Bike Score™ Methodology
Bike Score measures whether a location is good for biking on a scale from 0 - 100 based on four equally weighted components:
- Bike lanes
- Destinations and road connectivity
- Bike commuting mode share
Like Walk Score and Transit Score, our goal with Bike Score is to provide an easy way to evaluate bikeability at a specific location.
If you'd like to use Bike Score data, see our data services.
Bike Lane Score
The Bike Lane Score is based on shapefile data provided to Walk Score by city governments. Bike lane infrastructure currently includes all on and off street bike lanes/paths but does not include infrastructure such as bike parking, bike sharing, etc.
Bike lanes are divided into four categories: on-street bike lanes, off-street trails, cycletracks (separated bike lanes), and residential bikeways (a.k.a. bike friendly streets or greenways).
While examining the data, there were variations in how cities reported their infrastructure, especially with regard to residential bikeways. In cities like Vancouver, residential bikeways are bike priority streets with traffic calming, signage, and on-street markings. In some US cities, a residential bikeway might have very little infrastructure. Because of these, we collapsed the four bike path categories into two categories, on-street and off-street.
For a given location, we sum up the length of all nearby bike lanes. We apply a distance decay function to each segment, where no value is given to segments further than 1,000 meters from the origin. We weight off-street lanes 2X as valuable as on-street lanes. This creates a raw value that we normalize to a score between 0 - 100 based on an average of the highest Bike Lane Scores that we sampled.
To calculate the "hilliness" of an area we look at the steepest grade within a 200 meter radius of the origin. A grade of 10% - 2% is given a score of 0 - 100. Our data source is the National Elevation Data set from the USGS.
Destinations and Connectivity
To measure destinations and road connectivity, we use a modified version of Street Smart Walk Score. Street Smart Walk Score measures the network distances to a diverse set of amenities and calculates connectivity metrics such as average block length and intersection density.
Bicycle Mode Share
Initially, we planned to include only the three previous components in Bike Score (bike lanes, hills, and destinations + connectivity). However, looking only at these components doesn't account for the number of bikers in a city like Portland, OR vs. other cities with similar infrastructure scores.
Biking is social. Many biking experts argue that there is a strong social component to biking. The "safety in numbers" research indicates that more bikers on the road makes drivers more aware of bikers—and more drivers have had the experience of biking. We believe as more people in your social network bike, there's a stronger chance that you will bike.
Bike infrastructure is important, but it's not the whole story. From 2007-2010 a lawsuit prevented the city of San Francisco from adding any new bike infrastructure—and yet cycling grew faster than any other mode share. Also, a slow residential street with no bike infrastructure can be better for biking than a high-volume fast street with a bike lane.
We decided to add a fourth component to Bike Score, bike commuting mode share from the US Census, to try to capture some of this social component. We create a 1 km moving window over the census tract level data and normalize bicycle mode share from 0 - 10% to a score between 0 - 100.
The Bike Score methodology is based on thousands of votes from the Walk Score community and expert advice from Professor Meghan Winters at Simon Fraser University, Professor Michael Brauer and Professor Kay Teschke at the University of British Columbia. A portion of the Bike Score methodology was developed with a grant from the Canadian Institutes of Health Research.