Guest Post: Craig Stern on graphics, imagination, and immersion
Posted by Rampant Coyote on February 21, 2013
While Jay is out of town, some very talented folks have graciously offered to help out with the articles. Today’s article is brought to you by game author Craig Stern. And the letter G. Craig Stern is an indie developer currently working on the turn-based tactical RPG Telepath Tactics. He is the founder of IndieRPGs.com, and can often be found rambling in short, 140-character bursts on Twitter.
Nathan Grayson of RockPaperShotgun recently had the opportunity to ask a question of Bethesda head Todd Howard: how does he plan to make the world of the Elder Scrolls games feel more tangibly alive? Mr. Howard’s response:
“Everybody always wants more power,” he said. “As a developer, you always want more. How good will it look with more memory? How many people can we put on screen if we have more processing power? But even with all of those things, I think people discount graphics. They’ll say, ‘Well, the gameplay’s what really matters,’ and it does. But I do feel that graphics and your ability to present something that feels new, real, and believable puts people in that environment where they can really enjoy what they’re playing.”
This is not the answer I would have given; not for RPGs in general, and certainly not for the Elder Scrolls series.
The fact is, graphics are not necessary to make a player feel that a game’s world is “tangibly alive.” Game worlds don’t feel alive to us because we’re so stupid that we can’t distinguish between realistic images on a monitor and real life—they feel alive because we willingly suspend our disbelief.
In actuality, you don’t need graphics at all to make someone feel that a fictional world is alive. Books immersed people for centuries before the requisite technology for video games even existed, spawning the popular idiom “lost in a book.” Deeply engaging with the characters and setting of a novel makes the world of that novel feel tangibly alive. When we suspend disbelief, our brains play along—no graphics necessary.
Video games have not changed that basic fact about us. The one game most notorious for completely sucking players in also has some of the crudest graphics in gaming: Tetris. It’s nothing more than a bunch of four-block shapes falling, row by row, down a grid. And yet, Tetris immerses players so fully that it bleeds over into the real world in a phenomenon called (you guessed it) “the Tetris Effect.”
So, does this mean that graphics aren’t valuable? Absolutely not! Good graphics are valuable in a game for the same reason that a particularly skillful painting is valuable: aesthetics give us pleasure. In games, specifically, aesthetics can heighten an already compelling experience, giving us shocking and beautiful things to look forward to when exploring the world.
But Mr. Howard isn’t making a broad argument for aesthetics here: he’s aiming specifically for “real” and “believable.” He’s making a narrow appeal to more photorealistic graphics by leveraging computers’ ever-increasing RAM and processing power.
Photorealism represents a narrow subset of aesthetic approaches to games. Arguably, it’s the one approach which demands the smallest investment of vision on the part of the creator (to say nothing of the smallest investment of imagination from the player). It is, in essence, demanding that things look as much like they already look in the real world as possible. It’s a particularly dull approach to take for a game meant to transport the player to a fantastical world. It says to the player: “Yes, yes, we want fantasy, but…let’s just make it as mundane and familiar as humanly possible.”
If the push for photorealism were merely inadequate to make a game come alive for its players, then Howard’s answer could be dismissed as myopic but harmless. However, the nearly industry-wide insistence on photorealism by big studios explodes game development budgets, which is anything but harmless. Like a giant leech sucking blood out of the host studio, a mandate for photorealism siphons off needed resources, directly impacting a developer’s ability to hire people working on other areas of a game—areas which are at least as important as visuals for engaging players.
The Elder Scrolls series itself is instructive. I remember my first time playing The Elder Scrolls: Morrowind. The wide-open nature of the game was striking to a first-time player. The whole world was open to exploration from the word go, and what a world! The changing weather, the way chance encounters felt like organic moments…I spent weeks exploring Vvardenfell, suspending my disbelief without a problem.
Of course, Morrowind’s graphics are miserable by today’s standards. For the time, however, they were good—better than good, in places. I was absolutely floored the first time I saw the game’s water rendering. Morrowind’s use of real-time reflections and displacement effects was well ahead of its time. (Morrowind was released in 2002; one year later, its contemporaries were still rendering water like this or this.)
Still, there did eventually come a time when I stopped suspending my disbelief, when Morrowind ceased to feel like a real place. That moment had nothing to do with the graphics. It was the characters. They had started repeating themselves. I could travel from one end of Vvardenfell to the other, talk to completely different NPCs, and get the exact same dialog in response. They had revealed themselves as hollow marionettes, something that no amount of graphical fidelity could fix. Thinking back, this brought on one of those moments of cold disgust that only a gamer can feel; I’d overinvested myself in this world inhabited by empty nothings. “I’ll never get that time back.”
