neighborhood boundaries are my obsession here in the district, and nothing gets me more heated up than seeing (just about) everything east of the river shown as "anacostia".

looks like another problem with their rankings is that area inside of rock creek park, the arboretum, and the old soldiers' home are included. these areas should really not be counted, as they aren't neighborhoods. you might want to walk to them in order to visit them (which is tough if you don't live in the neighborhood next to the arboretum, for example), but they shouldn't drag a neighborhood's "walkability" score down, IMO.

by IMGoph on Jul 17, 2008 9:20 am  

This map highlights the potential and obstacles that the MacArthur Blvd. has. I think it's a beautiful neighborhood that is very walkable, but its transit connections are dreadful. It's a real shame, because the neighborhood offers so much: great historic homes, great public schools, a good (not great) concentration of restaurants, and a gorgeous streetscape.

There is a lot of potential to increase density along MacArthur without destroying the character. The obvious lynchpin to that would be improved transit. Of course the tragedy of that is that the neighborhood had great transit with the Cabin John trolley. I really wish the city would consider rebuilding at least a section of the Cabin John trolley along the right of way that still exists.

by Reid on Jul 17, 2008 9:53 am  

But there are official neighborhood boundaries, freely downloadable in either ESRI or Google Earth formats from the DC Department of Planning GIS website. (There are several free programs that can read the ESRI format, by the way.)

Choose view all, then note that there are "assessment neighborhoods," "assessment sub-neighborhoods," "neighborhood clusters," and "neighborhood composition" datasets.

by thm on Jul 17, 2008 9:57 am  

Definitely some problems with the algorithims, but very cool nonetheless. I'm pretty sure that the neighborhood boundaries come from Zillow, partially because they have to get boundaries all from one data source, rather than picking and choosing what data to plug in for each city. So that's why the DC OP boundaries aren't used (which some people would quibble with anyhows).

But David's right, it gives you a great birds-eye view of walkability in DC at large, and by general neighborhood when you drill down, whether the exact number for each exact neighborhood is spot on.

If you're used to checking out their regular tool, it's just beyond cool to me to have the color coded temperature scores for the entire city that you can view at large or small scale.

Don't miss their petition to improve America's walkability if you're interested.

www.walkscore.com/transportation-bill.shtml

by Steve Davis on Jul 17, 2008 10:11 am  

Yes, walk score is real a good idea! But why only walk score?How often do we walk nowadays? More and more people drive cars.I have also tried one more service at http://drivescore.fizber.com/. It is called Drive Score.With the help of it you can see how close establishments are by car. Try! It's really a perfect idea!

by Tina on Jul 17, 2008 10:45 am  

you're kidding me, right tina?

by IMGoph on Jul 17, 2008 10:48 am  

Tina, I think you are looking for www.worserworserwashington.org...

by NikolasM on Jul 17, 2008 11:01 am  

I see Tina's point. The fact is that most people outside of the central core will not live in walkable neighborhoods and converting to one will take several years if not generations. So if you are looking to live outside the core, why not find a place where you can use your car the least. What we really need a Gallons/year map that shows how much fuel one uses to during a given year.

by RJ on Jul 17, 2008 11:16 am  

I've noticed "Tina" posting on Walk Score posts here before, maybe not with the exact same text, but with something very similar. I wouldn't be surprised if "she" finds blog posts all over the web that are tagged with Walk Score and writes the same comment on them.

Am I calling her a sockpuppet? No. Troll? Yes.

by Adam on Jul 17, 2008 12:07 pm  

"Yes, walk score is real a good idea!"

At least we can agree that it is truly real a good idea.

"But why only walk score?How often do we walk nowadays?"

Lessee, I've lived here ten years and I've walked somewhere every single day. So I guess my response is "really really often."

"With the help of it you can see how close establishments are by car."

A device which allows one to see spacial distances between two locations? Why, it must be ENORMOUS! What's that you say? It's all shrunken down and in proportion--"to scale," you say? It's really amazing that no one ever thought of such a thing before.

by Jake H. on Jul 17, 2008 12:36 pm  

Fun toy, but not ready for serious planning and investment decisions yet. Hopefully they'll figure out how to include offices, other workplaces, religious and civic uses, transit stops, and in general balance out the heavy retail slant of the score.

Note Bolling AFB and the South Capitol Street corridor is all red (car dependent).

The Takoma Park neighborhood designation needs some work...

by Laurence Aurbach on Jul 17, 2008 12:47 pm  

I haven't been able to figure out drivescore's algorithm. I ran some test cases:

My home (near East Falls Church Metro):

Walkscore (WS): 49 (this would be much higher if the EFC parking lot were redeveloped to include some retail, as is planned).

