Has anyone used this site to measure their runs? A lot of the loops I'm running in Minneapolis have been measured by a wheel (around lakes) and are coming up short on this site. Just wondering if anyone else has had problems with this? Kind of a cool tool though for creating new runs.
curves are going to be longer than tangents. good tool for right angle roads though.
It measures tracks pretty accurately
I tried making the individual sections small enough to accomodate for curves. must not have worked, or maybe i started running 15 seconds per mile slower this week on training runs? sweet.
It's pretty damn accurate
Google Earth has a far superior measuring tool (that follows the topography in google earth; ie accounts for elevation changes that are large enough)
The error due to elevation changes is insignificant relative to the error from other sources.
If you are careful the gmap-type sites are accurate to about 0.5% when compared to calibrated-bicycle measured courses. I haven't done much experimenting with Google Earth. It might be a little more accurate.
Using a measuring wheel isn't particularly accurate. It is hard to keep it straight and maintain constant ground contact. Plus, the wheel itself may not be that accurate. I checked a cheap one I bought against a steel-taped calibration course and it was off by about 4 inches per 100 feet. I would trust your gmap measurement more than the wheel measurement.
How accurate do you need to be? How much is it off from your wheel measurements?
FYI, since google has changed to TeleAtlas (I believe?) for their map data the accuracy has decreased.
badgoogle wrote:
FYI, since google has changed to TeleAtlas (I believe?) for their map data the accuracy has decreased.
true. the street on which i live disappeared off of gmaps. i had been using it, but have since ceased.
Nope. Google Earth does not account for elevation changes, and in steep topography this can account for errors of as much as 5% on a given course, and even more.
When your gmaps route is going around curves, you need to zoom way in and click a lot of short distances. It's easier to do this in satellite mode instead of map mode. If you go short-cutting in your map by clicking across huge chunks of land, then naturally your map results are going to be shorter, just as your run would be shorter if you actually ran the giant shortcuts that you clicked.
usatf.org/routes works exactly the same way.
To illustrate:
Here's an example of a clumsy mapping job that is definitely going to come out short. Zoom way in and look at how the map shortcuts huge chunks of ground, instead of following Burke-Gilman precisely, even though following Burke-Gilman was obviously the author's intent. This is especially obvious just north of Gas Works Park.
http://www.usatf.org/routes/view.asp?rID=4716
Now, by contrast, check out this map; zoom in in Satellite mode and observe how the curves around the Shedd Aquarium and the Adler Planetarium are more or less precisely drawn.
http://www.gmap-pedometer.com/?r=211430
Okey dokey?
Note that this degree of accuracy is harder to attain away from urban areas, where the satellite photos on gmaps do not zoom in as much as they do in the cities. Out in the country you're going to have to trust the map mode. Still zoom in as much as is practical, and follow the curves precisely as you click through your map.
I use this to check if my garmin 205 is accurate and over every course I measured, it was exactly the same. what do marathons use to measure there courses? i feel in the 70's 80's what ever method they used, couldnt of been exactly accurate.
As far as I know USATF course certifiers still use a "Jones Counter", which is a device that goes on a bicycle wheel after having been calibrated (on that bike, with, hopefully, a consistent tire pressure) using a short distance that is known via steel tape measure.
Keith Stone is the real expert on this stuff.
MarathonMind wrote:
Nope. Google Earth does not account for elevation changes, and in steep topography this can account for errors of as much as 5% on a given course, and even more.
Can you outline your math for this 5% error due to topography?
Even if you were running up (or down) a slope that gained/lost 200 feet every 5280 horizontal feet (pretty steep for running), your actual distance would be 5283 feet, or 0.07% off measured.
I used the Pythagorean theorem. What did you use?
are you sure about this? I really hope so, cause I've always felt I was running faster than gmap-pedometer told me I was going...particularly in the hilly areas around my house
I've been using gmaps all the time lately - i find that if i take the time to zoom in and use smaller increments - the distance is as accurate as anyone could possibly want. Even if I'm being lazy and don't use increments that are quite as small, the distances are still very accurate - we're talking 7 to 20 miles - now, if you're worried about a few feet for a 1/2 mile then you should be on a track anyhow...i'm pretty anal about my running - always timed, measured distances, daily log etc...but give me a break - if you think that gmap distances are way off, you've probably just been running alot slower than you thought you were - (by the way, this miscalculation is normal human nature)....
Asterix,
You are certainly correct (no surprise there). I remember answering this elevation thing in another thread a month or two ago. For any reasonable course, ignoring elevation produces less than .5% error. For example, using a one mile long hill that has a 10% grade (5280 feet without elevation) yields approximately 5306 feet, or an extra 26 feet, for just under .5% long. How many people run races or courses with an AVERAGE grade of 10%?
To get 5% error you'd need a hill with about 32% grade. That's approximately the grade of the steepest roads on the planet. So, unless your running route includes rock climbing and sky diving, you'll never get an error of 5% or more.
Math is your friend.
Wolfgang Grajonca wrote:
When your gmaps route is going around curves, you need to zoom way in and click a lot of short distances. It's easier to do this in satellite mode instead of map mode.
When using mixed mode, I find the roads on the map don't always match up with roads on the satellite photo. This alone has changed some things for me.