Quick deep-linked url to look specifically at San Francisco for those interested (this map really only starts getting interesting when zoomed pretty far in to a specific city)
> some other non-professional runner can run a 5:37 mile over a hilly span that takes you 7:30-8:00 is humbling
Another big Strava user here. I suppose it's somewhat inspiring and certainly fascinating, but it doesn't really strike a chord with me. I've also don't really connect with the "crush your friends!!" narrative they have as the main thrust of their marketing.
I'm sure it's 'pops' more than "make long-term and measurable, and improvements to your fitness and mental outlook through exercise and a personal training log".
People use strava for many more activities than just running and biking.
Some people uses if while snowboarding/sking. I've personally used it while power kite-bugging.
Yes, Yes, and more yes. This is awesome. for someone who likes to roughly map out routes before trips, being able to see "Do locals ride this road or this one" is huge. Might convince me to create a new strava account (lost my old one when I quit facebook a year ago)
In line with http://xkcd.com/1138/, i'd request this map be adjusted by the population density in the mapped areas, so be to be able to get more data than "a lot of people live here" and "a lot of Strava users live here".
If you look at connections between medium to small sized cities though, often in google maps, even with street view, something seems like it'd be a totally fine route. This takes that, and when paired up with the frequency people tend to feel comfortable riding at speed on them to track, can show you out of ten various routes, which is the best to take. Zoom level :)
it's not so useful for seeing what regions have the most active people, but it is still useful if you zoom in on any city to see where the trails are. even in the most populous regions, there's lots of good info to be had on this map.
This criticism only applies when zoomed out. It provides a lot more information than "where the people are" when sufficiently zoomed in on a specific location.
Why does that make sense in this case? People probably don't care how many bike paths there are in an area per capita. They would just want to know how many bike paths there are and where they are.
I was puzzled by that abrupt change at the border as well.
It could be market penetration.
Maybe bike infrastructure like the bike knooppuntnetwerk plays a role as well. http://www.nederlandfietsland.nl/knooppuntroutes/overzicht-k...
Although dedicated bike lanes fall of much more gradually at the border f.i. https://www.google.de/maps/@51.6623341,6.998291,9z/data=!5m1...
As a Dutchman I never realised how unique this dense network of dedicated bike lanes and route signs is. Until I met some Canadian bike head expats who couldn't stop raving about it.
It seems that if you're into touring on bike and getting around by bike the Low Countries are without equal.
As I'm living and cyling here [0] most of the year, my impression is: It's a mix of things.
* In Germany cyclists are not treated well on streets [1]
* The nature is much more beautiful in NL/Belgium/Luxemburg [2]
* It is population density and "hilliness", too. I.e. there are
more people using a bike, when it's flatter, imho.
* The infrastructure and law in Netherlands and Belgium are much
much more cycling friendly.
* The culture in Netherlands and Belgium is much more outdoor
sports friendly, thus the average person and car driver
treats you better. Guess where all the "classics" of road
cycling are?
[1]: Mainly due to the focus of the media (ARD, ZDF) on the doping problem, using road cycling as a scapegoat and the sacred car culture in Germany. Each time, I'm training on German roads, my life gets threatened at least once since 2006. Thus I train in Netherlands and Belgium for 99,9% of my time.
You mean competitive sports cycling. Everyday commute or leisure cycling is awesome in Germany. (Well, I often get aggressively almost run over by ruthless sports cyclists who treat every way as their highspeed training center every now and then...)
Good point. I know many fellows, who forget that they are moving way faster than expected and it's the competitive cyclist to blame. Oddly enough, I know a few groups who basically race in the traffic; However, the ones I know, reside in Netherlands. Most of the time, when I see conflicts between cyclists and cyclists or pedestrians it is due to bad design (of the pathways of each of one).