Years have passed; graphics have improved. Skyrim is far more realistic-looking than Morrowind ever was, but the old flaws remain. The “arrow to the knee meme” exploded because it represents the same old flaw. It’s meant to sound like a personal anecdote to give the game’s guardsmen character, but every single guard in the game says it. It pulls back the curtain, reveals the fraud, erupts the suspension of disbelief. Better models, better textures, normal mapping, facial animations: what does it matter? They’re still just hollow husks.
Perhaps I put it best back in 2011:
Imagination transcends technology. It does not depend, and has never depended, on a 1-to-1 graphical representation of the thing being imagined. To say otherwise is to conceive of immersion as a mechanism by which the game disengages a player’s imagination and replaces it with direct sensory input.
Imagination replaced by direct sensory input. That’s a sad way to think of the role of video games. It doesn’t represent the promise of RPGs, and it certainly doesn’t represent the kind of industry I want to be a part of.
Filed Under: Biz, Guest Posts - Comments: 10 Comments to Read
Anon said,
Skyrim photorealistic? Are you kidding me?
It’s graphically elaborate for an open world CRPG, sure – but not in the direction of photorealism. In my eyes it’s a step nearer to those Brothers Hildebrandt fantasy artworks than any fantasy CRPG before it (with a 3D engine).
Note, that I’m not defending the game – I agree that the game behind the “illusion” is disappointing in a lot of places (extremely linear dungeons, stability on PS3 platform etc.) but it’s unfair to hit it for something that is common to CPRGs: Generic answer from generic NPCs.
Yes, I mean the knee-meme – there are how many guards in this game? Two hundred or more – and the sick & spoiled brats who play this want each and every single one of them have artificial intelligence to respond with infinite variety? WTF?
Xenovore said,
I agree. For a few years now, we’ve been at the point where the graphics are plenty good enough. (Rage’s visuals proved that, at least when it was working properly.) I really enjoyed the look of the Borderlands games; that’s how it should be done. World of Warcraft: that’s how it should be done. Dishonored went for a more realistic look, but still has a painterly quality to it. Whatever the game, the developers should put their own style into the visuals, making them interesting and unique, not just try to mimic reality.
What’s still lacking are things like:
• Large seamless, populated areas (biggest I’ve seen lately is only a couple square kilometers; give me 100s of square km at least), not broken up into “zones”.
• Smarter, more responsive AI. (But yeah, make NPCs talk a lot less; in real life you don’t have people running up to you and saying “Hi” every 10 seconds or carrying on inane conversations with themselves.)
• More emergent game-play. Most games still want to railroad the player along. Instead, give the player options to play and solve things his own way. And don’t penalize him if he does something unexpected. (As long as it’s not leveraging a bug, of course.)
• More dynamic story and world. Let the player(s) bring about actual change in the world setting and plot direction. And have things happening even when the player isn’t directly participating.
• #1 for me (and something of a corollary to more emergent game-play): Better interaction with the world. Let players pick up whatever they want, push stuff around, break stuff, etc. with good feedback and control so they can see exactly what they’re doing. There have been some attempts at it, but with terrible control. E.g. in real life stuff doesn’t dangle from an invisible thread in front of me when I pick something up, nor does it bounce around wildly when I run into something with it! We’ve got two hands — give players at least a modicum of the fine control they have in real life.
Craig Stern said,
Xenovore gets it; I suppose one out of two ain’t bad.
JTippetts said,
I’ve gotten to the point that I broadly (and possibly unfairly) categorize games purely based on visuals. If a game touts its visuals, if a studio brags about how awesome its graphics are, I’m not interested. In essence, they are telling me in detail exactly where they spent most of their budget, and it’s not on gameplay.
Don’t get me wrong; I do like visuals. But I also like abstraction. I like visuals that complement the gameplay, that work well with it, not visuals that are photo-real purely for the sake of someone being able to scream “look how real these graphics are! Ain’t they FUN?” Real life is messy, un-glamorous, dirty, and not very fun at all. Why any developer would really want to try so hard to mimic reality, I honestly couldn’t say.
Felix said,
Er, no offense but that’s not such a good idea. Imagine having to walk for 15 minutes of real time between two in-game locations 1Km apart. Repeatedly.
And how do you propose to achieve that with a keyboard and mouse? 🙂 With a controller it would be at least technically feasible, but imagine having to carefully, very carefully, move the analog sticks around, then hold the shoulder buttons to grab things. You’d be like an arthritic robot sifting through a box of Legos with plungers for manipulators.
Abstraction is as good for gameplay as it is for graphics, as Wii developers discovered to their dismay. Gestures are just too vague and difficult to read.
