Don't get me wrong, I love WalkScore, but the fact that it doesn't take transit into account is a glaring omission. I know that can be hard to map, but aren't just about all rail transit stops part of the info that's in Google Maps?

My other gripe is that it doesn't distinguish between, say, a 7-11 and a full-service Safeway, or between a KFC and a nice sit-down restaurant. But that would be nearly impossible to do in an automated system.

by jfruh on Jun 13, 2008 1:44 pm • link • report

I agree WalkScore is a great idea, but it really doesn't come close to getting things right for me. The stores listed as closest grocery store, bar, movie theater, school, drug store and hardware store just aren't those things.

Luckily my area IS very walkable so it's right in the end, just not in the details. But I can't wait to obsess over the DC Map.

by Ryan on Jun 13, 2008 2:04 pm • link • report

jfruh, I agree walkscore has it's issues. Besides ignoring transit obvious issues including measuring distance "as the crow flies". It also does not consider topography, street width, block size, etc...

I think they want to advance their product and will enhance the methodology over time. It's just a work in progress right now. They are receptive to feedback. Within 24 hours I received a response to an issue I emailed them about. The tool is having difficulty with the D.C. quadrant system. It believes certain businesses in SE and SW are just a couple blocks away from my NW address.

by FourthandEye on Jun 13, 2008 2:07 pm • link • report

I agree with all these criticisms, and WalkScore recognizes a few themselves. But the biggest flaw, which I haven't seen mentioned, is that offices and work places don't appear in the score. Also, it is almost totally consumption oriented -- no churches or religious institutions, no day care, clinics, police stations, etc.

by Laurence Aurbach on Jun 13, 2008 8:27 pm • link • report

Walk Score isn't perfect, but it does a darn good job of estimating walkability for any address in the country based on limited data. Google Maps' business listings are imperfect.

Just getting the nearest grocery wrong doesn't throw it off too much; it only shows the one nearest one, but if you click on the little triangle you can see that it actually looks at several.

Yes, it would be great for it to quantify transit, proximity of jobs, etc. But having the ability to put a number to something is very powerful. For all we talk about what's important in a city, being able to color in the city based on those factors, even imperfectly, brings the issue home very well.

by David Alpert on Jun 13, 2008 10:24 pm • link • report

David Alpert wrote: “But having the ability to put a number to something is very powerful.”

Since putting a number to something is very powerful, it would be irresponsible to do so unless one was certain that Walkscore algorithm and the Google Maps business listings are providing a reasonable measure of “walkability.”

As noted in earlier comments, there are serious problems with the methodology and the data used for computing this score. In addition to the failure to capture offices, workplaces, and many other non-retail, non-entertainment destinations, the score does not take into account some of the very basic characteristics of a walkable area. Given the same distance to the included amenities, a neighborhood which lacks sidewalks and has long winding blocks, such as the example of an unwalkable neighborhood on their website, would earn the same score as a neighborhood with sidewalks, shorter blocks and a grid layout. Similarly, there is no reduction in the score if a walk to the supermarket would require climbing a steep hill or fording a stream. Topography, street layout and existence of sidewalks are simply not considered in the score.

Aside from the weight given to retail and entertainment destinations in the calculation, many of the businesses listed in each category simply didn’t fit their description. For example, for one address that I entered, the “schools” included an Opus Dei study center, a day care center, an upscale restaurant and a massage school. One bookstore listed had been closed for more than a decade. Supermarkets included mini-marts that were so insignificant that I wasn’t even aware of their existence, and perhaps they served only the residents or employees of a single building.

So, attaching a number to something is powerful, but using the numbers generated by this algorithm based on these business listings and labeling it a measure of “walkability” is irresponsible.

Since a visual representation can be even more powerful, mapping this measure and labeling it a description of the “walkability” of neighborhoods in a city is even more irresponsible. At best, and only if the business listings were more accurate and if non-retail, non-entertainment destinations were given appropriate weight, the algorithm, using relatively long "walking" distances measured as the crow flies, might measure the availability of basic services or a choice of basic services within a short driving distance of a particular address.

by JR on Jun 15, 2008 10:31 pm • link • report

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

fizber.com. It is called DriveScore.

With the help of it you can see how close establishments

are by car. Try! It's really a perfect idea!

by Nike on Jun 16, 2008 11:41 am • link • report

@JR - calling the walkscore irresponsible seems rather over the top to me. It could possibly be prone to irresponsible use but I don't see how the score itself is irresponsible. They company that publishes the score is upfront about it's limitations and they are working to address them.

Let me give a baseball analogy. There is a metric called OPS which is On-base Percentage(OBP) plus Slugging Percentage (SLG). It is used by fans and sportswriters as crude tool used to measure batting production. From a pure Arithmetic/Stats perspective it isn't that valid. OBP and SLG have different denominators and scales and

probably shouldn't be added together. But the convenience of doing so to wrap value in one number to fans and writers outweighs the shortcomings in their minds. That in and of itself does not make OPS irresponsible. It's still accurate enough for conversations and debates. However, if the MLB Players association said that only players with the top OPS at their positions can start the all-star game that would be an irresponsible "use".

by FourthandEye on Jun 17, 2008 2:29 pm • link • report

FourthandEye: Use of the walkscore to measure "walkability" is irresponsible, even though the company that publishes the score is upfront about the fact that it doesn't take into account topography, exisitence of sidewalks, or whether it would be necessary to swim across a lake as you carry your groceries home, and they acknowledge that it is based on very inaccurate Google Maps business listings data. The walkscore simply doesn't measure the characteristics that are important in determining what percentage of a family's trips might be done on foot or public transit, rather than by private behicle. While the site can be fun to use simply to see how absurd the suggested destinations are for various starting points, it is not "over the top" to view use of this index for policy recommendations or promotion of this index as irresponsible.

by JR on Jun 17, 2008 7:00 pm • link • report

Hi, I'm the founder of Metroseeq.com, I just launched it 1 month ago as a variation of yelp and walkscore but focus more on local deals. It offers a different way of searching - predominantly using Google Maps. I'd appreciate it if you can check it out and give me some feedback. Thanks a bunch!

by Kevin on Jun 20, 2008 4:16 am • link • report