The paradoxical nature of traffic jams

Following on from the ULTra transit post, here’s a question about urban transport: what’s the best way to solve sluggish traffic flow around a busy street? Well, you could try shutting the street down entirely

To mathematicians, this may be a real-world example of Braess’s paradox, a statistical theorem that holds that when a network of streets is already jammed with vehicles, adding a new street can make traffic flow even more slowly.

The reason is that in crowded conditions, drivers will pile into a new street, clogging both it and the streets that provide access to it. By the same token, removing a major thoroughfare may actually ease congestion on the streets that normally provide access to it. And because other major streets are already overcrowded, diverting still more traffic to them may not make much difference.

There are links to some research papers and reports on traffic flow studies over at MetaFilter, but you might want to start with the more accessible Wikipedia article on Braess’s paradox. I don’t know about anyone else, but I find it strangely comforting to realise that the world doesn’t always work the way we expect it to… though that could be because I don’t drive.