Modran said,
And don’t forget the dreaded Uncanny valley. There’s a moment when graphism are too close to reality for comfort but not enough for believability. That’s when you notice all the flaws you would have otherwise overlooked (“his skin is too pasty to seem human”, “he never blinks”) if the art was “inferior”.
Scott McCloud explains that with crude art, our brain “fills in the blanks”, but as art get prettier, there are less blank to fill, and there commes a point when the blanks to fill are too small for the thoughts you’d put in. Like trying to get a square block in a round hole.
And then you can”t not see the holes…
Maklak said,
For me the graphics hit a “good enough” several years ago, and that’s for FPS-like games. I can’t play Gothic 1 anymore due to how it looks, but most things from the last 5 or so years are fine. Heck, I even think that about the only good thing about consolisation poisoning the AAA games is that it forces them to work just fine on older PCs.
As for other gernes, I care about graphics even less. Heck, one of my favourite games of all time is Dwarf Fortress on Phobeus graphic pack. Everything is represented by 16×16 squares.
WoW and Borderlands are both pushing me away with their graphics and for me they are firmly in the “don’t” category. But then they are popular, so maybe I’m wrong.
> > give me 100s of square km at least)
> Er, no offense but that’s not such a good idea. Imagine having to walk for 15 minutes of real time between two in-game locations 1Km apart. Repeatedly.
This much space would also either have to be almost empty, filled with buildings that can’t be entered or procedurally generated and then edited by level designers. I agree that even with unrealistically high running speed in games, it would be a chore to walk between distant locations in such a world.
> > (snip) We’ve got two hands — give players at least a modicum of the fine control they have in real life.
> And how do you propose to achieve that with a keyboard and mouse?
Well, I suppose this would be doable with VR gloves, but I prefer a more abstract system with a mouse and keyboard that we have now.
alanm said,
Awesome graphics are great. Every five years I buy the latest big commercial glitzfest, just to check on the state of the tech. Graphics algorithms are cool stuff, worth appreciating in their own right.
But games. Games are about mental models and systems and gameplay, not graphics. The golden rule for me is: the art shouldn’t get in the way.
It’s time the game industry drags itself into the 21st century of software development and admits the obvious: game art is just another usability requirement. I mean usability in the UX sense of the word. Artwork is a facet of UX because it impacts the experience of the user. If your game art is getting in the way of users of your system achieving their goals then it needs to change, simple.
Phrased like this it’s not a mystery – UX is a well researched and understood field. Games like Skyrim have an explicit use case of “user wants to ogle the lavish graphics”. Dwarf Fortress and Nethack don’t have that use case. Different targeted user populations throw up different use cases. A certain style or quality of artwork may be needed to satisfy some use cases (“user wants ascii-art graphics so that he becomes enmeshed in the rougelike tradition and feels l33t”). Just another day of real life software development.
Tesh said,
I’m an artist in the game industry. I appreciate good visuals, as my huge library of screenshots will attest.
Still, I avoid the Uncanny Valley like a plague, and sweet visuals won’t save poor gameplay. Yes, I’ll get a couple of hours of sightseeing out of Dear Esther, and a bit of rumination on the story, but the experience is fleeting. I’ll happily play Minecraft for many, many hours, very content with the visuals because they work with the *play* instead of competing with it.
On the dev side, it also makes me sad when pushing the graphical envelope is a higher budgetary priority than making a great play experience. Mind you, I love chasing the bleeding edge professionally; I trained in the Pixar movie level of tech, after all… but if it’s not in service of the game, it’s inadvisable.
Xenovore said,
@Those assuming 100 square miles must be empty and we’d have to walk everywhere…
A) I stated “populated”, i.e. there is content there, and yes, of course there would need to be some procedural content generation. (And “…buildings that can’t be entered…”; LOL! How is that a requirement anymore? Maybe 10 years ago…)
B) Why do you assume I was talking about walking?? Why do you assume it would even be walkable?
@Those reading “2 hands” as an indication that we must don VR gloves, or start using two mice, two joysticks, whatever you were thinking… No, that’s not what I was saying at all; good extrapolation though. =P
My point was that when we pick up and manipulate things in real life, we can have two stable points grasping the object, and we can generally handle it with good control. Conversely, object manipulation in games tends to be a pain in the ass, with a single, unstable grasping point, and only the most crude control. E.g. my only options are to drop an item from 4 feet up, or chuck it across the room? I can’t rotate it? No feedback about where I might potentially place it? Why do I always have to pick items up to move them? Why can’t I just nudge things if I want to? (Something that could be easily done with a mouse and/or keyboard.)
I want better. If a game is going to let me manipulate objects in the world (and yes, it should), then it should better mirror real life and not make it into a fiddly chore.
@Tesh: Agreed!