In Germany often simply a few red stones in the sideway are defined as cycling path. Recently, cyclist traffic gets detached from the sideway and placed on the road. What is good and basically copied over from Netherlands and Belgium. A downer at this point is, that the state doesn't change the width of already existing roads, though it would be possible sometimes. Thus cars constantly are driving on this "cycling lane" or simply forget it is there.
Basically, I'm speaking of (road)cycling and cyclo-cross. Not necessarily competitive. Considering commuting it gets better, as nature and sustainability is more in political focus these days. However, it is _really_ far away from the infrastructure you see in Netherlands and Belgium.
But it is also depending upon what you compare here. So, considering a global view, I guess that Germany is way over the top. Considering good design, Netherlands and Belgium have my wholehearted appreciation. Netherlands even much more than Belgium.
On a side-note: I do not believe Strava neither has much commuter nor too much "leisure" data, imho, most of them are somewhat ambitious.
There is some information to be gained at a high level. For instance, in China you see that only Beijing, Shanghai, and Hong Kong have high concentration, so this indicates the most international cities in China.
Looking at Iowa, it is pretty cool to see how many bikers do ragbrai - the 2013 route is what connects Omaha and Des Moines, 2012 is to the north of that.
Each pixel has a count. Each tile gets a 90 percentile value. Every corner gets the average of its 4 tiles. Every pixel in the currently requested tile gets a value [0, 1] based on the bi-linearly interpolated "max". It's then colored based on that final value.
It looks like it's losing a lot of dynamic range when you zoom in to a specific road and it's all completely pink (or whatever the hottest color is depending on the chosen theme).
This map is cool, but it doesnt help me use it for what I imagine is a pretty common use-case, finding a popular route from Place A to B. Would it be possible to drop placemarkers between two places and do routing based on Strava popularity? Right now Google bike directions are pretty bad where I live, and it would be nice to know what route the "insiders" are taking to do a popular route (MIT to Concord is one Im thinking of, but I can imagine other alternatives)
Would Strava be OK with people using this data to trace paths for Open Street Map?
I noticed there are quite a few paths through Golden Gate Park in SF (as just one of many examples) that are pretty clear running/biking paths that are not in any of the typical basemap providers' maps and not currently in OSM. If you reached out to the OSM community and gave the go ahead, this would be a great way to improve the trail data in OSM.
I'm amazed that there hasn't been more conversation about this - kudos to both you and Strava for this. These tools and this data set seem to have huge potential for crowd-sourced mapping.
Better still if it could be released an an official layer for Open Street Map. Any chance of that happening given the licensing conditions of the various data sets?
Unlike a traced layer, the heatmap is self-updating and objective, so I think has a value of its own.
Per my other comment -- can you do a version which show avg speed of the route goers? It would be awesome to see the fastest path via bike or run between points... if you noted each stop the people made, it could be the diff between a smooth, fun ride vs a stop-heavy route.
The overlay tiles zoom weird in leaflet. They get treated like a base layer so when you zoom, they get scaled 2x and the new tiles load on top. That doesn't look right when the tiles are transparent. There must be an option to fix this but couldn't find it.
I couldn't get the styles I wanted with mapbox so I just defaulted to google.
Thank you for this map! Two questions:
- what implementation of quadtree are you using?
- will you ever consider opening the data for download as shapefiles/any other vector format?
Thanks again, great job!
Around MIT, the route around the river is pretty popular as the map shows. Except at one point you have to choose between a fork in the road -- do you take a right and run right next to the river and enjoy the water, or do you keep running straight and enjoy the rush of cars on Storrow drive as you run. Im more of a car person, but my girlfriend likes to run by the river. Strava shows me that the Storrow alternative is indeed less popular :(
Can't remember the terminology they use but Strava allows you to define privacy areas (within a configurable radius of a point) where no points are shown on your individual routes. Exactly to stop someone working out where you live/work from browsing your history.
Either they haven't honoured this when calculating the global heatmap or, more likely (and harder to fix), such locations are given away by friends uploads (who don't have the same privacy areas defined).
