Include Hills and Elevation Gain/Loss
Steep hills should hurt the Bike Score of an area. I'd suggest analyzing the elevation gain/loss around the radius of a location. Maybe use 3.25 miles or a 15 minute bike ride.
radius of nearest hill is pointless, I live at the bottom of a hill but the bike lane and bike paths all go around the slope at my location and are level, accessible, and continue so for miles around. My walk score and bike score are completely different. It is like living at the top or the bottom of a cliff or bluff. I live where all the amenities are and I live on the street designated a BIKE path which connects to several bike routes. Yet my bike score is below 50. So hills and lack of data applied to my microcosm is false and lowering my entire block's value because of it.
Hills happen -- and all you can do is go around. Seattle is a great biking city but a negative factor for hills wouldn't reflect that.
Sorry to say but...ALL world class bicycling cities (25% by bike) are flat as pan cakes.
i don't think hills should be factored into the score. i think it'd be a "nice to have" in addition to the score. if you lived somewhere with hills and you wanted to get around on a bike, you'd eventually learn to live with and destroy those hills around you. personally, hills are a + for me.
Yes, elevation change is important info to many cyclists, but a city council can't vote to have fewer hills. I think a city's BikeScore should be something that its elected officials have an ability to change, at least in theory. A neighborhood/city/entire region shouldn't be penalized for something that it has no ability to change. Include the data, but don't use it to compute the score.
David Rowan Armor commented
Whatever you do, don't use Googlemaps' calculations for bike routes. It regards hills/elevations to be of such importance that in a city like Minneapolis, it never offers a useful direct route. Some of us are okay with climbing and coasting.
What is the 3.25 mile/15 minute bike ride? The distance, time, and intensity at which people are willing to travel by bike varies greatly among people.
Hills are very important to me in determining bike routes, however, it can easily become very complex for longer routes, and depends on the rider's preferences. A very fit cyclist may not care about hills, where a more causal cyclist will happily go and 1/2 mile to avoid a significant hill.
Living at the top of a hill sucks for bikeability because you always end with a hill climb. I'd rather start with a hill climb and empty panniers and coast back home with grocery laden panniers.
So something implicit there is that the reachability of a destination is different depending on whether you are considering a one way versus a round trip. One-way trips come into the picture if you ride down the hill, and throw your bike on a Metro bus for the return hill climb.
I have to agree with FM, that a hilly versus flat route, within limits, doesn't have a huge impact on my on average speed. The elevation delta between the start and end points has much more of an effect. The "limit" is once you start needing brakes on a downhill, you waste the potential energy you gained climbing up.
And yes, Google Maps does consider grade, though I don't recall the exact formula. Because of this the computed route from A->B may be very different from the route for B->A.
(My frame of reference for dealing with hills: I live on top of one in the Seattle area.)
I think Google Maps does weigh elevation changes.
I don't think that steep hills hurts the bike score as much as flatlanders think it does. California people think bad weather hurts bike score, but look at Madison.
This is essential to include, but is going to be difficult. Topography is variable, so a radial method won't really work. You'll have to calculate the elevation changes for each individual route. And it can't just be start/end -point elevation change, because going up and then down a huge hill and ending at the same elevation is very different from biking on flat terrain.
maybe a certain amount of elevation change adds "distance" from home to a shop. e.x. 100ft = .2mi or something like that.
totally agree. Amount of elevation change over the route should be a factor. I wish google maps weighed this when doing bike routes!!
Except the point is to get somewhere ~ so supermarket or grocery story and drug store "close", post office "close", bus stop "close", means within a certain distance AND a certain change in elevation. Too far or too steep means scoring access as "possible" but not "easily accessible"