powered by UserVoice

Walk Score Algorithm Improvements Forum

Help prioritize improvements to the Walk Score algorithm.

  1. 846 votes

    Use walking distances instead of as-the-crow-flies distances

    Best would be actual walking directions (recognizing walking paths and other pedestrian amenities.) Next best would be using driving directions. Next would be using "Manhattan" distances, where you assume a certain street grid + multiply the crows-flight distances by sqrt(2). All of the... more

  2. 726 votes

    Transit should be factored into the algorithm

    Access to transit, expressed as some type of Transit Score, should be factored into the Walk Score algorithm. Please join the discussion here.

  3. 537 votes

    Include "Bikability"

    Walking is great for most things, but sometimes what you need isn't around the corner. That's when the bike comes out. I want to know what's in range.

  4. 492 votes

    Improve Data Quality of Amenities

    Use a better data source than Google Local Search (e.g. infoUSA) so that you can filter out convenience stores counting as grocery stores. Could also use education.com or another service to get better lists of schools.

  5. 317 votes

    Allow users to correct bad data

    Create a user-generated content interface for Walk Score users to correct bad data or add missing amenities.

    How could we prevent people from gaming it?

  6. 304 votes

    Analyze road network when measuring walkability

    Instead of straight line distances, use actual road routing.

    Algorithm should also look at road width/speed as a proxy for pedestrian development. Consider looking at intersection density or block length as other proxies for pedestrian friendliness.

  7. 289 votes

    Trails

    Urban trails do not appear in the WalkScore calculations. But these urban, off-street pathways provide highly desirable routes for biking and walking. Cities like San Jose with nearly 53 miles of trails do not see their walkability scores impacted by these imporant resources.

  8. 207 votes

    Allow users to customize their own scores by prioritizing commerce they value

    I may not have any interest in being able to walk to pet stores and others may not have interest in walking to a school. Personal customizability.

  9. 183 votes

    Hills (topography) should hurt score

    Inclines of more than X% make it hard to walk. Steep inclines could penalize the Walk Score.

  10. 163 votes
  11. 144 votes

    Include safety from crime and traffic accidents

    Some measure of pedestrian safety from crime or traffic accidents (or both) could be included.

  12. 116 votes
  13. 100 votes

    Create walkable neighborhood maps (the heatmaps) for top 250 cities

    The San Francisco Peninsula isn't a large city and isn't quite covered by SF or San Jose heat-maps. Without the heat-map it can be very hard to find those walkable neighborhoods. Can you expand the coverage to the next 200 largest cities?

  14. 90 votes
  15. 87 votes

    Account for urban villages and increase the score if one is nearby

    If several destinations are clustered near each other, that should increase the Walk Score. I might not walk a half mile in three different directions for food, books, and banking, but I would be more likely to walk if they were together in an urban or suburban village.

  16. 78 votes

    Allow the walkscore to increase if more than one destination is found in each category.

    Currently, Walk Score gives maximum points if there is just a single example from each category in walking distance. While this is somewhat accurate, Walk Score should allow for bonus points if *more than one* destination in each category is present. After all, one restaurant is nice, but a dozen... more

  17. 58 votes

    Contact the OpenStreetMap Foundation to start a collaboration

    OpenStreetMap is a world project to create and mantain a world Map open and free with the same collaborative principle as Wikipedia.
    In the OSM Map everyone can add or update any information (amenities, street names or speed limits)
    The Walkscore calculation should be easy if OSM DB will be used ... more

  18. 54 votes

    Figure out how to add work-commute to the walkscore.

    The biggest problem I see with walkscore is that it can't "know" users daily work-commute from their houses. Unlike shopping and some public services now on Walkscore, there is no way of building-in a personal commute to work. So, it would need a separate input box, like on Google Maps,... more

  19. 54 votes
  20. 53 votes
powered by UserVoice