Monday, March 29, 2010

Reality Augmented Games

When a designer creates a video game, a video game being a technical product, a designer, like all designers, engineers, or people who make technical products, are critiquing society. They are looking for a gap or a question that isn’t being asked in reference to some technology or in the design of something, In asking this question, that designer creates a new piece of technology, an answer to that question; a bridge. This idea comes from Bruno Latour in his work Aramis.

The designer of a video game, then, is looking to answer some design question; some technological issue that the previous designer of a similar product came up with. It is this mechanism, this process that keeps things changing. It is this process that keeps design in a constant state of flux.

However, money enters the equation at a wide variety of places. Will this design change impact the sales of the new product? How will a precedent influence consumers? Actor-Network-Theory tells us that the answer to those questions lie within the actors themselves. However, those particular answers are measurable. We can tell how much brand influenced sales with two similar products (in moves, for example: 13th Floor vs Matrix or Wyatt Earp vs Tombstone). These designs almost universally rely on a similar idea; an established norm; an accepted means through which business is done. In movies, it’s the structure of a story, what characters do and do not know vs the audience, etc.

In video games, these processes exist as well. There are a set number of ways we’ve decided video games are. While a new way appears, it is typically couched somewhere in between two previously established types. Any ‘radical’ game will not be so much radical, but a new translation of an old idea, with a twist.

In video games, as Ian Bogost tells us, we rely on procedure. Procedure dictates how machines do things like draw rocks or how fast an object can thrust upward, or how different mushrooms interact with a particular character. There is no part of a video game that escapes these procedures. It is this predictability that allows us to practice, to become good at a game; at other games like us.

These procedures are always based on translations of realistic things. Rocks fall, like gravity. If I jump, I’m also going to come back down. The procedures of nature become the procedures of things. However, in video games, you can mess around with those procedures. Only, you can’t stray too far or it just doesn’t seem believable. While you can put fantastic things in games: Flight, Lazers, Surfing down from an exploding spaceship on a piece of metal inside a mechanized suit being piloted by the president of the united states, and other such things, those things are always couched in reality.

Reality is inescapable. You cannot design something without referencing it. The point of this post then, is why spend time designing things like cities, artificial intelligence, and other such procedures when the data for those things are out there, 24 hours a day, 7 days a week? Why should a game like GTAIV worry about designing patterns of traffic for its citizens when expansive studies of New York City’s traffic patterns are done constantly?

Hundreds of thousands of hours are spent programming things we already have data for. Census data, ethnic data, demographic data about sicknesses, traffic patterns, consumer spending, public opinion, gun ownership, if guns make you happier, crime data, property values, voting records, salary data for public officials, and more. Social science drowns daily in the data it has unearthed. You have data going as far back as the origin of statistics.

I suppose, as Lucas Zen Hannon put it, what about reality augmented games as opposed to augmenting reality?

Shenmue tried to do this with almanac data and the weather. Grand Theft Auto IV tried to capture the feel of New York City by having research teams watching the sky:

Because we were working in high definition and we knew we’d need a shitload of research, we wanted to be somewhere where we had a foothold. We have a full time research team here off doing things every day, from how one building looks over another or photo shoots of hundreds of different kinds of pedestrians or video in the sky over 12 hours to see if how we make sky move is correct. There’s so much weird research: The ethnic makeup of every neighborhood. The traffic patterns. To try and get a feeling. We’re not trying to be 100% accurate, but we’re trying to capture the essence of the place.
From Ben Fritz's interview

However, imagine if Grand Theft Auto IV contained up to the minute traffic data for New York City? Concurrently, imagine if Grand Theft Auto IV also held events in the same event schedule as New York City does and created traffic patterns based on the location of those events? If game makers want their cities to feel alive, what better way to do it than to augment their game with live data from a live city?

This opens up a whole new realm of escapism (for good or for bad), whereupon you play yourself, in your neighborhood, in your house, with the normal amount of money you make. What would happen if you, living where you do, suddenly gained some sort of super power? What if aliens came to Earth? Instead of a player having to relate to a character through narrative, the player plays themselves in a fantastic situation. Further, friends living nearby could come over and all take part in what is transpiring. The living, breathing representation of the sum total of social science research can be shared by everyone, or just 1 person. Modular game experiences based on proximity and actual travel to that particular city. The game generates maps and buildings based on real life housing data, maps.

