Strava and planning

[bikinchris]

I attended an area planning commission meeting this week and was surprised to learn that planners use STRAVA to help find preferred bicycle routes.

Just thought you might find that interesting.

[sandyrs]

Makes sense. This seems tailor made to that purpose:

Strava Heatmap (http://labs.strava.com/heatmap/#6/-120.90000/38.36000/blue/bike)

[Exonerv]

On a related note:
http://www.wired.com/2014/06/strava-sells-cycling-data/

[MattTuck]

I find it a little troubling to be honest. Yes, strava has good data for the cyclists that use their service. But that is a self-selecting group. I'm not sure how well it extends to the broader cycling public.

Just use their ride planning tool (which alleges it uses the heat map data) and you'll see that sometimes it routes you on main roads with lots of auto-traffic. In most cases, it would preferable to take a mile detour to a parallel route that has fewer cars. The problem with the heat map is that it counts everyone's ride as a single vote.

Experienced cyclists and those that do more miles/rides at certain times of the day, should be over weighted. A better process for a planning board would be to actually talk to cyclists and figure out what locals want, rather than cough up money for this data.

[bikinchris]

I find it a little troubling to be honest. Yes, strava has good data for the cyclists that use their service. But that is a self-selecting group. I'm not sure how well it extends to the broader cycling public.

Just use their ride planning tool (which alleges it uses the heat map data) and you'll see that sometimes it routes you on main roads with lots of auto-traffic. In most cases, it would preferable to take a mile detour to a parallel route that has fewer cars. The problem with the heat map is that it counts everyone's ride as a single vote.

Experienced cyclists and those that do more miles/rides at certain times of the day, should be over weighted. A better process for a planning board would be to actually talk to cyclists and figure out what locals want, rather than cough up money for this data.

They are aware that STRAVA cyclists are not representative of all cyclists.
As for other cyclists. Therein lies a bigger problem. They don't have enough money to do cyclist counts on roads and can't count every single road in the area. They can do traffic counts with unmanned counters, but can't tell if it is a cyclist.
They need something and STRAVA cyclists are much better than nothing.

And yes, they pay STRAVA for that data.

[Schmed]

.... A better process for a planning board would be to actually talk to cyclists and figure out what locals want, rather than cough up money for this data.

True, but if you talk to bike racers, you get one answer, if you talk to enthusiasts, you get another. Mountain bikers, recumbent riders, commuters, etc. etc. etc. all have different wants and needs.

Statistically, the Strava data seems like a good start. Maybe take that data, then talk to cyclists about the plan?

[bikinchris]

The hardest part would be finding out what the people who require sidepaths to ride want.

[jimoots]

Yes Strava does have inherent bias towards more experienced cyclists but there are plenty of ways to interrogate the Strava data to provide some meaningful outputs for a planner.

Want to see what routes recreational cyclists use? Drill down on trips with lower average speeds.

Want to see what routes commuters user? Drill down on the time of day.

So on and so forth.

The people that buy this data aren't buying a single heat map and using that. They're buying an API that they can generate insights from.

[MattTuck]

The hardest part would be finding out what the people who require sidepaths to ride want.

That is actually an incredibly good point.

There is a sort of latent reservoir of 'almost riders' that strava does not capture at all. These are the people that could be convinced to ride if certain criteria were met.

By its nature, the strava data is using the set of people for which the existing infrastructure is 'good enough' to get them over the threshold of riding. The insights you generate may help improve some things, but I'm not sure if it will help bring the 'almost riders' on to the road.

[unterhausen]

You made me go look

I have a pet peeve as far as the bike path from my house into town, and Strava heat maps illustrates it very nicely. There is a point where they want you to ride on a sidewalk instead of having a path, and people ride through parking lots instead. What I didn't realize is that most people ride through a different parking lot than I do. There are also a lot of people that go around the bike path as much as they can, which is something I often do. But the bike path gets lots of traffic on Strava. I think if they looked at that honestly, they would see there is something wrong with that path. It's actually a lot more dangerous than riding on the road, but it avoids conflict, which is something I don't need