I forgot to mention some things in my post on Virginia. I’ll express them here, and discuss the wider gaming context.
Games are increasingly borrowing various techniques from the medium of film. What’s going on here? Without delving into the motivations (expanding the audience to achieve financial growth, envy of film’s recognised cultural significance) there’s lots of stuff to examine:
- Technological improvements mean games can look better. Often, this means photorealism. Indirectly making games more film-like in appearance.
- Camera techniques discovered by filmmakers are relevant for 3D games.
- Film-like storytelling techniques, and emphasis
That’s what I’ll focus on here. Stories can be about anything, but stories about people are the most interesting, and challenging. Take Virginia, it’s largely about the characters.
Integrating good characterisation into a videogame is hard design challenge. You can make game characters look cool, give them interesting backstories, write compelling lines of dialogue, have talented actors to voice them (and even play their roles physically, with motion-capture). But what about the most important character, the protagonist? That’s where the difficulty is.
Protagonists are controlled by the player. What can a player do with a character? That’s going to be limited by the input mechanism. Games are great at making controllable character run, jump, punch, kick, and shoot. But if you’re making a real story, you need some more subtle sorts of human interaction to be possible.
So personal interactions, like conversations, are usually handled with cutscenes. Or canned dialogue, played in-game at relevant moments. Maybe a forking script with multiple choices.
Compared to the degrees of freedom offered in your typical game, in terms of physical action – movement – this is going to feel limited. A silent protagonist is one way to sidestep this problem.
This is perfectly serviceable when your protagonist is alone in the world. But when he’s surrounded by other fully-voiced, gesturing, and emoting characters, there’s a dissonance there. What’s wrong with Gordon Freeman? Characters talk to him and he doesn’t reply. You can’t reply. Is Gordon just shy?
Maybe you don’t want to have a shy protagonist.
The ultimate solution: let the player speak into a microphone, or type speech on a keyboard. Why that can’t work: it requires solving hard artificial intelligence.
Look at physical interaction, think of small, high-precision actions. Fine motor control. Say you want a character to shuffle a deck of cards. For an NPC, you make the animation and control it with a script. With a player character, you’d typically give the player a context-specific prompt. When the player character is in the right place, facing the right direction, the player will be able to press the ‘interact’ button, and then the card-shuffling animation will play.
A much more limited sort of interaction, compared to the protagonist’s standard repertoire of abilities which can be used with a high degree of flexibility, blending actions together (running + jumping), controlling the duration of the action, providing nuanced modification on an action–shooting isn’t just shooting, it’s shooting + aiming, and you can aim at anything (or nothing).
The ultimate theoretical solution for this stuff is more feasible than solving hard AI. Your deck of 52 cards, and all the other crap on the tabletop, could be physically simulated. And you could use an Oculus Touch controller or Leap Motion to detect hand and finger position to control a physically-simulated game hand. Then you can freely touch, grab, lift and push game objects in a natural way.
But it’s still a hard technical challenge to get natural, adept movement. Motion controls are notoriously imprecise. Modelling physics in games often exposes the imprecision of game movement control, making your character seem incredibly clumsy, bumping into tables and knocking chairs over. Many games play this up for humour, like Surgeon Simulator or Octodad.
If you’re trying to tell a serious story about complex human relationships, you’d have have a cast of wonderful animated characters relating to each other in profound ways, according to your masterfully-crafted script. But a mute, clumsy protagonist running around like an idiot would seem… a hindrance.
So, Virginia’s approach. For dialogue, our silent protagonist doesn’t appear strange, because everyone’s silent. There’s no dialogue. Like a silent film.
And every player action (except movement, which is kept to slow, natural walking speed) is just a context-specific canned animation. Point and click, and see what happens.
They embrace the constraints of a limited possibility space. Expansion of interaction-freedom seems to be the natural evolutionary pathway for game design, and Variable State moved in the opposite direction. The player is quite restricted. The player’s action isn’t always the motive power of the game’s progression. The game has its own agenda, you’re along for the ride.
Unsurprisingly, Virginia’s design choices and execution have been polarising. Critical reactions to this have been mixed–the PC version’s Metacritic score is 74. I maintain that Variable State has done a damn good job.