On this map I took out any point within 1km of the start or end of every ride. This is stronger than what users have set and eliminates 1000s of "hot spot".
I've looked into a couple of these cases and it's friends "coming by for a ride" or them stopping at their house mid ride.
I'm working on further eliminating these points by just not included "stopped" points. I've also tried doing some image processing to remove these hot spots but have had mixed results.
Found a lot of residential building footprints with activity inside just browsing around at z17. Maybe you could use slide to check local activity hotspot's and cross reference them against OSM building footprints or remove those gravity spots with no uniform direction or evenly weighted exit path slope altogether.
Another approach might be only displaying common paths of 2 or more unique client id's.
I commute with Strava every day without privacy, but my address and even my street are completely empty. Probably because of the volume of users in London is so high that no one person can create a visible trail.
Awesome idea and data set, however I'm very skeptical about its quality. Random analysis: the city where I live in Brazil (Curitiba/PR) shows major avenues and bus lanes as heavily used by bikes and for running, which is partly true but it's far from as heavy as the map makes it seem; my hometown 500km from here (Praia Grande/SP) has 7x more exclusive bike lanes and running trails and yet its usage is just average while everybody I know there rides a bike (not to say the most used bike/run path the maps shows there is actually a interstate freeway).
My family almost got killed once because of a bad judgement based on computer maps.
My wife and I were on a cross-country bike trip, with our dogs in tow. On this particular day, we were shooting for the Chicago area and had a long way to go from Ransom, Illinois where we woke up. There was nothing but corn, windmills, and an Air Force special training center. But we had a friend in Oak Park, and we were determined to get there and take a rest day.
After following a number of scattered trails and highways, we wound up lost in a kind of industrial alcove, trying to cross one of the canals. After a little backtracking (which is the worst), we stopped to ask for help from a bald guy in a pink tee shirt whom I'd noticed cycling earlier. He was a German guy named Rolf and he apparently biked about 60 miles a day just for fun. He told us we still had pretty far to go, and he rode with us about ten miles or so to his car, where he had a map book. He gave us detailed directions saying exactly what would be the best and safest route. We thanked him and went our separate ways.
We followed his directions until after dark (at which point we would normally have stopped). We'd already traveled close to ninety miles that day, and were getting anxious to get off the road. At some point, we called our friend to see if we were close. We told her Rolf's directions and asked if that would get us there. She said she would look online.
She called us back with a different set of directions, and, figuring that she should know her own neighborhood, proceeded to follow them. This took us onto a road that, we later learned, wound through what is essentially a wildlife refuge, and which quickly became extremely dangerous. There were no street lights, and cars were coming around turns at high speed. We decided to stop and barely got off the road in time to avoid being hit. We then had to get our bikes, dogs, and trailer across the street so we could walk the path. It was one of the scariest experiences of my life.
I later found out that our friend had never been on the road she recommended, and was going only by what it "looked like" on Google maps. (This was 2008, by the way.)
So yeah, quality. And there's no substitute for human expertise.
EDIT: and yes, it was our bad judgement as well, for taking bad advice... and maybe for not having our own maps in the first place. Although I've done thousands of miles of cross-country (in the US), and I think you really do have to wing it.
Haven't done much bicycle touring, but I've had motorcycle wheels in all of the lower 48 and parts of Canada. Something I learned years ago is to never ask locals for directions. First, a lot of people know how to get to work and to the grocery store, and that's it. Oh, they'll pretend to offer directions because no one wants to look like an idiot who doesn't know their own town. Second, even if the local knows the way their directions will be filled with all kinds of assumptions that are valid for someone who lives in the area but not for someone who is seeing the area for the first time.
Now, convincing my wife of this truism has been something I've worked on for twenty years of marriage ("why don't you stop and ask someone?" "Because it likely won't do us any good."). Thankfully GPS navigation units have made the debate somewhat moot. :-)
One big difference with this data set is that it's pulled from tracks traced by Strava users. So, modulo malicious uploaded tracks, these are all tracks that have been ridden, by definition. And if a given route is "hotter", that means more people have ridden it.