A game that took these things into account would quite possibly be the greatest application of statistical data of all times. The only weakness is the reality of sampling and the distortion of data associated with the weaknesses of statistical analysis.

Please steal this idea.

6 comments:

L. Z. Hannon said...

Call me what you will, but I believe there are messages within calculating a player's in-game starting money on the average income of their neighborhood so hilariously profound, it is physically painful.

Progressing forth, I do think there is a certain lack of specificness to these things. I don't believe your onscreen actor could ever sufficiently represent you via data -- rather, it is a representation of the data that most likely represents you. Your avatar becomes the embodiment of the closest statistical information that can be obtained; it becomes a caricature, a gross reflection of everyone in your neighborhood. Like closing your eyes and rolling a ping-pong ball around in attempt to capture the essence of objects, nuances are lost.

Otherwise, yes. 100%.

Before Game Design said...

I have to admit there is a certain fascination I have that is based on a perverse wonder of statistics. What would a statistical world look like, if those statistics were applied in a video game to represent everyone. Would it look the same? How would it look different?

For specific people. It would be fun to slowly gather game player statistics and use that to populate the world with NPCs for other gamers. It could slowly populate different regions for players to visit. It's a shame that urban environments would fill up.

Over on facebook where this feed is imported, a friend tells me:

As for your paragraph on 'playing yourself'... I think we will have games more closely mimicking every day life as the technology continues to advance (games like Flight Simulator is a good example of a game getting pretty close to what you envision), but I think you ignore one major point. People play games to ESCAPE reality... One may play an Orc Wizard in a quasi medieval setting felling dragons with powerful spells not because it seems 'real', but because it seems totally and completely UNREAL. You are right that it needs to take make some references to reality, less players wouldn't understand it nor be interested in purchasing it, but the key difference between games tends to be the elements in which it differs from reality, not those to which it is LIKE reality.

The fun of emulating an environment populated by realistic data is how you can turn that data on it's head. Given the infinite possibility of reality with pre-programmed restrictions, i'd venture to say the possibility of escapism is far greater than the normal modes of escapism.

Casey O'Donnell said...

Some games are using data like this. Spiderman 3 used building and street data from Google Earth. I know other developers are attempting to figure out precisely what you're talking about. The biggest barriers however are CPU and data.

I'll start with data, since it is probably one of the biggest. Even if a developer knows where to find the data for a given domain, that data may be in a format the requires significant work to decipher and put to work within an existing engine. In many cases data may exist, but getting permission to use it in a commercial application may be difficult. What if a data source moves?

With regard to CPU, a "pattern" may be less CPU intensive than an actual simulation drawing on live or baked data. As long as your pattern provides verisimilitude, why spend more engineering and effort making a reality augmented game? Will users care if they have to trade some other aspect of the game for those reality augmented aspects?

Even if GTA IV could have real-time traffic data, what if it came at the expense of some other aspect of the game? What if it required dropping in game ad-streaming, it would be unlikely to make it past upper management. So there are a host of issues that complicate the issue.

I think you're right though, it is a great idea. I think developers will likely do some of this, but not to the extent that a game set in NYC is drawing real-time data into the game for all aspects of a virtual space. Instead you'll see selective use of these features or a single feature used as a core game mechanic that a game is then based around.

I'd rather see a game that took this kind of data and used it in ways that were non-obvious. What if weather at your current location affected the spawning monsters or loot (rather than ... the weather ... )? Severe weather causes better loot?

Before Game Design said...

I'd rather see a game that took this kind of data and used it in ways that were non-obvious. What if weather at your current location affected the spawning monsters or loot (rather than ... the weather ... )? Severe weather causes better loot?

I really like this idea. I am married to the idea of using this data in obvious ways so I never thought about using it in subtle ways! Extra points for getting across town at 5pm game time! Also, giving control of the weather in a way that affected local crop growth. You could really do a Sim City in ways that would be insane.

Dee said...

Whoa! Sociologist-in-training plus gamer? I caught you on Twitter a long while back, but I've been knee-deep in a thesis. Ugh.

Anyway, good show on you. You're getting RSSed and once this semester wraps up, I'll be back in touch. Take it easy!

Before Game Design said...

I hear that about the semester! During the summer i'm going to be trying to get some structure in what I want to explore in games down. It should be fun! Talk to you then!