Map my ride vs google maps elevation values

C R

Veteran
Location
Worcester
I wonder if anyone else has noticed this. I use map my ride to track my mileage, but I normally check my routes in google maps before a ride, only recently I have noticed that the elevation reported by map my ride is significantly different from the elevation reported by google, even though the distance is more or less the same, for instance, this is the route I did yesterday according to map my ride:

upload_2018-7-23_14-47-20.png


whereas according to google the elevation is

upload_2018-7-23_14-48-37.png


The difference in distance is about 0.5 miles, but the difference in elevation (~300feet more in map my ride) is a lot. I would like to believe what map my ride says, but somehow I expect google is more likely to be correct.

Anyone have any ideas what might be going on?
 

Globalti

Legendary Member
I asked the very same question a few weeks ago on here and received the very same answer: "No system can be guaranteed - just be sure always to use the same system".
 
  • Like
Reactions: C R
OP
C R

C R

Veteran
Location
Worcester
There are various ways to guess the total amount of ascent in a ride. It's much harder than just working out the overall distance. The different ways each have their own systematic bias so will come up with different results. But they will be consistently different.

Route planning software will have fairly coarse data on terrain elevation. So if the route goes from A to B, it will try to figure out the elevation at A (probably by searching for the nearest known elevation to A in its terrain database), do the same for B and subtract A from B. This is dependent upon how good the elevation data is. If there's a dip or a hump between A and B, it won't notice it. So route planning elevations will be smoothed out and will underestimate the ascent. And different route planning sites, with different algorithms, and different mapping elevation data will give different results.

Ride tracking software from GPS units has elevations recorded with the track. So that should be easy, right? Well, no. If the device just uses GPS elevations these are notoriously inaccurate. If the device has a barometric altimeter this may not have been correctly calibrated and may be affected by pressure drift during the ride. So the elevations on the tracks will be very variable and will over-estimate. Most sites use some kind of smoothing algorithm to calm them down. And different sites use different algorithms giving different results.

Q: Which is right? A: All of them. Or, perhaps none.

Never compare elevations returned by different methods. Choose one method and stick to it.

Also ask yourself. What do you mean by "correct". How can you determine the "gold standard" measure of amount of climb? Contour counting? Nope, to coarse, there could be ups and downs in between the contours. A GPS? Nope, too variable as noted above. A mapping site? Nope, they are all different.
So I don't need to feel guilty if I quote map my rides values ;).

Thanks for the explanation, it is more or less what I had guessed, just I didn't expect such a difference considering the distance tracks much more closely. I use my phone for recording my rides, so I imagine map my ride is using some kind of gps estimate.
 
OP
C R

C R

Veteran
Location
Worcester
I asked the very same question a few weeks ago on here and received the very same answer: "No system can be guaranteed - just be sure always to use the same system".
Sorry, I did search before posting, but obviously my cc search fu is lacking a bit.
 

MontyVeda

a short-tempered ill-controlled small-minded troll
I posted a thread a few years ago wondering why Map-my-ride's elevation graph was completely inaccurate... I much prefer google's.

edit... The reason i knew MMR's elevation was inaccurate was because I'd ridden eight miles along a canal tow path with not a single lock or roaming bridge so the graph should have been dead flat, but it showed around 36ft of ups and downs.
 
Last edited:
OP
C R

C R

Veteran
Location
Worcester
The Audax folk have a standard for climbing points. Does this make sense? http://www.aukweb.net/results/aaa/aaaclmb/
Thanks @twentysix by twentyfive, it makes sense, I will have a play with it later. I somehow expected that all apps would do some estimate based on contour plots, what I find strange is that the discrepancy is so large.
 
Top Bottom