I'd fight with all my soul to prove that's utterly wrong, at least regarding the issues I've mentioned above. If the data is not 100% open it's impossible for it to be so reliable. And even if it is, that doesn't mean it's good data at all.
I think you are being a bit cranky about it, the text on the control panel says exactly what it is. The mechanics of the colorization aren't very apparent though.
It would be interesting to see different splits of the data, I imagine Brazil routes normalized nationally would do better with your example.
my conclusion from this is that a lot of people in Brookline and Allston need to buy "Great Runs in Brookline"[1]. No affiliation on my part, just the knowledge that book put me onto a lot of great running routes that aren't covered very well at all by these Strava wearers. There is more to the area than Beacon St, Comm Ave, and the Chestnut Hill Reservoir.
Wow, this is incredibly useful for planning touring routes.
I often ride 100+ mile trips to areas I don't know very well and the hardest part is knowing what roads are more-or-less bikeable to plan a route.
Google Maps bike directions are basically worthless on long trips since they divert you way too often into neighborhoods and side streets. Strava is a perfect data set for this use case.
This is damn useful as-is, but if someone could turn this data into long-range bike directions I would absolutely pay for it.
The Strava route builder incorporates the popularity of other users routes when you plan your own. So, in an extension to the normal find-shortest-path algorithm between two points, penalties are given to paths not used by other users or bonus given to those that are popular. The result is that you end up with the better roads to ride on. Especially on the bike this is really useful as I can plan new routes in unknown areas and they are pretty good.
For running I found it remarkable how these heatmaps show the inclination to run along water bodies or in forests. It can almost perfect predict where I would or wouldn't run.
http://strava.com/routes let's you create your own route based on the popularity of segments. It's not as fast as creating a google maps route but combining the heatmap with the routes tool is a pretty surefire way of creating a solid route.
This is a trail-poacher's dream. It'd be wonderful to have this cross-indexed with any extant publicly available data sources (state/national parks, property lines etc.) to try to determine the legality of a given route. Illegal trails are likely illuminated by this heatmap, but their illegality is not.
I follow a guy on Strava who is in the process of riding every street in St. Louis City. He's been all over the Northside. But, as this map shows, it's not a popular place to go cycling.
There's a difference between "cycling" and "cycling while using Strava". I suspect that few people who cycle because they can't afford to fix their car, or they lost their license, or their wife commutes in the single family car, etc. belong to the latter group.
It's interesting to see that. My first hypothesis was probably wrong, but in the interests of open discussion, I thought, "Wait, do the poor people up north not bike or run?" I realized immediately that this was unlikely, just they don't own smartphones to track themselves, or some variant of skewed sampling. If they DON'T bike, that's sad, because biking is a money-printing fountain of youth (http://www.mrmoneymustache.com/2013/06/13/bicycling-the-safe...).
SO, as a 100% bike commuter for the last ~4 years... I find this map to be interesting and definitely beautiful -- but utterly useless.
What I would love to see is: Given two points on the map - what is the most common path between them based on multi-path tracks of others...
What were their avg speeds along those routes.
How frequently did they stop.
Also, looking at SF, clearly there is no bike path across the Bay Bridge (as an aside rant, how did we spend $7.5 Billion dollars on half a bridge, adding no additional capacity and no full span bike/pedestrian throughway)...
It would be great for this map to ID bus/bart/public transit options that connect across such things like the bay bridge.
Also, the fact that the dark red is the more trafficked route(?) that was completely non-obvious to me....
"What I would love to see is: Given two points on the map - what is the most common path between them based on multi-path tracks of others..."
Check out strava.com/routes, as others have said in the thread. It doesn't have your feature requests, but it does point to point routing based on popularity.
