Chart of the Day: Walkscore vs. Streetscape Quality

Via Jeff Speck’s twitter, here’s a chart from a recent report titled “Painting the Picture: the validity of Walk Score in addressing subjective urban design qualities in the built environment“, which analyzes whether or not Walkscore does an adequate job of actually capturing the elements that make for a high quality pedestrian experience.

Here’s one of the charts, comparing “design score” to “walk score” across different points in downtown Detroit :


Contrast this correlation (.62) with a similar analysis from suburban neighborhoods, where the correlation is far lower (.11). According to the authors, “conducting a correlational analysis by neighborhood type revealed significant differences in the strength of correlation between Walk Score and subjective urban design qualities.”

In other words, Walkscore is pretty crude, and at a few points there were big differences in quality of sidewalks with similar scores. This is especially true for auto-oriented areas. Walk score can’t really tell the difference between a comfortable sidewalk and a STROAD-facing one.

Check out the whole 62-page study to really dig into the data!

7 thoughts on “Chart of the Day: Walkscore vs. Streetscape Quality

  1. Jim Ivey

    Walkscore started as a good idea, and became a steaming pile of shit, gamed with an immense amount of bad data by realtors looking to pitch their neighborhoods and properties. Try looking up “grocery stores” in your neighborhood of choice, and see how many gas stations and garbage convenience stores get mischaracterized.

    They could make progres on reversing this by organizing a mass crowd-sourcing campaign to clean up the data, but it’s become a commercial tool that doesn’t seem to care about quality,

    1. Alex CecchiniAlex Cecchini

      To be fair, I value the convenience store 2 blocks from my place even though they may not have fresh vegetables and fruit. When I need a half gallon of milk or some chips for a party or even the gyros they sell in the deli, it’s a life saver.

      I agree the problem to Walkscore’s approach has always been that they never cared about the details that matter for destinations. Breadth of products offered, quality of products, quality of service, hours they’re open, etc. There are plenty of ways to get this data for destinations (Google/Yelp/etc for hours/reviews, some natural delineation for gas stations/c-stores vs superettes vs full on groceries). I can’t believe it would be that hard to weight the destinations based on that data to give a more accurate WS.

  2. Walker AngellWalker Angell

    I wonder if Walkscore simply lucks out in urban areas while its lack of value shines in the suburbs.

    In suburbs its common for a 1/2 mile walk along a 50 mph Ramsey county road with no sidewalks and narrow potholed shoulders to score much higher than a 3/4 mile walk along a 10′ wide side path with 10′ of grass between it and the road.

    Given that Walkscore (and Bikescore) are owned by real estate giant Redfin you’d think that they’d understand that there’s more to an appealing walkable neighborhood than distance.

  3. Matt SteeleMatt Steele

    The two most useful things about the walkscore website:
    1. The walkshed/bikeshed map generator. That thing is awesome, better than the walkscore itself.
    2. Walkscore, when it is approaching zero. That is, Walkscore is a much better tool to show how hopelessly unwalkable a location is, rather than showing how good a location is. The difference between a WS of 40 and 70 may be tough to identify, but we sure know what a WS of 0 to 10 looks like.

    1. Walker AngellWalker Angell

      Ummm. My office (in a residential area of Shoreview) has a walkscore of 0. Yet we ride bikes to breakfast, lunch, and dinner several times per week. On nice evenings we’ll often walk to dinner or for ice cream. Until Rainbow closed we rode bikes for shopping at Rainbow (and I still often ride to Fresh & Natural). I’ve ridden to the local hardware store, bank, and other places within 2 or 3 miles quite frequently. We’ve also ridden to numerous places within about a 5 to 7 mile radius. This is all because we have safe and comfortable (well, comfortable between curb cuts anyway) bikeways throughout the city. Certainly not hopelessly unwalkable.

      Our house has a walkscore of 16. I know that’s not great but better than 0. Yet we never walk anywhere from here and only rarely ride. There’s too much Ramsey County road with no bikeway required. I know people halfway between our house and restaurants who have walkscores of 40 and 50 and yet they don’t walk for the same reason — Ramsey County roads.

      Agree with you about the graphics tools though.

      1. Joe

        Well notice you didn’t say you ever walked anywhere from your office. A place can have a walkscore of 0 and a bikescore of non-zero, which is clearly where your office lies.

Comments are closed.