Google Maps accuracy for running

Depends on how accurate you are. If you're mapping out a trail run you'll probably map some straight lines where you're running a curve. So it'll be too short. If you're mapping out a run in the city it's probably too long since you'll be cutting some corners when you run.

why not just use something designed for it like http://www.gmap-pedometer.com/ instead?

if you're running on the sidewalks then probably pretty accurate

Every time I measure a track with gmap-pedometer, it's within 1 meter of 400.

Does gmap-ped assume we run the tangents?

I don't know about the US version, but I know up north on GMaps, the auto map probably has set amount of points on its route. It definately does NOT assume that you are running the tangents as evidenced when I use manual I get a much shorter route. Also, sometimes the road on 'map' version is 10m or more off than what it is on 'satelite'. That doesn't happen often though.

Unfortunately where I live, they took the shot in an April so I can't get good access of the running trails.

^ meant to say September, not April

Mr Crane wrote:

Does gmap-ped assume we run the tangents?

You can draw the tangents yourself, manually.

You also probably run tangents less than you think, unless you're racing.

I use Google Earth to map out my road courses prior to the actual measurement (bike-jones counter). If you are carefully and zoom in, you can get within 10-15 meters of the true distance of a 5K course - even if it has several turns.

How accurate do you need to be? For general training purposes google maps is accurate enough.

I agree, it's accurate enoough for road training purposes.

My thoughts have wondered about steep elevations and the distance reading. If I take a spot before a steep hill and after a steep hill will it give me the straight line distance or the actual longer distance of going up and down the hill between the two spots?

Any smart guy out there know the answer? (lets assume its mapmyrun and its the follow road function)

Assuming a hill has no more than 1m elevation per 5m run (20%), lets bring out pytagoream's rule

sqrt((1*1)+(5*5)) = sqrt(26) = 5.0999

so if the hill is 300m long in this regard, then it is only 6m long. Granted this type of thing might add up on a run, but if the rise is smaller than the run, then the true difference is almost identical to the run distance.

I doubt GMaps or any other of these online tools accurately helps measure hills because they are measured based on highly accurate GPS pictures (relative to the actual layout of the cities itself), not off of terrain. I have definately found the GMAPs terrain to be way off at times (ie, 20m elevation knowing its at least 30m high at times).

As we are discussing this is relative to training and in this case, only a headcase needs the wheel to measure any road distance greater than 300m (and that is provding there is no nearby track).

Now if we were talking about race preperation (race directers only), then yes it is better to bring out the wheel or bike. Also, in the case that your nearby trails are covered by trees, it is better to bring out the wheel there as well since it is way more accurate than a garmin (which I have found to be severly off at times)

I know for certain that gmap doesn't account for elevation, because I went to the site and checked by measuring a hill and the flat ground next to it.

It took 2 minutes, you lazy bum.

Bad Wigins wrote:

I know for certain that gmap doesn't account for elevation, because I went to the site and checked by measuring a hill and the flat ground next to it.

Then how does GMap have the capability to give you an elevation plot (granted, it's not completely accurate)?

See the previous poster's example of the Pythagorean theorem. 20% grade is a crazy grade and your measured hill is nowhere near that.

Say you went on a 10 mile run and were constantly climbing a 5% grade (1m gain every 20m). After covering 10 miles (as measured on a flat projection, you are now 0.5 miles higher up and the actual distance you ran on the road is 10.0125 miles or about 20 meters longer than 10 miles.

Don't they teach this stuff in school anymore?

Google Maps/ Gmaps can give you a rough estimate of the elevation profile but it does not use this data in calculating the distance.

Um, yeah wrote:

Then how does GMap have the capability to give you an elevation plot (granted, it's not completely accurate)?

See the previous poster's example of the Pythagorean theorem. 20% grade is a crazy grade and your measured hill is nowhere near that.

Say you went on a 10 mile run and were constantly climbing a 5% grade (1m gain every 20m). After covering 10 miles (as measured on a flat projection, you are now 0.5 miles higher up and the actual distance you ran on the road is 10.0125 miles or about 20 meters longer than 10 miles.

Don't they teach this stuff in school anymore?

Strange Times wrote:

Google Maps/ Gmaps can give you a rough estimate of the elevation profile but it does not use this data in calculating the distance.

But the point is that it doesn't really matter. Who is going to be able to accurately plot a pancake flat 10 mile loop within 20m? Whose training is genuinely dependent on whether their run was 10.12 miles or 10.15 miles?

There are bound to be plenty of little errors as well with the image stitching as well.

The accuracy of elevation of course depends on the density of waypoints. It draws straight lines between waypoints so if you're manually plotting a course it's up to you to zoom in close enough and click every few meters if you're running in hilly terrain. If you use the "follow roads" feature it does it for you.

Google maps seems way more accurate than my Garmin based on the waypoints.

It is very very accurate. Last year I was able to use it to show that the Lakefront Marathon in Milwaukee was about 56 feet too long and as a result my time should be adjusted. While we disagreed as to the amount of time to be taken off, we settled on 2.5 seconds, which got me into Boston.