Thunderbolt logo

Uncanny Valley: The Real Problem With Progression in Our Industry

When the first pictures and videos started rolling in the for this current generation of consoles, we were all floored by the graphics. Interesting points were raised; just how real could games get? The common problem that a lot of CG movies face is “uncanny valley”, the point where digital images become so close to real, they look bizarre. Now that we’re all used to the graphics standards of today, it’s easy for us to pick apart visuals – a low resolution texture in Gears of War, an ugly piece of lighting in Crysis. Video games are still behind movies when it comes to looking real. But what about feeling real?

screenshot

If you were to pop in the original Half-Life, and then play Half-Life 2 directly afterwards, I’m willing to bet the first thing you’d notice was the visual upgrade. Valve’s classic shooters, however, are just that – shooters. While Half-Life 2 upped the ante with great graphics, realistic physics, and characters who emoted with a passion akin to real actors. What do all of these things have in common? They’re all differences the player sees. Aside from the interactive physics, Half-Life 2‘s achievements were entirely cosmetic. In fact, the same thing can be said about nearly every game in every genre, ever. Sure, plenty of titles innovate, but we’re still far off from a truly massive impact.

The difference is control. Originally, games were played side-to-side or up and down, two-dimensionally. The shift to three dimensions, in classics like Star Fox for the SNES or Doom on the computer, was absolutely staggering. Really though, all these years later, are we doing much more with it? Moving in three dimensions has really just stayed that, moving in three directions. When games can truly control in three dimensions, we’ll see the next revolution. But how will it happen? What we need is a game that allows for true interaction on a level that’s never been seen before. Many games have successfully drawn us in with one feature that connected players to the game world; the next one will have to use all of them and more.

screenshot

What we need:

Interactive objects.
Ever since games started incorporating basic physics, ragdoll and environment physics have almost become the norm. Half-Life 2 capitalized on the impressive Source engine, with realistic body ragdoll and plenty of clever puzzles based on the environment. Another Source game, Portal, brought the actual player into the puzzles, incorporating the physics of the main character into the game, not just objects in the environment. Both of these games rely on their engine as a gameplay element, however. The best examples of physics are always accidental. They appear more realistic when games don’t use them specifically for puzzles or gimmicks; for example, a horrific crash would probably have more impact in an adventure game than Burnout. We go to see action movies to see all guns blazing, but the impact of a single shooting is far greater in a drama. It’s the same principle here. Realistic physics should surprise the player, simply being a simulation rather than a sideshow. It’d be nice to see a physics engine used for more than just barrel-chucking, too. The recent Crysis is getting closer, with nearly every object being interactive. It’s still nowhere near an all-encompassing physics representation, however.

Real characters.
Mass Effect was dead on with its representation of cinematic conversation, with hundreds of characters that were fully-voiced and eerily expressive. This will be hard to take, further, however. Imagine a game with as many people as say, Grand Theft Auto, where everyone had a unique voice and story? Ridiculous, say the skeptics. Imagine recording all of those voices! How would they even fit on a disc? What would characters say? Well, for one thing, storage shouldn’t be an issue. If the PS3 is using the massive Blu-Ray disc, wouldn’t that be a perfect way to start using some of that space? Granted, voice and writing talent would be a challenge, too, but I’m sure that the developers of Pong never dreamed that games would become as large as Grand Theft Auto III. This is an area that will always see room for evolution. Once we’ve given millions of characters voices and stories, then what? What will they say when their dialog trees have all been used? Until we invent a program that can accurately think like a human, and put words together like a human, we’ll always be able to make progression in the area of character interaction. For now, though, clever writing and good acting is what we need more of.

screenshot

Better worlds.
Morrowind and Oblivion came close to depicting entire societies, but ultimately came up short in the census. There are less people in Cyrodil than in an average neighborhood, but at least they were all developed and written (more than 5 voice actors would have been nice, however). But society still hasn’t been depicted in a game realistically, outside of perhaps Sim City. Would Grand Theft Auto‘s cities really let criminals fly helicopters and wield machine guns? How come every store in Final Fantasy sells the same stuff? Economies, politics… it’s all confusing stuff, and games have barely scratched the surface. This could branch two ways; we could see a linear, story-based game that simply depicts more social issues, or an open world game with a society that evolves along with the player. MMOs like World of Warcraft let players interact and form guilds, as well as trade. This allows the game to constantly evolve; imagine if something like this could happen offline, as well as online.

Control.
The Wiimote is a huge innovation for controllers, and yet we’re still using it for inane waggling minigames? Realistic physics would be made better with realistic control. Physics doesn’t have to just apply to things that are required to progress, like puzzles; physics just needs to simply exist. The most immersive world is one where, like ours, everything exists. It doesn’t need a purpose, it just is. I don’t need this can of Sprite to solve any quest, and spilling it wouldn’t reveal the secret to opening some locked door. Stuff just happens. Until we can move away from “press A to open door”, games will control in relatively the same way. The trouble is, how do we control real? Where on the Xbox controller does it let Master Chief fiddle with a blade of grass? Does the PS3 Sixaxis controller allow Ratchet and Clank to sidestep an attack fluidly, without relying on a “dodge” button? Until now, we’ve relied on pre-animated moves for everything. We’re far away from allowing on-the-go animation and control, but some solution will arise. Will it be in a year, or twenty years?

screenshot

Graphics aren’t the only thing we have to consider when arguing how real games are. However, I feel that it’s pretty clear that we’re nowhere near the point where we need to worry about the “uncanny valley”. Games are rooted in concepts so basic, so linear, that even the largest Grand Theft Auto or Elder Scrolls can’t fool us. Until significant and arguably ridiculous evolutions are made, video games are stuck strictly in their own context. Which, I add, is hardly a bad thing. Innovation inside of a box is still innovation; especially in an industry with a box this big. It will be the greatest games, however, that find their way outside.

The author of this fine article

is an Associate Editor at Thunderbolt, having joined in October 2006.

Gentle persuasion

Think you can do better? Write for us.