3 ms·

This writeup would be greatly improved by a detailed real-world example.

• bradrn 5y ago
  I mentioned in another comment that I once read a book which mentioned this paradox. That book gave the example of a major street in Manhattan which was closed for a while; this surprisingly made traffic generally better. Unfortunately, I can’t remember the street, or when this happened, though it might be possible to look it up. (I’d like to know!)
• davikrr 5y ago
  https://en.wikipedia.org/wiki/Braess%27s_paradox#Traffic https://en.wikipedia.org/wiki/Braess%27s_paradox#Traffic
  • p_l 5y ago
    Some of the examples are rather badly chosen. Specifically the one about Warsaw. Yes, a lot of people might have used public transport while the bridge was unavailable, but the congestion was insane - at one point it took hours to travel what used to be maybe 10 minute trip. You see, the bridge was a major artery linking two sides of the city with no other good links nearby (best option was by rail, but it required considerable detours - effectively I never used it when going to work, only when coming back). Especially since the loss of a bridge meant closure of exclusive bus lanes, as even if they continued the jams involved in reaching the other bridges would prevent any speedup. There were essentially no other ways to travel, as in N+M notation (N = river crossing capacity, M = spare capacity) Warsaw has been long time operating at below required capacity. Closest example of adding more crossings not reducing congestion involves a bridge which couldn't take up increased traffic because the roads to it were congested...
• mrow84 5y ago
  The “Possible instances of the paradox in action” section, “Traffic” subsection, has several examples.