Tips for VR Games

Enough with the philosophical considerations, for now. Here are a few practical tips:

Scale 1. The scale of the world has to be realistic. You should feel like you are the right height (unless you want your player to be a child, as in Among the Sleep). Head movements should not be amplified (unless you're using redirecting techniques).

The easiest way to achieve Scale 1 is to make sure that 1 world unit is 1 meter. The field of view should exactly match the field of view of your HMD. In an ideal world (or big industrial VR system) the distance between your two eyes should also be correctly measured and used. Your brain picks up all these cues; you might not be able to create or maintain presence -- and even make people feel sick! -- if you don't strictly follow this rule.

Know your hardware. Know the range of tracking: Can my hardware track translations, or only rotations? If the tracker also reports positions, up until what distance? What's its precision? When does the tracking data stop being usable? Know the field of view: As you are supposed to follow the "Scale 1" rule, you shouldn't distort the virtual field. If the field of view is narrow, the user will have to move her head much more to see around than with a bigger field of view, and might miss some important action in the periphery. Know the resolution: if you want the user to read information, she will have to come much closer with a low-resolution display than with a high-resolution display. As with Android development, your game might end up running on different hardware. We will soon have a HMD war with lots of HMDs, each with different characteristics. Using tools like MiddleVR will help you work with many different VR systems.

Have a consistent viewpoint. If your game is a first person game, avoid playing cinematics or making the player drive a vehicle from a third person view. It breaks immersion.

Break habits. Longtime video game players have bad habits: when they wear an HMD they will stand still, as if they're seated in front of a TV. Those who are less experienced with games people will naturally look around. Gamers need to unlearn the constraints of current games. In a tutorial, you should force the user to look around and move his hands. Your game should also take advantage of those new possibilities. For example, in a recent game prototype I worked on, we had enemies appearing to the right, left and above the player; there is no joystick/mouse to navigate and look around. It forces the user to look around and aim with his hand to get all the enemies. In another game prototype I worked on, the only interactive object is a candle in a very dark environment. This was a great way to force the player to explore: he naturally took the candle, and used it to explore the dark environment, pushed some objects, and burned others to solve the puzzles.

Try to keep the player active. In Heavy Rain, for example, you're almost always playing. There are numerous cutscenes that look like videos, but suddenly you'll have to perform an action. If you don't have the game controller in your hands at this moment, you'll fail the action. This forces you to always be alert and ready.

Another very interesting feature of Heavy Rain is that the game happens in real-time, which means you sometimes have to take decisions and act fast: do I have to shoot this guy before he kills my partner? You're forced to act quickly -- and as in real life, once you make a decision, you'll never know if it was right or not.

Create realistic puzzles. Still another example from Heavy Rain: You're in a rush and have to phone a room in a motel. Can you remember, in less than 15 seconds, the room number? Just as in real life, you have to dig it from your memory while experiencing a lot of stress.

Finally, work hard on presence. Presence is not easy to build. Start light, test often. Build presence slowly, make small additions, test again. The experience is what happens in the user's brain! Your simulation enables the experience: it is not the experience! Presence should be natural. Observe the reactions and modify the game. Don't throw in every possible gimmick just because it will make a cool video. A lot cool videos really end up being terrible experiences.

Conclusion

There is of course a lot more to be said about developing a VR application, but I hope this article got your attention on some fundamental points. I leave you with this quote that I hope you will apply:

"Our approach is to treat virtual reality as something quite new with its own unique conventions and possibilities that provide a media where people respond with their whole bodies, treating what they perceive as real."  - Mel Slater