The Bay Bridge will have a bike bridge to Treasure Island. The old span needs to be removed before it can be completed. Currently one can ride part of the path as an out and back route. http://baybridgeinfo.org/path
I found something very interesting in Minneapolis (where I live); it's a path that you wouldn't be able to find using google maps or probably even any bike-specific route finder.
I'll have to check it out in person but I'm guessing it's a mountain bike trail or something, which is what this heatmap would be great for helping people find.
A similar effect is evident in my area, in which there is an unofficial mountain bike track. The track is well known in the mountain bike community, and the understanding is that it's location can be passed on by word of mouth, but it should generally not be publicly documented, for risk it will get closed down by the landowner. The landowner is a government owned utility and the land is unused, so bikers and local residents don't feel too may qualms about using the land. The track isn't on any of the online mapping services, but stands out like a sore thumb on the heat map. On the plus side, I can now find any tracks that I wasn't aware of!
Agree, and these almost look like GPS tracks. This would underlying mapssets basically irrelevant (except for exposition). I bet quite a few folks are able to pick up visuals of familiar off-piste routes with a couple of glances at familiar but not well-publicised areas.
Those are the official mountain bike trails at Theodore Wirth park. Google Maps is bad for finding mountain bike trails--it doesn't show the other Twin Cities trails at Lebanon or Murphy-Hanrehan either. They're the sort of trails you could find if you wanted to, though, by some searching or asking at bike shops.
I agree in general, though. Even if you have your own routes or know friends who have theirs, this helps break out of that rut and find some new rides. This would have been great for finding rides just across the river in Wisconsin (though as-is it wouldn't help find hills. I've always wanted a map similar to this that showed all roads in terms of both ridability and steepness).
Yep, came here to say pretty much the same thing. There are also some mountain bike trails of so-so quality down by the "blue bridge" in North Minneapolis over the Mississippi. I grew up around there so haven't checked it out in 15+ years, but it was both fun and kinda sketchy at the same time back then :)
Yup, I can also confirm those are mountain biking trails.
I used to bike past here every weekend 1.5 years ago. I clicked on the link because it said Minneapolis but I didn't expect it'd actually be a place that I've been before.
Interestingly, at least for the cities in the southeast where I've lived, these maps are amazingly good proxies for gentrification. For example, have a look at Houston, TX or Memphis, TN.
In the Midwest, the trend is much less clear. For example, Madison, WI is covered with bikepaths, so there's no clear relationship. (Same for Anchorage, AK. You can see the main trail system, and it pretty much circles the city.)
In a lot of other places, the pattern is controlled by topography/parks (e.g. Birmingham, AL)
I'm limiting things to cities where I've lived, but does anyone else notice similar patterns?
There seems to be a chinese decree that all maps must be skewed so that in the case of being attacked, an attacker can presumably not find its way around the chinese capital.
Bike maps are likely created directly from GPS paths, which are unskewed. When you look at Bejing, you can easily see that difference.
I've built map applications while based in Beijing. Google is definitely skewed here, by decree of the Chinese government. The interesting thing is that the skew is different for various regions in China, but it hasn't changed for years. If you correct for the offset in Beijing, everything will be fine as it is the same across the whole city. You can create a map of offsets for the major cities and be done with it.
They claim that the offset is for security purposes but this simple fix shows that the excuse is bullshit. In reality the Chinese put all kinds of barriers to entry for foreign companies to make it difficult to penetrate the market. Another example is that you have to dual home your servers to both china unicom and china Telecom, as the traffic between the two networks is degraded on purpose. You thus have to purchase a special DNS service that differentiates between requests on the two networks.
In the same vein, the purpose of the great firewall is more economic than political, being a highly successful form of protectionism.
Why don't non-china mapping companies have a table of these offsets and simply apply them everywhere (well, maybe except chinese IP adresses) ?
