Monday, November 22, 2010

The Blind Nature of Walkscore

I've written ad nauseum about the shortcomings of WalkScore. It is a step in the right direction, but like all statistics, an abstraction. Or in some cases, lies or damned lies. WalkScore even admits where their ratings often go afoul and are working to correct this in beta versions that measure diversity of neighborhood services rather than the mere presence of some/any.

Furthermore, it is a quantitative analysis to express the qualitative, which can't be quantified. Follow my drift? In other words, they express walkability strictly by proximity (which is necessary), but not by quality of the walk. Are there sidewalks? Are they wide enough? Are the buildings constantly changing and engaging to maintain interest and foreshorten the walk? Are cars a threat? Are others around to ensure some measure of common trust and safety?

In a way, it measures the skin to explain bone structure. Not quite accurate or scientific, but better than nothing. And, if anything, it represents a collective shift in priority and awareness that it has caught on at all.

One specific example to even further the disparity between superficial and subcutaneous, Main Street in downtown Dallas is considered the most walkable in all of DFW. Not a stretch, but that it is a 98 out of 100, might be. Is it really as walkable as Center City Philadelphia? Portland? Old town Alexandria, VA? Surely not, but I can accommodate my daily needs here. It is inflated because of the issues mentioned above and it counts tunnel businesses which are mostly only open 11am to 2pm, serving the office population which swells above 150,000 each day from a night time (permanent resident) population of about 5,500.

So I'd like to add two things. First, an email from a friend, occasional blog contributor, and reader:
Just thought it interesting to see Walkscore got metro-wide heat maps now, but still amused to find my old neighborhood scores higher (86) than my new neighborhood (83). Attached pics.

In fact, they made a correction a few months ago that I submitted that accurately made the park much closer than they indicated, boosting the neighborhood up to an 89. I just checked now and that's gone.
Old Neighborhood (86)


New Neighborhood (83)


Now that they do have heat maps, let's look at some:


Austin:


Dallas:

You'll notice the green dots as downtown (brightest), with uptown and deep ellum latching on, lakewood, fair park, and bishop arts as some of the others. Striking how poorly Ross Ave corridor ranks particularly in relation to all that is around it. Seems like a great opportunity to stitch several parts of town together...wonder where I got that idea...

Houston:


Philadelphia:


DC: