I generally like to stay on quiet paved roads with a shoulder. Here's a "trick" I use with Google maps to help determine this. I go to "street level" to see what the road the looks like. If no street level view is available, then this almost always means the road is dirt.

Yes. It's harder.

Routes for motor vehicles are easier to map (one just has to drive on them).

The roads that tend to be preferred for routing for driving (highways) are the easiest to map (and cyclists can't usually use them).

The market for motor vehicle routes is also much larger (more lucrative) and there's more competition.

Motor vehicles don't care (too much) about hills. They don't care (too much) if a route is not quite as short as it could be (few drivers would really care if a route is a few miles longer).

Cyclists generally have a wider range of purposes (leisure/exercise/transport) and preferences than drivers (drivers generally want the shortest/fastest route).

There are lots of roads that drivers have no problems using but cyclists would, very much, not want to use.

I wouldn't call that path/road "generally usable" for bicycle riding.

The location of these sorts of paths/roads should be mapped but the properties/nature of them need to be mapped too. That way, you can actually look for them and people who aren't can avoid them.

It's a deficiency if a map doesn't allow you to avoid these types of roads. It's a deficiency (for cycling) if they are not on the map.

I agree. It's a flaw. But, there is a silver lining.

There is a growing population of tourists who are actively seeking rougher paths. The section here was the worst of it. I would take the entire 102-mile route again on my cyclocross bike, lightly loade,d without hesitation. But, I probably moved faster on my mountain bike.

Heh!

On a trip from VT south into western MA I was following the idyllic, low traffic VT 142 along the CT River. Google Bike Maps suggested that I take Pond, Scott, and Old Vernon Roads instead of 142 out of Vernon VT...



Pond Rd soon became a rough and tumble double track similar to Max's picture! By this time I was too far along to abort the side route. I continued on until it re-joined 142.

You rode through a place called Satans Kingdom and expected a pleasant ride???

It's a silver lining to you and (potentially) a trap for others.

(Many flaws are "silver linings" to somebody.)

I realize that (and I think it's great). But it's a defect if paths/roads are misrepresented. I don't see short sections (that require walking) as a problem.

It's a defect to blindly follow a mapping program that is in beta if you have very specific preferences or needs for your route. If I was on skinny road tires and refused to touch dirt, I would plan my route more carefully and potentially select the "cars" option with "avoid highways" checked. That usually works for me.

When I visited upstate NY and western VT, I was impressed by the wide, clean shoulders on so many roads.

Here in SW Ohio, shoulders are rare, except for busy main roads, and those shoulders are often full of debris. But the country roads are often very quiet, with hardly any traffic, so 2-lane with no shoulder is fine.

It seems that Google Street View has very good coverage in many areas, but other nearby areas can have only a small number of roads included, and the selection is almost random there.

Bing maps and gravel
This NY State biking blog said that Bing Maps showed paved roads in white, and unpaved in brown. I just looked at the example area, and now it looks like paved roads are cream colored, and unpaved are thinner and white, a pretty subtle difference. EDIT--see the example below, this doesn't work.

When I saw this last year, I checked a different state for gravel roads, and it didn't seem to work there. Perhaps they get road status from the state highway departments?

The blog post: Meridale 43
Wait, it doesn't work on this map.
The Coe Hill Road is unpaved. So the wide/narrow differences must be more about main or secondary roads.
(I rode part of this route, and the gravel roads here were amazing, graded smooth, with very fine gravel. 10-12% grades at times, though!)
Bing "Streetside-explore at eye level" view covers all these roads on the map, even the gravel ones.

Checking my own area, the Bing Streetside only has coverage near cities so far. SW Indiana has nothing, for example.

It's a defect to follow any map blindly regardless of preferences and needs.

Part of the problem is that people assume their experience with using maps for driving translates into using maps for cycling. That isn't quite true (I listed some of the reasons in another post above).

Also, quite a few people have an idea that maps are perfect.

I'm going to assume that you know what you are doing. I'm talking about people that might need more help.

Of course, but that requires having good data too. You also don't always have the option to plan routes "more carefully".

Locally, I've seen (in OSM) a road that was classified as "cycleable" that wasn't and a road that is an important cycle route classified as "not cycleable".

I don't trust Google Maps for cycling trails outside of the city. If it leads you on roads, it's usually fine, but what it designates as "bike trails" are often trouble.

It once lead me and a friend to an abandoned ATV trail running through a private hunting reserve. We found out we were trespassing and there were people with guns and hunting dogs around us.

Another time it lead a friend onto a trail that ended in a swamp. Google Maps thought that the trail continued.

It's highly unlikely that Google is surveying these sorts of ways. Who knows where or how Google gets the data for those. So, yes, you should be wary about the quality of Google maps in those places.

Openstreetmap (OSM) quality is variable but OSM allows people with interest in less popular features to add them to maps. That is, OSM maps might have better information about cycle paths/trails than Google does.

Agreed!

In Iceland, there's a comprehensive cycling map that dictates the conditions expected for every road. Number of cars, gradient, road condition, etc. All keyed out - for literally every road in the entire country.

In the United States, our development density, population density, and total area are several orders of magnitude higher. The fact that Google Maps accomplishes as much as it does is a small miracle. The fact that the cycling option even exists astounds me from a logistics perspective. I don't think we have the manpower to properly update every road as might be ideal.

But, consider the following:

1. People did fine without any digital mapping for decades. Even a complete cycling luddite can just turn around if a road is unsafe. Nobody is perilously careening down riverbeds thinking "Google said this would turn out fine!"

2. Because of the vast spectrum of preferences among cyclists, a crowdsourced update feature in google bike maps may make things worse, not better.

3. Google is unlikely to send out a guy with a StreetView camera/backpack out on all of your local dirt paths and abandoned roads. Those, you'll just have to explore for yourself. Many local organizations create cycling maps of their backyard; a cursory knowledge of Google Earth can help you get to making the same. I've done some cycling map work in GIS before, too, if you have access to it through a local university.

I have found Google maps to be a useful tool for cycling routes on occasion. It's better than nothing, and often provides excellent supplementary information, even when you have other maps. Here are some examples. I rode the entire Erie Canal trail last summer from Buffalo to Albany. We had the official guidebook with maps, but it was cumbersome to keep handy and read while riding. We also had days when we rode off the official route, such as a side trip to Niagara Falls and a long detour in the Rochester area. Google maps on our iPhones was very useful in finding other routes. You also can use Google maps on your phone to pinpoint your exact location on a route.

Another example is from an organized group ride that I participated in a couple of years ago. The ride was supposed to be a metric century and it was mid-January when most people are not trying to set personal best speeds. It was a local club ride with the typical hotshots whose primary goal seems to be dropping other riders. The ride leaders did not have cue sheets or maps to distribute and quickly dropped 2-3 riders. Myself and 3-4 other guys hung with the leaders until about the 40-mile point but they dropped us on a very long hill. One of the remaining riders told us not to worry, he knew the route. Turns out, he didn't know it very well and got us very lost about the 50 mile point, after missing a turn. I finally pulled out my iPhone and found us the shortest safe route to the starting point. We ended up with about 67 miles for the day, which was fine because it was my birthday ride and I made my miles!

That is why it is beta. I have no idea where they get their data, but I'd be foolish to think that much of it isn't auto-generated based on bots. It is a good resource, but needs to be verified before use.

And yet none of that data is in Google Maps, OpenCycleMaps, or any other online maps programs I've seen. Google doesn't even offer biking directions there. I'm really missing the elevation profiles when planning.

I'd really like to believe that, but:

Apple Maps glitch directs drivers onto Alaska airport runway - Telegraph
Ontario lake GPS fail as woman drives 100ft into freezing Canadian waters | Daily Mail Online
https://www.cnet.com/news/gps-sends-...ut-of-her-way/

Interesting (not completely surprising). Iceland is the size of Ohio.

Not "every" road.

We don't really know how many did fine or have problems. Also, there might be more people doing things now and they might be going to fringier places than most people did decades ago. One issue is that maps encourage people to go places (few people choose to venture where there are no maps). If the maps misrepresent ways, then those people are encouraged to go where maybe they shouldn't.

GPS-based mapping systems provide a sense of certainty that isn't always correct (and isn't correct where it's important to be correct).

I believe cyclists need to think more like wilderness navigators (understand the basics of navigation) and less like drivers.

The problem can be that people convince themselves "it's probably only a short segment" and they might be too far along for turning back to appear as a reasonable option.

If nobody is doing this, then there's no point in mentioning it. Maybe, the issue isn't this at all. (Though, there have been people who have continued when they should have turned back.)

I'm not arguing for crowd sourced information.

Of course not. There are other sources than Google.

Yes. And Openstreetmaps provides a place for individuals to record what they find.

There are non-beta modes that can be problematic.

???

Any resource needs to be verified (if it's possible).

That this happens is interesting. In any case, you will always find people who misuse something in a spectacular way. That doesn't mean these events are particularly meaningful or useful.

It appears these sorts of things happen because people are just listening to turn instructions (taking the instructions as "commands" rather than "advice").

