It feels ridiculous to post a link found from Kottke (especially one that came via Alexis Madrigal) since I assume everyone has already seen it, but this article on analyzing cities like one would the molecular structure of materials intrigued me. Will this modeling actually yield value? I'm skeptical of any algorithm that puts Los Angeles and Seattle in the same bucket.

The premise is intriguing, but the question of the value of metaphor is even more important.

Are materials and metropolises really comparable? And if so, is the comparison useful as more than a metaphor? The urban planning community, which has its roots in the design world, has historically been wary of science’s attempts to capture the incredible complexity of the urban environment. (In her classic 1961 book “The Death and Life of American Cities,” Jane Jacobs lambasted modern urban planning itself as a “pseudoscience,” in which “years of learning and a plethora of subtle and complicated dogma have arisen on a foundation of nonsense.”)
But it is warming to these efforts. Today scientists are some of the leading investigators of urban design issues. “There have been ideas about cities since Aristotle and Plato,” said Luis Bettencourt, professor of complex systems at the Santa Fe Institute. “But the ways we can measure cities, test ideas, and compare cities across time and place and size has become so much more possible, that we can now test those ideas.” Bettencourt, who was trained as a theoretical physicist, published a paper in Science last summer proposing a new quantitative framework for understanding cities: They are a unique complex system, he argues, with predictable social, spatial, and infrastructure properties.

It's a worthwhile question to ask because I've long thought Silicon Valley and the technology world to be dangerously addicted to metaphor. If you look hard enough, almost anything can be found to resemble something else, but that does not mean outcomes in one system can be used to predict outcomes in another.

But pattern recognition is a reflexive habit for venture capitalists and technology prognosticators. When the future is unknown, we look to history as a guide because hindsight is rich in specific outcomes. This can be a dangerous trap when the similarity in patterns occur at the surface but result from differing underlying dynamics.

Modeling is one form of metaphor, and it can also be dangerous (recall the first chapter of Kieslowski's masterpiece The Decalogue, based on the first of the Ten Commandments: "I am the Lord thy God; thou shalt have no other gods before me."). However, the increased digitization and measurement of the world has made it possible to model many more natural phenomenon.

Look at the recent success of people with finance backgrounds moving over into the sports franchise ownership and management. Jonah Keri examined one such successful crossover in The Extra 2%: How Wall Street Strategies Took a Major League Baseball Team from Worst to First, but most teams not run entirely by finance alumni still employ quantitative analysts to pore over data and model out player and franchise performance, not to mention attendance and revenue. With advanced camera systems and statistical tracking come more data with which to build models of individual and team performance at greater resolution. Sports previously thought to be too complex to model (mostly team sports like basketball, football, soccer, and hockey, which lack the volume of discrete individual confrontations that baseball offers) are being understood at a deeper level using technology like the SportVU camera systems, and even baseball is being understood at a finer level by its own implementation of 3D camera tracking and systems like PitchFX.

When is there enough data to use a model for prediction? I thought of this when reading about the analysis of cities as molecular structures, and it recalled the legendary city of Magnasanti, the so-called perfect city.

If you haven't heard of Magnasanti, it's likely because it's not a real world city but a virtual city built in the game SimCity a few years ago by a young architecture student named Vincent Ocasla.

This video provides an overview of Magnasanti, as does this interview with Ocasla.

Godfrey Reggio’s Koyaanisqatsi seems to have been a big inspiration.
It very much was--I first watched it in 2006. The film presented the world in a way I never really looked at before and that captivated me. Moments like these compel me to physically express progressions in my thought, I have just happened to do that through the form of creating these cities in SimCity 3000. I could probably have done something similar--depicting the awesome regimentation and brutality of our society--with a series of paintings on a canvas, or through hideous architectural models. But it wouldn’t be the same as doing it in the game, because I wanted to magnify the unbelievably sick ambitions of egotistical political dictators, ruling elites and downright insane architects, urban planners, and social engineers.
I’ve a quote from one of your Facebook status updates here: “The economic slave never realizes he is kept in a cage going round and round basically nowhere with millions of others.” Do you feel that sums up the lives of the citizens of Magnasanti? (And you might want to set your Facebook to private by the way.)
Precisely that. Technically, no one is leaving or coming into the city. Population growth is stagnant. Sims don’t need to travel long distances, because their workplace is just within walking distance. In fact they do not even need to leave their own block. Wherever they go it’s like going to the same place.
Heavy.
There are a lot of other problems in the city hidden under the illusion of order and greatness--suffocating air pollution, high unemployment, no fire stations, schools, or hospitals, a regimented lifestyle--this is the price that these sims pay for living in the city with the highest population. It’s a sick and twisted goal to strive towards. The ironic thing about it is the sims in Magnasanti tolerate it. They don’t rebel, or cause revolutions and social chaos. No one considers challenging the system by physical means since a hyper-efficient police state keeps them in line. They have all been successfully dumbed down, sickened with poor health, enslaved and mind-controlled just enough to keep this system going for thousands of years. 50,000 years to be exact. They are all imprisoned in space and time.

If you look at how Ocasla achieved the perfect city in SimCity (I have not played the game in years but “perfect” in this case is measured by in-game metrics such as citizen happiness, crime rates, the number of abandoned buildings, etc), Ocasla came up with a symmetrical layout based on the Buddhist Wheel of Life and Death. The symmetry means everyone has the same amenities within a short distance of their residence, so there's no need to travel a lot for those because everyone has their version nearby. There are no roads, only subways, and there are lots of police stations, one of the things that gives Magnasanti its totalitarian feel.

If this is the ideal city in SimCity I have severe doubts as to the sophistication of the game's simulation. It looks one roadless suburb with the same strip mall in the center of each, all with the same exact stores: Chipotle, Starbucks, a Best Buy, maybe a Home Depot. The same suburbs that are seeing ghettoization and an outflow of residents into urban centers. Where is the art museum, the local sports stadium? How do you replicate restaurants from star chefs?

Despite all that, I could see China trying to build a Magnasanti prototype in their countryside.