Drivescore (DS): 80

Parents' House (Southern Reston):

WS 32, DS 68

Uncle's House (SE Pennsylvania, Exurb of Newark, DE):

WS 2 (only the corner grocery a mile away)

DS 65

1776 K St NW (near Farragut SQ):

WS 98

DS 81

by Michael on Jul 17, 2008 1:33 pm  

"Tina" left the same comment on our post this morning, as Nina I think. blog.smartgrowthamerica.org

I told her why not just calculate her walk score and if it sucks then that's probably a pretty good "drive score." Or it's like scores in golf. The lower the better. Driving more YAY!

(that said, I would definitely love to see walk score integrated with the data from the Center for Neighborhood Technology's Housing + Transportation affordability index released a few months ago. As someone rightly points out, choosing a neighborhood where you can drive less or make shorter trips can be a wise choice and an improvement for lots of people. The Joe Cortright study on the Green Dividend in Portland finds that drivers drive less there, saving everyone money. It's just just about walkers, but about fewer and shorter trips.)

by Steve Davis on Jul 17, 2008 2:09 pm  

Yes, someone (then going by "Nike" and using a similar Gmail address to this morning's "Tina") left almost the exact same comment last time I wrote about WalkScore.

NikolasM: Ha!

by David Alpert on Jul 17, 2008 2:13 pm  

The Walk Score seems to be a great example of GIGO, unless perhaps you are considering sending your child to school at the Opus Dei study center or upscale restaurant, rather than the local public elementary school, which isn’t even on the list, or perhaps you plan on patronizing bookstores that have been out of business for more than a decade. Perhaps, you plan on doing your grocery shopping at a minimart or delicatessen, rather than a supermarket, buying your hardware at the Countertop Company, or using the library at a psychoanalysis clinic or specialized museum, rather than using the public library.

And, of course, Walk Score is based on distances calculated as the crow flies, does not consider whether there are sidewalks, what the topography is like, whether there might be a lake or highway between the starting point and destination, and most importantly, does not consider the availability of public transportation in the algorithm.

So, I cannot agree that “The algorithm still is far from perfect, but it does a pretty good job of quantifying what areas are more or less walkable.” The data is garbage, and, by their own admission, the algorithm fails to use some of the most important variables in determining walkability. They recognize that they have omitted from the analysis public transit, street width and block length, street design, safety from crime and crashes, pedestrian friendly design, topography, freeways and bodies of water, and weather.

by JR on Jul 17, 2008 2:52 pm  

JR, everything that you say is true, and it's all part of the weakness of walkscore.

Nevertheless, if you look at that big map David has in his post, does it not coincide with your intuitive understanding of what parts of the District are walkable and what parts are not?

by Alex B. on Jul 17, 2008 3:01 pm  

Where I live in Silver Spring rolls up a 91, which would make it 9th in the District, which is where I walk to in order to buy liquor. Boo-yeah! It would be interesting to compare walk scores for greater metropolitan areas as well.

And I agree that we should not get penalized for Rock Creek Park et al.

by Lindemann on Jul 17, 2008 7:42 pm  

Alex B--the Walk Score is probably useful at the level of precision illustrated in the map (essentially, green areas = walkable, red areas = not walkable). But drawing distinctions between a neighborhoods or cities with tiny differences in scores is worthless.

by alexandrian on Jul 18, 2008 8:20 am  

Oh, of course, alexandrian. I completely agree.

I'm just saying that despite the poor data and an incomplete methodology that doesn't include things like transit or sidewalk design (data like this would be really hard to get and apply on a broad basis, btw), the final result at the 'big picture' scale is remarkably accurate.

Furthermore, despite the many flaws, it's worth noting the attention this has drawn to walkability and alternative transportation in general. That alone is invaluable.

by Alex B. on Jul 18, 2008 8:53 am  

Would it really be difficult to include permanent transit like metro stops (and one day trams) in the algorithm? It seems that would be easier than retail business that can close up from one week to the next. I any case I have not needed Walk-score as a tool.

Isn't there a function whereby one can type-in neighborhood amenities and Walk-score will calculate a score?

by Bianchi on Jul 18, 2008 9:05 am  

"Would it really be difficult to include permanent transit like metro stops (and one day trams) in the algorithm?"

Not to mention bad stuff (from a walkability standpoint) like highways, overpasses, and on-and-off ramps--which can be counted just by looking at Google maps. If that stuff were factored in, it would make the area around the Kennedy Center go from green to red in a hurry!

by Jake H. on Jul 18, 2008 10:28 am  

What can I do if the walk score is very low? You know, I’ve found rather interesting way called drive score. Using the software developed by Fizber company http://drivescore.fizber.com/ I you have the opportunity to choose what to use – walk, drive or bike score. It’s very useful while selling a house. You can show it from advantageous position. You’ll try to.

by FSBO on Sep 27, 2008 12:06 pm  

Hooray, repeat spam!

by Michael P on Sep 27, 2008 1:51 pm