Last week, our guest poster, Evelyn, wrote about her experiences with multi-user interactions in VR. This was super interesting to me, and it made me start thinking more about a different kind of multi-user VR experience. It’s fair to say that very few people have VR devices right now. As a result, every time I’ve gone to any kind of VR meet-up or demo it’s been a weird experience where many people sit around unable to interact with the one person who is actually wearing the headset. The experience seems fundamentally solitary – even when you are surrounded by people in physical reality. Sure, those people *could* interact with you in VR *if* they had their own headsets (as Evelyn discussed last week), but they probably don’t, and, anyways, they are actually right “here”. This has gotten me thinking about VR experiences involving multiple users, but a single headset.
Last week, M proposed one such multi-person experience. Using Tiltbrush, a VR drawing app, M used Vi as a physical model in her virtual space to create a flamboyant virtual outfit perfectly fitted to Vi’s real life body shape. One could easily imagine such custom drawn clothing being a standard for avatars, and perhaps even being custom printed for wear on physical bodies. (And we streamed the process live, which you can see archived on twitch.)
Watching M’s outfit drawing happening in the third person, I started imagining Pictionary-style ‘picture guessing’ games for Tiltbrush where the use of the headset and the controllers was split between two people – one person who could only view the virtual world and one who could only interact with it. We tried playing these games together earlier this week.
In the first variation of this game that we played, the controller user draws a mystery object (without being able to see it themselves), and the wearer of the headset (and the only one that can see the drawing) has to guess what was being drawn. You can see Vi’s curated video of the twitch stream of us playing that game below.
As a group, we turned out to be much better at this game than we anticipated. The ability to make drawings life-size and to scale, as well as the ability to take advantage of all three dimensions by doing things like drawing a beach on the ground and a castle up into the air seemed to more than compensate for the inability of the sketcher to see what they were drawing themselves.
In the second variation of ‘picture guess’, the headset wearer tries to get everyone else to guess a word by directing the controller user how and wear to add lines to make an illustration of the object. Everyone watching and the person with the controllers could then try and guess what was (supposed to be) being drawn. I found the physicality of this version fascinating. The headset wearer would say things like “draw a cylinder into my hand” or “draw a circle around my head” or “draw a line from here (pointing) to the ground”. They were gesturing in physical space at something that only they could see in virtual space. This version was also much more conducive for a big crowd, as none of the guessers could see the object in question, so nobody had the big advantage of the headset wearer in the first version.
One thing that I liked about these games is that they allowed multiple people to interact with the same setup by splitting up the headset and the controllers. Thinking about this idea, the idea of a visual version of the game Marco Polo came to mind. I like to think of this as the first version of Marco Polo suitable for deaf people. For those unfamiliar with it, Marco Polo is a version of tag, where the tagger has their eyes closed and has to tag the other players by sound only. To aid with this, they are allowed to say “Marco” whenever they like, at which point all other players are required to respond “Polo”. For VR Marco Polo, I made a simple VR world, where pulling the trigger of a controller caused the controller to appear for 1 second in the virtual world. The controller players were required to respond to a cue (we continued to use the calling of “Marco”, but it could be a non-audio cue such as a hand movement) on the part of the headset player to pull the trigger on their controllers and reveal their positions. You can find the simple code for this on my Github and try this out for yourself here.
Despite the incredibly simple mechanic, we had a lot of fun playing this game and quickly developed strategies. Players using the controller would try to hold them far out to one side when they triggered them and then quickly move in the other direction once the trigger period was over. Initially, this seemed like the ability to “fake” your position a little bit might be giving the controller users a bit of an unfair advantage, but this was compensated for by the highly constrained gamespace provided of the VR environment. The headset wearer was generally able to corner controller players in an amount of time that felt neither too long nor too short. I liked the active-ness of this game and the fact that it enabled up to three players to interact in the environment at once with a single setup by splitting up the controllers between people – something that wasn’t really possible in our improvised TiltBrush games.
In addition to the games above that we played and explored last week, I know of one existing game where multiple participants can play a game involving a single headset, “Keep Talking and Nobody Explodes” (also playing without VR). In this collaborative game, the headset wearer can see a bomb and needs to deactivate it. However, only the non-wearer has access to the convoluted bomb deactivation instructions. The headset wearer describes what they see and the ‘bomb deactivation expert’ uses that information to walk them through the deactivation steps, with the goal of deactivating the bomb before it explodes. The interaction between the players is entirely verbal. The mechanic is based on the idea that each player has access to different information that they need to share with each other.
This idea of having players with access to different information seems natural, if not fundamental, for games involving only one VR headset as that person really is partially in a different world from everyone else. Incompletely shared information between players is a fundamental part of many games both in and out of VR. It’s an essential component of all of the games that I’ve described so far in both their VR and non-VR variants. The incompletely shared knowledge in all versions of Marco Polo is the locations of the players. In regular Marco Polo, this is enforced by simply closing ones eyes, while in VR Marco Polo the blindfold is a VR headset which gives the player access to incomplete information about the real world controller location. In all picture guessing games (in and out of VR), the secret word is an example of hidden information known to only one player. This is interesting as it makes the most fundamental bit of hidden knowledge not dependent at all on who can see what. Indeed, that fundamental knowledge switched hands in the two versions of picture guess that we played, from the person drawing but not seeing the VR to the person seeing it.
We’ve been spending some time around the office describing other VR games and activities of this format, from more variations of classic games to collaborative activities to competitive shooters. This includes a lot of games that a family might play together after purchasing their first VR headset. In the short run, the ability to have a ‘games night’ involving a single VR headset engaging several people at once could be an early selling point of VR equipment, by making VR no longer necessarily solitary or requiring a large initial outlay of purchasing a headset per person. It seems to me that this style of half-in/half-out semi-shared VR experience is one that has been mostly ignored and is deserving of a lot more thought.