It seems a tiny amount of work compared to, say, gathering the actual street data itself; and it would be an useful thing to make this data available to the public like openstreetmap data.
Based on what the map says about my local area, the results seem quite er... surprising... in that they don't really match what one sees on the ground (especially for cycling). And when I say "don't really match" I mean "are completely out of wack"...
I suppose it's because the results only reflect users of this app, and there's a correlation between app users and certain types of usage.
145 comments
[ 4.7 ms ] story [ 216 ms ] threadhttp://labs.strava.com/heatmap/#13/-122.45242/37.76345/blue/...
Another big Strava user here. I suppose it's somewhat inspiring and certainly fascinating, but it doesn't really strike a chord with me. I've also don't really connect with the "crush your friends!!" narrative they have as the main thrust of their marketing.
I'm sure it's 'pops' more than "make long-term and measurable, and improvements to your fitness and mental outlook through exercise and a personal training log".
I find it perfect for that purpose.
http://labs.strava.com/heatmap/#17/-208.82432/-33.87246/gray...
http://labs.strava.com/heatmap/#8/6.26125/52.11946/gray/both
Especially there is no gap between the Ruhr area and the Netherlands
http://imgur.com/SLF6dIM Source: http://www.ivie.es/es/actividades/noticias/2012/ws_res01.php
[1]: Mainly due to the focus of the media (ARD, ZDF) on the doping problem, using road cycling as a scapegoat and the sacred car culture in Germany. Each time, I'm training on German roads, my life gets threatened at least once since 2006. Thus I train in Netherlands and Belgium for 99,9% of my time.
[2]: Especially in the link provided by mxfh in https://news.ycombinator.com/item?id=7688974
Basically, I'm speaking of (road)cycling and cyclo-cross. Not necessarily competitive. Considering commuting it gets better, as nature and sustainability is more in political focus these days. However, it is _really_ far away from the infrastructure you see in Netherlands and Belgium.
But it is also depending upon what you compare here. So, considering a global view, I guess that Germany is way over the top. Considering good design, Netherlands and Belgium have my wholehearted appreciation. Netherlands even much more than Belgium.
On a side-note: I do not believe Strava neither has much commuter nor too much "leisure" data, imho, most of them are somewhat ambitious.
The data is aggregated into a quad tree based on number of GPS points in each pixel. Tiles are then served on the fly using Go and C using CGO.
Cloudfront tackles most of the load, but the load balanced i2.xlarge instances can do about 300 tiles a second.
Maybe unnecessarily complicated, but it does provide really good local normalization. You also get cool looking shading like in this location http://labs.strava.com/heatmap/#16/-121.76487/38.49134/gray/...
(edit: Your color scales are not really intuitive, I have no idea what red or blue means or red vs green.)
Is this now the complete dataset from strava Run/Bike activities?
It's not updated on every upload, but pretty much the complete dataset.
http://www.strava.com/routes
- Where's the source data from?
- You should think about producing an overlay with bike and run in different colours to see where these differ.
An overlay would be nice, but there is a version for both running and riding.
Also how accurate is the bike vs run stuff? I'm seeing a a surprising amount of people biking activity in wilderness areas.
I noticed there are quite a few paths through Golden Gate Park in SF (as just one of many examples) that are pretty clear running/biking paths that are not in any of the typical basemap providers' maps and not currently in OSM. If you reached out to the OSM community and gave the go ahead, this would be a great way to improve the trail data in OSM.
I've also built another tool, Slide, to help use this data for faster editing, see http://labs.strava.com/slide/
Unlike a traced layer, the heatmap is self-updating and objective, so I think has a value of its own.
I couldn't get the styles I wanted with mapbox so I just defaulted to google.
We're not considering opening the data. But we are working with cities to help improve their bike infrastructure
http://bikeportland.org/2014/05/01/odot-embarks-on-big-data-...
http://labs.strava.com/heatmap/#17/-71.07863/42.35498/gray/b...
