Braess Paradox [video]

I don't see any paradox.

There are so many variables it's difficult to generalize without stipulations.

For example, if routes are known to close, some traffic may cease due to perception of lost opportunity, where lowered utilization inproves transit for the remainder.

Imagine the effects of broadcasting a notice of closure without actually closing paths? If transit times improve, does this count as paradoxical?

Depending on the scheduling of intersections, opening up new paths may degenerate efficiency at crossings leading to aggregate decrease in flow (scheduling turbulance).

There's an assumption of shared endpoints (sources and destinations) maybe partial en route. Consider that completely independent paths can't affect each other by definition. If they did, that would be paradoxical, but better to suspect hidden contention.

Imagine opening up all area within an arbitrary boundary between source and destination, so the road capacity expands arbitrarily, what is stipulated about the paths chosen? At the most open and free is it Brownian? Under such analysis, what is the meaning of travel?

It seems to me that early traffic modeling may not have considered orders of effects from first principles and found some edge cases which are surprising only at first order analysis.

If it truly were paradoxical then new road capacity would have ever been built, as every road would further impede transit!

What about too much choice? Travelers never fully embark because they have to many options and can't decide?

Is there a degree of transit speed which erodes any distinction between source and destination? (Instantaneous transport booths)

How time efficient should transit be?

There are surely many more ways of looking at this

Latest Veritaseum video is about Braess Paradox (https://en.m.wikipedia.org/wiki/Braess%27_paradox).

What intrigues me about this video is the analogs in how organizations work (or don’t work) together. What is your experience with the parts of how your team/group/company/organization works that create serial dependencies, where if cut, would lead to more optimal operations due to the parallelism that would emerge?