[conductr]

Thanks for doing the deep dive this is interesting data that I didn't want to research :)

> There is a steep drop off in people cycling/motorcycling to work (for some reason combined in the dataset) at the 30 minute mark

I live in Dallas which is one of your samples. I think on the above point what you would find is above a certain threshold and commuters are using a highway/high speed road and probably see it as a necessary component to keep drive time down. I just mapped my commute and it's actually 12.2 miles of which 9.3 are on a highway. It takes me a consistent 20 minutes to get to/from work; or about 35 if I avoid the highway for some reason. Average speed is probably a small factor and stop lights/merging traffic being the dominant slow downs.

On a bike, intersections still slow you down and you don't see people bike across a highway's path very often at all here because all of those intersections tend to be high speed, several lanes, and quite dangerous too. Since biking isn't a normal activity for most people here, I'd venture to guess most people would impose an additional danger to themselves until they got familiar with things.

The shear number of roads and infrastructure to be upgraded to make it feasible is daunting.