Turn on & off recording right at your house & you're advertising:
http://labs.strava.com/heatmap/#17/-121.98018/37.23843/gray/...
Either they haven't honoured this when calculating the global heatmap or, more likely (and harder to fix), such locations are given away by friends uploads (who don't have the same privacy areas defined).
Oops.
I've looked into a couple of these cases and it's friends "coming by for a ride" or them stopping at their house mid ride.
I'm working on further eliminating these points by just not included "stopped" points. I've also tried doing some image processing to remove these hot spots but have had mixed results.
Another approach might be only displaying common paths of 2 or more unique client id's.
http://labs.strava.com/heatmap/#17/-208.78860/-33.89894/gray...
Any idea what these are? I've only seen them in the bike layer
Edit: I found more in a university http://labs.strava.com/heatmap/#17/-208.81832/-33.88863/gray...
My wife and I were on a cross-country bike trip, with our dogs in tow. On this particular day, we were shooting for the Chicago area and had a long way to go from Ransom, Illinois where we woke up. There was nothing but corn, windmills, and an Air Force special training center. But we had a friend in Oak Park, and we were determined to get there and take a rest day.
After following a number of scattered trails and highways, we wound up lost in a kind of industrial alcove, trying to cross one of the canals. After a little backtracking (which is the worst), we stopped to ask for help from a bald guy in a pink tee shirt whom I'd noticed cycling earlier. He was a German guy named Rolf and he apparently biked about 60 miles a day just for fun. He told us we still had pretty far to go, and he rode with us about ten miles or so to his car, where he had a map book. He gave us detailed directions saying exactly what would be the best and safest route. We thanked him and went our separate ways.
We followed his directions until after dark (at which point we would normally have stopped). We'd already traveled close to ninety miles that day, and were getting anxious to get off the road. At some point, we called our friend to see if we were close. We told her Rolf's directions and asked if that would get us there. She said she would look online.
She called us back with a different set of directions, and, figuring that she should know her own neighborhood, proceeded to follow them. This took us onto a road that, we later learned, wound through what is essentially a wildlife refuge, and which quickly became extremely dangerous. There were no street lights, and cars were coming around turns at high speed. We decided to stop and barely got off the road in time to avoid being hit. We then had to get our bikes, dogs, and trailer across the street so we could walk the path. It was one of the scariest experiences of my life.
I later found out that our friend had never been on the road she recommended, and was going only by what it "looked like" on Google maps. (This was 2008, by the way.)
So yeah, quality. And there's no substitute for human expertise.
EDIT: and yes, it was our bad judgement as well, for taking bad advice... and maybe for not having our own maps in the first place. Although I've done thousands of miles of cross-country (in the US), and I think you really do have to wing it.
Now, convincing my wife of this truism has been something I've worked on for twenty years of marriage ("why don't you stop and ask someone?" "Because it likely won't do us any good."). Thankfully GPS navigation units have made the debate somewhat moot. :-)
https://news.ycombinator.com/item?id=7688369
I think you are being a bit cranky about it, the text on the control panel says exactly what it is. The mechanics of the colorization aren't very apparent though.
It would be interesting to see different splits of the data, I imagine Brazil routes normalized nationally would do better with your example.
[1] http://www.amazon.com/Great-Runs-Brookline-Vicinity-Lowenste...
http://labs.strava.com/heatmap/#15/-119.86045/34.43307/yello...
Here's one I've raced in a few times. Looks like it's not quite as popular, though! http://labs.strava.com/heatmap/#16/-89.48204/43.05660/yellow...
I often ride 100+ mile trips to areas I don't know very well and the hardest part is knowing what roads are more-or-less bikeable to plan a route.
Google Maps bike directions are basically worthless on long trips since they divert you way too often into neighborhoods and side streets. Strava is a perfect data set for this use case.
This is damn useful as-is, but if someone could turn this data into long-range bike directions I would absolutely pay for it.
