what else did I miss?” And they go back, and they start opening drawers and looking inside.

It’s not about what’s in the drawer, it’s about knowing that you are actually truly there in that environment.

On Experimentation and Polish

Tom Kaczmarczyk - Co-founder, SUPERHOT Team: Everything about VR is still new and unexplored—even now, years since this current wave of VR began! We still don’t even know for sure if the currently available solutions for finger tracking open up any break-through design opportunities, let alone all of the less market-ready stuff like augmented reality, eye tracking, myo-electric sensors, or EEG electrodes strapped to your forehead. Everything here is still fresh and exciting and chock full of opportunities for design innovation and for new forms of immersive experiences.

Approaching all these opportunities with the mindset to explore and to look for something that literally can’t be done in any other medium is lots of fun and, in my opinion, a better idea than just trying to translate an existing game genre from conventional platforms into VR.

Callum Underwood - Director, Special Projects, SUPERHOT Team: We’ve learned that polish (on the game, not the country) is absolutely key, store-based marketing is more impactful than traditional VR, and that players will always try to break your game, and that’s OK.

The games that do well are the games that treat your body as a controller, have strong game design principles, clear art direction, and lots and lots of polish. Arizona Sunshine, Job Simulator, Beat Saber, Red Matter—they all managed to pull off different things and excelled in their class.

Mark Choi - COO, MIRAGESOFT, Inc.: It’s critical that every release is polished and ready to be played through, whether it’s a playtest or an official launch. First impressions are one of the most important factors in whether people keep playing or not—so make it count.

You also need to touch base with your fans and understand what really appeals to them. It’s hard to develop and communicate at the same time, but we still read and reply to all the feedback we receive. We ran quite a lot of beta tests before releasing Real VR Fishing. One of the testers suggested turning off the UI to increase the game's immersion—something we’d never thought of, and it actually worked like magic for us. That’s how “Expert Mode” came to exist in Real VR Fishing, and it’s become one of the most beloved features among our players.

Holden Link - Director of Button Operations, Turbo Button: I had the opportunity to share a bunch of our failed prototypes at Oculus Connect 3 a few years back. For every game we make, there are probably a dozen prototypes that we set aside, and plenty more ideas that we end up removing from the games for one reason or another.

Brian Allgeier - Director of User Experience, Insomniac Games (and Creative Director, Edge of Nowhere): Developing in the early days of VR helped us improve how we make games. VR presented unique challenges that required us to get better at rapid prototyping and run frequent user experience tests. Each week we would plan, prototype, and test to learn what was working.

We created a prototype for Edge of Nowhere that allowed players to hide underwater. It was really cool in that you could see the water plane and then by ducking your head, you’d submerge below the surface, forcing the hero to hold his breath while out of sight from deadly monsters. I recall unconsciously holding my breath while playing it and being truly scared!

After several iterations, we ultimately scrapped the mechanic since it did not work well for the third-person game with a pulled-out camera. We also did not miss the leg cramps from doing multiple 10 second squats during a play session. Hmm…maybe there’s an opportunity for a Lovecraft Horror fitness app…

On Optimization

Ruan Rothmann - Lead Designer, GORN: Visually, being extremely deliberate with the amount and location of detail has proved crucial. High res textures can often look very bad on the relatively low-res VR displays, and it’s very hard to have everything look consistently good at high res, where that is something that you can traditionally get away with in 2D games. All particles and effects need to be in 3d as well—no billboarding!

Hannah Gamiel - Development Director, Cyan: Developing for VR has taught us some pretty valuable lessons in asset optimization. It’s one thing to make a game run on a min-spec PC, but it’s a whole different ball game to make your game run on a device that is rendering different images for each eye. Asset optimization (which in turn, makes the game be more performant) is absolutely crucial in that process to ensure that users won’t experience motion sickness.

Tom Kaczmarczyk - Co-founder, SUPERHOT Team: As it turns out, it’s pretty hard to optimize an indie game to run at stable 90 FPS in VR. Towards the end of the original SUPERHOT VR’s production, we were just barely meeting that performance bar on fairly high-end PCs. We didn’t even have enough spare processing power to capture gameplay footage for trailers without the whole game basically grinding to a halt.

Instead, one of our lead designers had to inflict hours of mild abuse onto himself by capturing footage from a special build that rendered one eye only. Those initial trailers turned out amazing but were bought dearly with a substantial amount of vertigo and headaches.

On Motion and Accessibility

Tom Beardsmore - Co-founder / CEO, Coatsink: I think we did just about everything wrong early on! We had a prototype back in 2013, which was (amusingly) titled “The Climb.” We had free climbing, zero comfort options, and fixed animations for death/falling. It was the most nauseating experience I’ve ever had…

I think one of the biggest lessons that we have learned (contrary to my last point) is not to assume that all users want the most comfortable VR experience. A huge percentage of players are happy to experiment and push comfort options to the edge in favour of a better sense of presence and engagement. This has been great to see.