I'd suggest that using a GPS isn't always "navigation" but often just following instructions.

The new US 36 Bike Path from Westminster to Boulder was completed last spring.
I was impressed when just a couple months later on a whim I tried using Google Maps to plot a bicycle route and it already knew about it!
This initially looks like it is taking the highway, but zoom in and you will see it is really taking the bike path which parallels the highway:

https://www.google.com/maps/dir/39.9...a=!4m2!4m1!3e1

Google, Facebook, etc. don't just have humans sitting there punching in data. They have bots constantly mining it, creating content based on it. A lot of that is less than quality, and only gets fixed when a human (or enough humans), generally a user, go in and flag it as wrong for a real employee to look at. I have no reason to believe that doesn't exist with building cycling maps, too.

Near my home, Google Maps shows some "bike trails" that are really foot trails that are not bike legal, but are commonly poached on MTB's. In fact all footpaths on the north side of the lake are not bike-legal, only the ARBT and fire roads. One of the wrongly marked "bike trails" is too steep for anyone but a committed downhill racer. There are more poached trails than the map shows... for instance, just north of this link there's a footpath to the west of the bike trail in the trees, and there are more trails than shown between the bike path and the lake.

Facebook? Keep in mind that we are only talking about map data.

What map data are these "bots" mining?

I believe Google bought the data they used as the base of their maps and have just continued to refine or update it (just like any other mapping company).

Openstreetmap data started with public-domain data (data owned/produced by governments). There is often restrictions on how this data can be used (Google might not want to be constrained by that).

Core map data (location of roads) has to be surveyed. (One can use satellite photos for surveying.)

Addresses for businesses might be (almost certainly is) mined by "bots" since that's easy to do.

book marked that bad boy

1- How has Doug not posted the obligatory picture of stairs that he and his wife had to ascend while on a tour in Randomcountrystan after Google showed the stairs as a road? Its a vital and key feature in every 'thoughts on google maps for touring' thread.

2- I use google maps all the time. Thats because 95% of my riding is spent within 300mi of my house and I have yet to be disappointed with the program.

3- Locally, there are multiple ways trails are displayed. Bold green line, light green line, dashed green line, and brown line. These each mean something different. I am shocked that so many of the issues mentioned so far are represented on google maps with a green line of any sort. Around me, all unpaved trails are marked as brown, as they should be. The trail of large rocks in MA which is looks like a dry creek bed is the extreme. I am completely surprised that such a trail would be represented with a green line on google maps.

4- Trails around me are so accurate that some which switch between paved and unpaved actually show up as changing from green to brown and back again, and its accurate. Any by 'around me', I mean in the surrounding states that I have explored and ridden in too.

Well-stated and I agree with most points, but I sincerely doubt that access to digital maps has resulted in a significant increase in cyclists getting stranded. I sincerely, sincerely doubt that the existence of the maps has had any influence on the number of incidents in any way.

People were stupid before, people are stupid now. And despite that stupidity, people are generally resilient and generally manage to figure things out, even if they end up with a bad story about how they had to walk their bike 5 miles in the dark.

This idea that Google Maps is shuttling people to their death (obviously exaggerating) or somehow encouraging people to go get into serious trouble seems MUCH, MUCH less likely than the alternative, which is that most people aren't having issues.

Put it this way; I see absolutely no indication, on this forum or otherwise, that Google Maps ever caused an actual incident. The problems you're inferring are hypothetical. The only problem that comes to mind in 3+ years here is my own thread on needing a route south of Denver a couple of years ago... and Google Maps helped me figure that out (along with cyccomute).

I don't disagree. I'm just trying to make sure other people are clear about the limitation of digital maps.

Some "stupid" people actually die.

In any case, the idea is to try to avoid "walking their bikes 5 miles in the dark" (not to treat that as acceptable).

"People being shuttled to their deaths" is more of an example (an extreme example) of the sorts of problems that people can have. And it's possible that "serious trouble" is not common.

But even being inconvenienced is something useful to try to avoid.

Other than being inconvenienced, I'm not really "inferring problems". I don't think that being "shuttled to death" is a practical problem but the things that lead to that uncommon result can also lead to other more common minor issues (thus, the "shuttled to death" stories are "cautionary tales").

This thread is full of examples of people having expectations of Google that are not realistic.

I'm mostly trying to point that out.

Yes, most people don't have issues or significant issues. Driving actually obscures many minor issues that happen.

I use a Garmin (with OSM maps) and minor issues are fairly common. I understand how things work well-enough to work around the issues (not only to work around them but to go and fix the OSM data later).

I suspect many other people are mostly confused and think that the Garmins "just don't work very well".