Chasing a Sound in your Head

I just can’t get behind John’s post about attention to detail in Red Dead Redemption, a game I gave up on a couple of weeks ago. Being the first game I’ve spent time on in quite a while, a couple of aspects of the game design really jumped out at me.

From the very beginning I got frustrated with how many false conventions game designers now – alright then, these days – expect players to simply accept and overlook. It’s no wonder that many non-gamers can’t understand the attraction of videogames; many of the most popular (and thus most complex) games make very little logical sense. Instead they obey their own set of rules: the way things work in videogames.

When I started out playing RDR, I tried to interact with the many random passers-by on the street, or search about the ranch settlement for useful items that my character could collect. What did I discover?


The zombie-NPCs traipse around and spout prerecorded lines endlessly, and the entire world is simply made of the usual wallpaper thrown over polygons. If you removed the textures from the 3D objects in the game, it would leave an incredibly sparse and simple geometric world with little or no understanding of its own contents. I don’t mean it’s lacking visual fidelity. I mean it’s lacking semantic meaning. Everything solid is just solid, and it matters not whether it’s a tumbleweed or a rock. Geometric shapes are immovable, and lack any other characteristics. You run full tilt into walls and remain there, nose to the bricks and legs pumping helplessly, maybe with half your torso atomically merged into the wall. Everyone and everything in this world is a wind-up prop, and dumb as a rock.

If you’ve played a game in the last ten years this should be familiar. And that’s my problem. Technical limitations like collision detection algorithms and polygon counts end up defining the feel of a game. In the tug of war between pure authorship and material inertia, the way games are built is winning. Design becomes a slave to form, and the seams really start to show. It’s funny that I only noticed this having not played a game like this in a while. This particular game is from Rockstar, so it behaves exactly like Grand Theft Auto; in fact, it’s quite shocking how slavishly formulaic the whole thing is. Replace GTA’s cars with horses and you’re halfway there. One the one hand, this makes things instantly familiar. But my eyes had been unaccustomed to the bright light of modern sandbox games (I passed on GTA4), and I noticed how many little cracks your mind papers over once you’re already familiar with the language of the medium. Film works in exactly the same way, but we’re all even more attuned to filling in the gaps there (nobody gets confused when it’s daylight in one shot and nighttime in the next; you know that time has elapsed). We all speak, or at least understand, the language of edited moving images fluently.

In an way – and this is the opposite of what everyone else seems to have said about it – RDR is a hugely unambitious game. The developers obviously accepted the established conventions of third-person open world videogames, and tried to do something that could be considered being “big” within those confines. You could say this is fair enough, since this is the same studio that single-handedly invented the genre. But it’s pure quantity over quality. Sure, the maps are huge and the sunsets look amazing, but it all exists within this universe that’s fundamentally based only on shape and texture, and lacking any real-world properties or intelligence. A rock is uniquely a rock, so it should behave like one. Games that try to portray the real world might be better served coming up with a model of a real world, a small but internally consistent model that they can actually spend time developing.

It’s clear how the game was made. The developers started with a framework: there are locations and characters within it and some basic interactions (things spurt blood and fall over when they are shot) but there’s no bespoke detail at all. It’s as if they made the game engine, then turned it over to the level designers who were told to work with the tools they were given. At no point does anything special happen. You can almost sense that all animals belong to a LandAnimal class, with modified properties to make them a bit slower or faster, and a few methods for running and dying. An animal will never look at you curiously in this game. An animal will never sleep. You occasionally stumble across something unique like the grieving husband that John mentions in his post, but then you see the same setup randomly spawn somewhere else an hour later and you realise you’ve been duped.

Probably these developers know something I don’t. Probably they know that their market already understands the conventions of this world, and that eventually, and through repetition, anything that deviates from these conventions seems somehow false. (I suspect that this might be what Tom means when he mentions games literacy.) But for me, this time at least, it never stopped feeling like a videogame.

This might not be a problem. I wonder whether cinema has the exact same issues, but that all of us are so familiar, so fluent in film, that we can’t see it any longer. Look really closely and you will indeed notice things like continuity errors and jump cuts, but it’s hard to detect much else. You’ll notice scene beats and act endings at regular intervals if you remember to try (look for them just before the ad break on TV). But it’s hard to watch a movie without forgetting about all about the other people in the studio, hunched just out of shot, or every scene as a post-it on someone’s wall before being ordered in a shooting script. It’s hard not to just get sucked in and have your disbelief suspended. And most of the time, the same thing happens for experienced gamers, for whom this game is a fully immersive experience. And why shouldn’t videogames enjoy the same level of enchantment as films?

Maybe RDR’s shortcomings would be a lot less noticeable if it didn’t spend it’s entirety aping every classic western trope in film history. Maybe it’s this crossing of streams that makes the whole thing hang together so awkwardly for me; if it didn’t call so much attention to the language of film, its own pidgin dialect wouldn’t be so glaring. In any case, the references to films are lame too; if you saw a real movie with a character like the snake oil salesman in RDR you’d think it was clichéd, derivative rubbish.

I’m on a bit of a rant here. Red Dead Redemption is one of the best-reviewed videogames of all time. But that just serves to highlight the dearth of ambition in so many games; it’s only a great game if you accept the existing conventions as a given. But really it’s just a popular old game with horses and six-shooters. A local maximum. It’s Back to the Future Part III: a fun twist on an old favourite, but hardly the greatest achievement of the medium.

Creating something shouldn’t just be about what comes out when you give it a go. Music that sounds like it came out of pressing record and just strumming away is usually boring. Much more interesting, on the other hand, is when someone has a sound in their head that nobody has ever heard, and they have to search high and low for a way to turn that idea into actual sound waves and record it. Quoth James Murphy on how a song can either develop organically from tinkering, or can be something that you hear in advance and have to seek out:

JM: Both of those things happen and they’re very separate. It’s not a balance between them. Sometimes I’ll just have a loop, then go play drums, then go play something on top of the drums, and things just start having a feel. Then other times I’ll think, “I want to have something that sounds like this.” Then you just have to sit there and try to chase a sound that’s in your head, which is a pain in the ass.

Don’t let the material guide you, designers. Think of something simple but rational and self-contained, internally consistent, and then chisel away at the material until it matches. It’s fine if it’s small. But don’t just settle for whatever comes out first try.

Or do and just call it a new language, I don’t know.

— 25 Jul 2010