In his talk at the DigiPen Institute of Technology, Valve VR developer predictably didn’t share any specifics on the company’s yet to be revealed VR game, but he did offer a glimpse into his approach to VR game design as it relates to the upcoming title.
As we noted earlier this month, Valve developer Kerry Davis held a talk at his alma mater, the DigiPen Institute of Technology, last week, in which he promised to talk about “developing the company’s upcoming flagship virtual reality title.”
Though Valve has confirmed their so-called “flagship” VR game is due to ship in 2019, the company still hasn’t actually revealed anything about the title. While Davis did offer an interesting case-study focusing on the process of VR interaction design, he made it clear up front that he wasn’t going to be sharing any specifics about the upcoming VR game.
“I will not be making any product announcements, and I will not be sharing new details about projects currently in development. Believe me, I am dying to share all the exciting things that we’ve been working on at Valve. It’s killing me, but today is not the time or the place for that.”
When asked whether Valve’s upcoming game would still launch this year, Davis said he didn’t actually know the release date of the title himself.
I do not know. And that is an honest answer. Because, like I said, Valve is a very autonomous department… that is not my choice to make, and I haven’t been in the meetings that are making that decision. My daily goal at Valve is to keep pushing the project forward, writing code, fixing bugs, making it awesome… and the day will come where the word will come through the office and they’ll say ‘Ok we’re announcing tomorrow’… and that’s when I’ll know about it.
Through the rest of the talk, Davis was careful not to share any revealing details about the title (other than the fact that it will include doors, apparently), though he did present an interesting perspective and case-study of the not-as-easy-as-it-seems process of VR interaction design.
Throughout, he focused on the difference between building a game that ‘simulates reality’ versus one which ‘simulates the experience of reality’.
Interactions have been pretty much the same throughout [non-VR] game history. It’s the big dividing line between the game world and the real world. If you want to interact with something in the game, you have to do something in the real world to make that happen, through an abstraction layer. So if you want your character to jump in the game, you have to press the B button on the controller to make him do that. Now with VR, I hypothesized that we can eliminate nearly all of these abstractions. Because if you can put the player into the virtual world themselves, why not just let them interact with the world directly just like they would in the real world?
But this is the question you have to ask first… I didn’t realize this until a long time later: what are we really trying to simulate? Are we trying to simulate physically accurate real-world interactions? Or are we trying to simulate the player’s experience of those real-world interactions? They’re actually two different things, and the distinction is important. To put it differently: are we simulating reality or perception?
Davis argued that the developer’s choice to focus on simulating reality or simulating perception should be intentional, so that the rest of the design can fall into place around that choice.
You can do either… if you’re making [the VR game] Hot Dogs, Horseshoes, and Hand Grenades […] you’re going for perfect physical reality simulation; that’s fine—that’s your choice—make sure it’s intentional. Know what you’re trying to simulate. Set your goals and make sure what you’re doing fits those goals. Increasing the accuracy of the simulation does not result in a more realistic experience […].
Focusing on the seemingly simple case of trying to make a usable door in VR, Davis outlined the iterative process of starting from a simple ‘press button to open door’ approach to something specially made for VR which included interactive physics, a handle, a latching mechanism, and one-way hinges. As the realism of the door’s functionality increased, it became harder for players to use because, Davis argued, VR is still missing tons of cues and haptics from the real world. Although the more complicated version of the door was more realistic functionality speaking, using it correctly was hard for players in a way that real doors aren’t.
He stressed the importance of playtesting interactions to put assumptions about player behaviors to the test, and make changes if things don’t work out as expected.
Only rigorous testing will tell you what suits the goal you’re trying to make. You have to test the heck out of it. And don’t get mad at your testers, don’t claim they don’t know how to work the real world. They do! It’s your fault they can’t figure out what they’re supposed to be doing. And we can’t assume that real-world interactions automatically translate to VR. We still need to teach players somehow what they can and can’t do in the VR world. Or if you choose not to do that, make that an intentional decision and know why you’re doing that.
Davis said that because Valve’s VR game is striving to focus on simulating the perception of reality, rather than reality itself, dialing back to a less complicated version of the door actually improved the player experience because playtests showed that it was easier to use and understand. In the real world doors might be functionally complicated but they are easy to use. Thus, a virtual door that’s easy to use, he said, more closely simulates the player’s ‘perception’ of the real world than a virtual door which is more physically and functionality accurate but is more challenging to use .
Davis’ nearly hour long talk offers up more nuance in this process than is reported here. If you’re interested in VR interaction design it’s worth a watch to understand how Valve is approaching the topic, and you can luckily catch a full recording of the talk here.