Steam Dev Days VR Update

As the developer of an upcoming Oculus Rift game, I was eager to see the VR sessions at Steam Dev Days. They ended up being the clear highlight of the conference for me. Since the whole “off the record” thing seems to have been interpreted pretty loosely so far, I thought I’d give a general update to other devs who were interested in VR but weren’t able to attend. This isn’t a complete overview; instead, it’s just the stuff that was most salient to me, someone who was listening for information that hasn’t come out before.

The conference featured talks from Michael Abrash and Joe Ludwig from Valve, Alex Schwartz and Devin Reimer from Owlchemy Labs (developers of Aaaah! For the Awesome), and Palmer Luckey from Oculus, whose talk was probably the most fun of the entire conference. A good portion of the content was advice for making a comfortable VR game, much of which was already old news to those of us who have been following Oculus for a while. Make sure your game runs fast, minimize player acceleration in-game, don’t override the user’s camera controls, etc. All this has been collected into a handy new VR Best Practices guide.

Michael Abrash’s talk was notable for the amount of confidence that he projected. Valve’s VR demo is pretty incredible, and Abrash used it to illustrate that most of the big problems (on the visual side, at least) can definitely be solved. Achieving a real feeling of presence in a virtual world is now just a matter of time and good engineering; it WILL happen. It’s not so clear exactly when, however. Abrash kept saying “in two years” or “by 2015 or shortly thereafter”. I interpreted this as his deadline for reaching this mind-blowing golden standard of VR; I’m hoping that Oculus chooses to ship earlier with a merely excellent product. [EDIT: And indeed, Palmer Luckey posted on reddit to say that "Valve's presentation is not necessarily a reflection of our internal timeline... It would not make sense for them to hinge their credibility on our ability to ship on time."] Palmer was pretty confident about the future of VR too, and it’s clear that both Oculus and Valve believe that it will be a huge part of the future of entertainment.

The nature of the relationship between Valve and Oculus was not entirely clear to me. They’re certainly working closely together. Oculus talked about getting R&D help from Valve, while Valve named Oculus as the clear front-runner among consumer VR solutions. (Valve said they had no plans for a consumer product and that their demo system was not shippable due to its requirement that users plaster their walls with tracking markers.) On the other hand, Valve made it clear that their VR software work was not Oculus-specific, and mentioned that they want to be ready for when competitors start to arrive on the scene. I don’t think that means that they’re betting against Oculus in the short term, though. My impression was that Valve was taking a practical (if enthusiastic) attitude toward cultivating VR gaming, and they recognize Oculus as the obvious vanguard in that movement.

Palmer’s talk was called “Porting Games to Virtual Reality”, and his message was that porting “doesn’t really work”. Instead, make content that’s designed for VR! We’ve heard this message before, but it was a pleasant surprise to see it presented so starkly and with such candor. Oculus doesn’t know exactly what they want from developers (that’s our problem to solve), but they have a high standard, and that’s ultimately great news.

There was also a Q&A session with all the speakers after the talks. Again, a large part of the discussion had to do with common mistakes in VR games and how to avoid them. The speakers seemed to agree that FPSs and similar games would not translate very well to VR, especially in their current form (with fast movement, circle-strafing, lots of forced movement from explosions, etc.). I consider that good news, since it points the way towards a wider variety of experiences in VR. There was also a consensus that games involving small, nearby objects (e.g. a virtual board game) felt great, especially with positional tracking. The speakers agreed that player input was still a hard, unsolved problem in VR, and Palmer made it clear that (for now) they’re targeting a seated experience with gamepads. Lastly, there was discussion about how great a multiplayer experience felt when multiplayer head-tracking and simulated eye contact was implemented.

One audience member asked a business question, wondering how to justify making games for a new, small audience of VR players. I was surprised by the honesty of the speakers; rather than project false enthusiasm, they were practical. The general answer was “it all depends on how many units you need to sell to make money”. It’s an easier path for a small team than a big AAA outfit (another reason why indies should favor the Rift). Even then, the Owlchemy devs mentioned that, despite a fantastic attach rate of around 8% among Rift owners, they still didn’t directly cover the costs of the extra month of development they spent on the VR port. Of course, the outlook will improve when the Rift actually releases. I’m still optimistic.

Overall, it was a very impressive presentation from Valve and Oculus. They convinced me that VR is going to be able to deliver on a technical level, and they left me feeling more inspired than ever to create games for the Rift. At the same time, they tried to be very clear about what wouldn’t work. They talked openly about what they had successfully accomplished, but they also talked about the challenges that lie ahead. It didn’t feel like a bunch of marketing folks desperately trying to recruit developers. Instead, it felt more like they were sharing the news of a fait accompli: VR is happening, and soon.