Selfish drivers and disciplined production
Traffic engineers in Korea have challenged convention by reducing the traffic capacity along a section of Seoul’s road network in order to increase the efficiency of the system as a whole. Somehow it worked. Researchers from the Sante Fe Institute recently concluded that a large number of network options countervails a large network capacity.
The “price of anarchy” is a measure of the inefficiency caused by selfish drivers. Analyzing a commute from Harvard Square to Boston Common, the researchers found that the price can be high—selfish drivers typically waste 30 percent more time than they would under “socially optimal” conditions.
The solution hinges on Braess’s paradox, Gastner says. “Because selfish drivers optimize a wrong function, they can be led to a better solution if you remove some of the network links,” he explains. Why? In part because closing roads makes it more difficult for individual drivers to choose the best (and most selfish) route. In the Boston example, Gastner’s team found that six possible road closures, including parts of Charles and Main streets, would reduce the delay under the selfish-driving scenario. (The street closures would not slow drivers if they were behaving unselfishly.)
It reminded me of something Brian Eno said in a recent interview about his role as a producer:
“But what I do can work for any artist. In modern recording one of the biggest problems is that you’re in a world of endless possibilities. So I try to close down possibilities early on. I limit choices. I confine people to a small area of manoeuvre. There’s a reason that guitar players invariably produce more interesting music than synthesizer players: you can go through the options on a guitar in about a minute, after that you have to start making aesthetic and stylistic decisions. This computer can contain a thousand synths, each with a thousand sounds. I try to provide constraints for people.”
Related: See bookmarks tagged counterintuitive by myself and others on delicious.