Braess Paradox
http://www.nytimes.com/1990/12/25/health/what-if-they-closed-42d-street-and-nobody-noticed.html
The Braess Paradox says that while adding streets to uncongested traffic networks does decrease average individual travel times, adding a street to a congested traffic network does not always decrease travel times. The reason for this is that the street typically is connecting two congested streets, which means it will cause congestion at the access point and exit point. Using queuing theory, Dr. Cohen, mathematician at Rockefeller University in New York, and Dr. Kelly of University of Cambridge showed that in a simple congested network, adding an extra street could cause an increase of travel times by nearly 50%.
The closing of 42nd street in New York, NY and opening of a new street in Stuttgart, Germany evidence this paradox. On earth day in 1990, New York City closed 42nd street. Many people thought this would create an enormous traffic jam, however, traffic flow actually improved. Similarly, 1969 in Stuttgart, there was such serious congestion that the city opened a new street, however, when they did so congestion only got worse. The city ended up closing the street.
This relates to the nash equilibrium we have been studying in class. Adding a street to one of our nash equilibrium problems would have either no-effect or a positive effect, because we assume people choose the best payoff. In the real world, however, people are not rational. If they are stuck in slow traffic on one street, they might try seeking out another street, which will likely not decrease their travel time and contribute to the congestion, increasing overall average travel time.