Chad Dezern - Head of Creative, Insomniac Games: We learned that free movement is possible, provided there’s a 1:1 relationship between the player’s real-world actions and in-game movement.

For Stormland, we set out to develop a traversal system that felt exhilarating, so that players would want to explore a big world just for the intrinsic fun of it. We thought at first that we’d need to set some aggressive constraints, but we ended up letting players climb every tower, cliffside, and mountain in the game.

Grace Morales Lingad - Creative Director, Sanzaru: Some of what we’ve learned is almost in defiance of convention. Like teleportation movement—that was not going to work in Asgard’s Wrath because I wanted free-flowing and high-intensity combat. It’s a game of inches and that wasn’t going to happen if you were always at arm’s length and popping in and out. I wanted the characters to be moving constantly, so we had to make free movement work.

Brendan Walker - Principal Engineer, Polyarc: When trying to make a new locomotion mechanic work in a VR game, you’re going to have to iterate a ton on your character physics to make it feel comfortable. This is of course also true for non-VR games, but you viscerally feel the difference when your head is attached to the character pill.

I once thought I was coming down with the flu when iterating on some character movement code. I then realized that I wasn’t sick—it was just that I was getting smacked around and launched into the sky all day due to my buggy code. Even when you have good VR legs, general discomfort will creep up on you. It’s wise to Invest early in making your game controllable in non-VR. Then, once your systems work the way you expect, you can iterate in VR.

Supporting non-VR controls has the added benefit of also making it easier for testers to regress non-VR related issues in your game.

Eric A. Anderson - Creative Director, Cyan: Our early Obduction VR prototyping on the Oculus Rift yielded some fairly fun navigation bugs, while we were still getting used to the best way to implement certain things. (It was the Wild West back then!) Falling through the ground, or getting launched across a map due to a bad physics interaction is a whole lot more shocking in VR than it is in 2D!

Nic Vasconcellos - Chief Button Polisher, Turbo Button: Before starting Turbo Button, Holden and I created a head-bobbing rhythm game called A Night at the Roculus. We were on a tight schedule, and that meant a lot of intense testing in a short amount of time. I don’t think either of us could move our necks for like a week after that. It was a great lesson in VR input and comfort, but people seemed to enjoy playing the game, so maybe there’s no lesson there at all.

Patrick O’Luanaigh - CEO, nDreams: We’ve found that things are often not as we expect. You really need to prototype! For example, in our original SkyDIEving demo, we assumed that collecting items while skydiving would be fine, but that landing on a trampoline that bounces you back upwards again would be horrible.

In fact, the trampoline felt fine (because it was totally predictable and instantaneous) whereas the feeling of hitting collectible objects with your head in first-person was truly horrible. (You naturally don’t want anything to hit your head!)

Thomas Van Bouwel - Developer, Cubism: VR's embodied nature provides an opportunity to make games and experiences that are more accessible to newcomers—not only newcomers to VR, but to games in general. It's easy to forget that in the midst of development though. Keep interactions simple and easy to learn where you can, and don't use more buttons on the controller than you need.

Early on I was mainly testing with other developers and VR enthusiasts, which meant I wasn't catching some fundamental accessibility issues due to most testers being able to quickly pick up how to use the Touch controllers. It was only when tests with non-gamers started going badly, that I realized I needed to simplify Cubism's control scheme. Sometimes tests going wrong can lead to the best improvements.

Bill Muehl - CEO, Twisted Pixel: The cool thing about this stage is that there's still so much potential we as developers can explore and innovate with VR. We're big fans of experimentation and pushing the potential of VR, and that oftentimes means we're at the edge of what feels comfortable, especially early in development.

One thing we've learned from players over the years is that there's no "one size fits all" for player movement as it relates to comfort. To maximize accessibility, while still pushing the envelope of player freedom, we've found it's key to give players as many customization options as possible. This includes settings to adjust turning style and rate, movement speed, tunneling and a bunch of other options.

Based on this approach, we've found that the vast majority of players have been able to find a setting that feels good to them. More generally, it's great that there's an increasingly diverse library of VR experiences to give everyone options across the full range of experience intensity.


If you enjoyed this, check out our lengthier oral history of Oculus from Rift to Quest 2, featuring many of these same developers alongside people who worked on the hardware. Our thanks to these developers for volunteering their time and accrued wisdom for this piece, and many thanks to every developer who’s released a VR game on an Oculus headset. You’ve made VR what it is today.