For running I found it remarkable how these heatmaps show the inclination to run along water bodies or in forests. It can almost perfect predict where I would or wouldn't run.
http://labs.strava.com/heatmap/#17/-118.29075/34.02060/gray/...
The noise may well be indirectly tracking building height. ;)
http://labs.strava.com/heatmap/#13/-90.28904/38.64610/gray/b...
What is the source data? I've ridden through North St. Louis using Everytrail and my rides are public.
Edit: It is from Strava itself. I didn't know they tracked routes. lol
There's a difference between "cycling" and "cycling while using Strava". I suspect that few people who cycle because they can't afford to fix their car, or they lost their license, or their wife commutes in the single family car, etc. belong to the latter group.
What I would love to see is: Given two points on the map - what is the most common path between them based on multi-path tracks of others...
What were their avg speeds along those routes.
How frequently did they stop.
Also, looking at SF, clearly there is no bike path across the Bay Bridge (as an aside rant, how did we spend $7.5 Billion dollars on half a bridge, adding no additional capacity and no full span bike/pedestrian throughway)...
It would be great for this map to ID bus/bart/public transit options that connect across such things like the bay bridge.
Also, the fact that the dark red is the more trafficked route(?) that was completely non-obvious to me....
Check out strava.com/routes, as others have said in the thread. It doesn't have your feature requests, but it does point to point routing based on popularity.
Here's the area in google maps:
https://www.google.com/maps/@44.9837242,-93.3338652,1914a,35...
and here's the same location in the Strava heatmap:
http://labs.strava.com/heatmap/#16/-93.33287/44.98897/gray/b...
I'll have to check it out in person but I'm guessing it's a mountain bike trail or something, which is what this heatmap would be great for helping people find.
edit: gps source>https://news.ycombinator.com/item?id=7688240
I agree in general, though. Even if you have your own routes or know friends who have theirs, this helps break out of that rut and find some new rides. This would have been great for finding rides just across the river in Wisconsin (though as-is it wouldn't help find hills. I've always wanted a map similar to this that showed all roads in terms of both ridability and steepness).
I used to bike past here every weekend 1.5 years ago. I clicked on the link because it said Minneapolis but I didn't expect it'd actually be a place that I've been before.
In the Midwest, the trend is much less clear. For example, Madison, WI is covered with bikepaths, so there's no clear relationship. (Same for Anchorage, AK. You can see the main trail system, and it pretty much circles the city.)
In a lot of other places, the pattern is controlled by topography/parks (e.g. Birmingham, AL)
I'm limiting things to cities where I've lived, but does anyone else notice similar patterns?
http://labs.strava.com/heatmap/#16/-122.11865/47.78666/gray/...
Bike maps are likely created directly from GPS paths, which are unskewed. When you look at Bejing, you can easily see that difference.
E.g. http://academics.wellesley.edu/Chemistry/Chem105manual/Appen...
The Chinese Street maps are very precise but consitently inaccurate (the skew). The GPS tracks shown are relatively precise and accurate.
The change to the GPS signals (by not encrypting as much of the timestamp data) improved precision.
https://productforums.google.com/forum/#!msg/maps/ambcCIPTQi... says that the satellite images are correct and that the street maps are misaligned.
They claim that the offset is for security purposes but this simple fix shows that the excuse is bullshit. In reality the Chinese put all kinds of barriers to entry for foreign companies to make it difficult to penetrate the market. Another example is that you have to dual home your servers to both china unicom and china Telecom, as the traffic between the two networks is degraded on purpose. You thus have to purchase a special DNS service that differentiates between requests on the two networks.
In the same vein, the purpose of the great firewall is more economic than political, being a highly successful form of protectionism.
It seems a tiny amount of work compared to, say, gathering the actual street data itself; and it would be an useful thing to make this data available to the public like openstreetmap data.
I suppose it's because the results only reflect users of this app, and there's a correlation between app users and certain types of usage.