Valve Talks Half-Life: Alyx And Why Arms Don’t Work In VR

Fans have had a long wait between Half-Life projects, but Valve is finally set to release Half-Life: Alyx, a big-budget prequel to Half-Life 2 designed exclusively for virtual reality technology. Before the game’s release, we sat down to talk with Valve’s Robin Walker and Brad Kinley about the lack of triple-A VR projects, how the coronavirus almost impacted their game’s release, and why Alyx doesn’t have arms. 

Sounds like about four years ago you guys started to work on a big VR game. Did you know right away it was gonna be Half-Life or did you mess around with other ideas?

Walker: You are correct in that it was about four years ago, I think, to this month. For reference, it was about six months before the release of the first Vive, and so we were looking at the state of the VR market. At that time, there were a lot of small studios doing a lot of really interesting creative work, but it was pretty clear that people wanted to see a large, triple-A title in VR. And there are a lot of really good reasons why that was hard for a lot of the companies, to expend the resources necessary to produce that kind of product. So we felt like we could be the company that would build a large, heavily-content focused VR title. We didn't start expressly with Half-Life, but it pretty quickly became Half-Life. We did what we always do, which is take everything we've already got and tried to use that as a tool to figure out stuff quickly. We built a prototype using a bunch of Half-Life 2 assets, and I think we stole glove models from Counter-Strike Go, and got a prototype up and running that you could walk around in.

VR is full of interesting indie projects, but are you guys surprised or even disappointed that other studios haven't tried to do more big game releases in VR?

Walker: I wouldn't say we're surprised at all or disappointed. I mean, the reality of most companies is that you need to do the thing that generates the most revenue, and the reality is the VR market is a smaller market than some other markets. It's growing steadily, and the rate of its growth still continues to increase as well and we're really happy with that. I think we've always been in the fortunate position of having no external investors, no one who has any control or ownership over our company other than the people who work at it, and so we're free to make what we consider good investments over the long term that may not necessarily pay off over the short term. As a company, we've been working like that for two decades. It's how we were able to build things like Counter-Strike, which started off not making money for many years before it became wildly successful. To us, this is more about investing in our long-term growth.

When you started working on a game that people could play for long periods of time in VR, what were some of the big problems you thought you would have to solve?

Kinley: We test as early as humanly possible, and so there was a point really early on, just after the team had gotten together, where we were throwing a whole bunch of stuff at the wall and then testing it nearly every day, seeing what was working and exposing problems.

Walker: There was an enormous number of problems we had to work through in terms of how to how to translate Half-Life gameplay into virtual reality. In all honesty, I think that part of the process turned out to be a little easier than we were expecting because Half-Life's gameplay just happened to lend itself well to VR. It was why we ended up going with Half-Life in the first place. I'd say locomotion was probably the biggest thing that we were expecting to be a problem, but it turned out a lot better than we thought.

I don't know if you ever played Budget Cuts, but it was one of the first VR titles. Before the Vive shipped there was an early version of that floating around, and we played that. Budget Cuts made us realize that we were wrong in assuming that teleportation was going to be destructive to the player experience. It was an opinion we'd formed from watching people play in VR, but there's a much greater difference between the experience of playing VR and the experience of watching VR. There are multiple places that can be quite jarring to a spectator, but when you're in the headset you don't notice them at all. Teleportation is one of them. It's not as big a deal as it seems when you're playing. Another one is arms. It's immediately apparent when you watch someone else play the game that you don't have arms, but when you're playing the game you don't notice that at all because in real life you don't really notice your arms either, your brain edits those out of your experience. 

Were there any aspects of developing a big game in VR that surprised you? 

Walker: In the end, it turned out we had a whole set of other problems we weren't anticipating. When you say you want to move somewhere, say you enter a room and you want to teleport through the room to go to some other point in the room. As a human, you internally, subconsciously understand how you're going to get there. If there was a table and a chair and a bunch of other stuff in the room, you'd know you would walk around the table. You're not going to jump on the table, walk across the table and jump down, right? If there was a Coke can on the floor, you're not going to trip over that. You can ignore that. You don't think about any of this, you just do it automatically. But, in every room in Alyx, there's like 20 or 30 physically simulated objects, and so we really had to figure out when a player says they want to move to a point, what are they thinking? Where do they want to go? The reason we had to figure that out is because there were constraints we didn't want to put on you. We could have easily said that you can't get on tables, but we didn't want to do that. So that took a lot of code.

You mentioned that people don't notice arms when they're playing. Did you playtest a version of the game that had arms at some point?

Walker: Yeah, we messed around a bunch with what the actual representation of your arms were. In the end, we settled on invisible arms that we use for physics detection, so we can tell if you put your hand in a drawer and shut the drawer on them and stuff like that, but we never reached the point where we were at the accuracy level that we felt we needed to be so they work with everyone. We don't know where your arms are. We know where your hands are, and we know where your head is, but there's actually a significant amount of variation in humans about the various length and movements between those points. The unfortunate reality is if you get it right, people don't notice it, but if you get it wrong, it stands out an enormous amount. However, it was interesting how much of people's awareness around various things like that fell away the moment they were in dense environments. The less dense the game was, the more you had time to think about those sorts of things like your arms. 

Do you have any interesting examples of something that happened in a playtest and how that inspired you to change the game? 

Walker: There are the large-scale ones, like the amount that we expect you to explore your environment as you move forward. Another one was seeing how much people really responded to using both their hands at the same time. We went all in on that. There are a whole bunch of places in the game that require you to use both hands at the same time while also moving your view around. But there's also just lots of little moments, like people trying to open something that we hadn't been planning to make openable, or people picking up a hat on the ground and trying to put it on their head and we say, 'ugh, why can't they put hats on?' The game is full of them.

Kinley: Reloading pops out to me. It's not a button press. It's a physical gesture, and that took tons of testing and there was all kinds of feedback around it. People that played the game and got a little bit of mastery around being able to reload and doing the gesture really enjoyed it. Being put under pressure is way different reloading than just reloading if you're in an empty room. 

Talking about the narrative, Half-Life hasn't traditionally been very funny, but it seems like there are quite a few funny parts to Alyx. Was that an intentional choice to make this game funnier or is that just a product of the Valve staff at this point?

Walker: There are two parts to that. One is that Half-Life has always had a dark sense of humor. Like no elevators has ever not killed anyone in the Half-Life universe. So we knew there was going to be some humor there anyway, but this product needed it a little more than previous products. One of our goals was to build the VR game that hopefully helped everyone understand why VR was an interesting platform to have experiences you haven't been able to have before. On the narrative side, we wanted to make sure that there wasn't some reason you might want to play it but couldn't. One of the things we were worried about was the degree of horror or fear in the game. We know that VR is much more immersive and compelling than non-VR games, but at the same time, we didn't think that you could eliminate those elements from Half-Life and still make a Half-Life game. We employed a whole bunch of design tactics to deal with that and help players get through the scariest parts. One of those tools was the narrative use of humor. It was a conscious choice to do some of that in certain places, but it wasn't an overall, 'Oh, we should make a funnier game,' or anything like that.

Obviously, the Coronavirus is the big headline these days. Has that affected Half-Life Alyx's release in any way?

Walker: No, it could have been. I think we all realized, in a way your brain sort of musing on things, that if it had been one week earlier it would have been really scary in terms of release date, but we happened to just finish our last real content production stuff and then the week after that was the week we realized, 'Hey, everyone should start working from home.'

Now that you have made this full big game for VR, are you excited to do some other projects in VR or are you're excited to go back to normal PC game development for a bit?

Walker: Valve is a place where you sort of need to ask everyone on the dev team that same question. I think overall, this was really fun to make, in a way that was like returning to an old friend. I think there are a bunch of us who would like to keep building things like this, but one of the strengths we have as a company is the flexibility to respond to what our customers think of the work we've done, and so we've tried not to make any decisions about what we're going to do next until we get some more data. 

Kinley: People have been very excited to return to this IP, and it has been enormous fun making this game and we're all really excited to see people play. I've got a bunch of time booked next week to just cruise the internet and watch streams.

Half-Life Alyx releases on March 23, so stay tuned for our full review. In the meantime, be sure to watch this collection of gameplay videos or read our feature on how several fans are working together to create their own version of Half-Life 3.