Results 1 to 12 of 12
  1. #1
    RoadBikeReview Member
    Reputation:
    Join Date
    Sep 2011
    Posts
    2,888

    Strava versus Garmin Connect - data discrepancies

    I have a Garmin 500, and I notice that when uploading the data to Strava and Garmin Connect, there are discrepancies. For example, some discrepancies are:

    - Strava showing my max speed as greater (and probably wrong) than Garmin Connect
    - Strava showing elevation gain to be less than Garmin Connect
    - Strava showing total ride time to be more than Garmin Connect

    For those using both Strava and Garmin Connect websites, are you guys finding this to be the case? I tend to think that Garmin Connect website is more accurate.

  2. #2
    RoadBikeReview Member
    Reputation:
    Join Date
    Aug 2008
    Posts
    1,937
    It's a lot of data. They both use different smoothing algorithms so as to display data that's useful. Different methods of dealing with the data yield different results. I don't know which is more accurate.

  3. #3
    RoadBikeReview Member
    Reputation: mann2's Avatar
    Join Date
    Oct 2012
    Posts
    306
    Yes to the above. Calorie counts vary greatly as well. However, I tend to rely more on the Garmin side of things for the simple reason that i have more historical data there. That said, I like Strava for its challenges and other motivational benefits
    check my review page below!

    www.roadiemanila.com

  4. #4
    RoadBikeReview Member
    Reputation: Alfonsina's Avatar
    Join Date
    Aug 2012
    Posts
    734
    Calorie counts are only worthwhile with the HRM. The 500 almost doubles my calorie count with no HRM and I can tell you that info is worse than useless when it so overestimates like that.

  5. #5
    Moderator
    Reputation:
    Join Date
    Sep 2007
    Posts
    3,481
    How big are the discrepancies? Everything is typically derived using math from lat/lon/elevation recorded every few seconds.
    Quote Originally Posted by aclinjury View Post
    - Strava showing my max speed as greater (and probably wrong) than Garmin Connect
    That's strange. It may be that Garmin is using the max speed actually recorded by the device, while Strava is figuring out from the raw data.
    Quote Originally Posted by aclinjury View Post
    - Strava showing elevation gain to be less than Garmin Connect
    Not surprising - this is very dependent on algorithm.
    Quote Originally Posted by aclinjury View Post
    - Strava showing total ride time to be more than Garmin Connect
    Elapsed time versus moving time?
    Old La Honda in less than 20 minutes! Or you can watch race video from the low-key hill climb on Welch Creek. More at www.biketelemetry.com.

    "I think," said Christopher Robin, "that we ought to eat all our Provisions now, so that we shan't have so much to carry.", Winnie-the-Pooh, A. A. Milne.

  6. #6
    RoadBikeReview Member
    Reputation:
    Join Date
    Apr 2013
    Posts
    8
    As others have said the smoothing algorithms employed can make a huge difference to the data.

    I use sport tracks which allows you to custom set the degree of data smoothing using sliders. On long rides the values presented, depending on the settings can be wide ranging if you play around with them.

    The trick is to use something realistic, but there is always a trade off between the variables.

  7. #7
    RoadBikeReview Member
    Reputation:
    Join Date
    Sep 2011
    Posts
    2,888
    My moving time shows 5h:58m on GC, 6h:01m on Strava. That's a 3 min difference.

    Elevation gain on GC is about 10,400ft. Strava shows about 10,300ft. This is actually a pretty acceptable difference, right?

    But here's something weird. My Garmin 500 has a built-in barometric sensor, so as such, GC and Strava website will tend to use the data from the Garmin 500. However, GC website has a feature where you can enable "elevation correction", and by doing this, you are forcing GC to use topographical survey data, well if I do this, then my elevation gain shows over 16,000ft!!

    Very fuzzy math going on here.

  8. #8
    Moderator
    Reputation:
    Join Date
    Sep 2007
    Posts
    3,481
    Moving time is probably different algorithms for detecting zero speed. Basically they will have a cut-off speed and below that you are considered to be not moving. The reason is that GPS wanders around even if you are still. Riding at very slow speed (e.g. dawdling waiting for a traffic light to change) would expose the difference.

    Elevation gain is very very close. This is competely acceptable.

    The topo elevation can vary widely depending on the terrain. Well designed roads for long climbs tend to be designed for a continuous grade. A small error on the GPS data can move you around and the topo elevation points are not very dense leading to false ups and downs on a continuous climb. Also if you are riding near a cliff (e.g. a canyon) then small positional errors can give huge topo elevation errors. They use smoothing to mitigate this, evidentally not too successfully on your ride.

    GPS users often think that they should get 100% accuracy on their data. Garmin have done very little to educate them to set reasonable expectations. Note that people have crossed oceans and travelled to the moon on less advanced technology than this!
    Old La Honda in less than 20 minutes! Or you can watch race video from the low-key hill climb on Welch Creek. More at www.biketelemetry.com.

    "I think," said Christopher Robin, "that we ought to eat all our Provisions now, so that we shan't have so much to carry.", Winnie-the-Pooh, A. A. Milne.

  9. #9
    RoadBikeReview Member
    Reputation:
    Join Date
    Feb 2002
    Posts
    19,237

    Calories are basically a guess

    Quote Originally Posted by Alfonsina View Post
    Calorie counts are only worthwhile with the HRM. The 500 almost doubles my calorie count with no HRM and I can tell you that info is worse than useless when it so overestimates like that.
    While having the HRM might improve the accuracy of the calorie count there is little correlation between HR and calorie expenditure. At the same power output HR can easily vary by 30 bpm between two riders. With a decent algorithm speed, distance, and elevation gain would allow a much more accurate calorie number than is usually provided by HR based calculations.

  10. #10
    RoadBikeReview Member
    Reputation:
    Join Date
    Sep 2011
    Posts
    2,888
    Quote Originally Posted by ukbloke View Post
    Moving time is probably different algorithms for detecting zero speed. Basically they will have a cut-off speed and below that you are considered to be not moving. The reason is that GPS wanders around even if you are still. Riding at very slow speed (e.g. dawdling waiting for a traffic light to change) would expose the difference.

    Elevation gain is very very close. This is competely acceptable.

    The topo elevation can vary widely depending on the terrain. Well designed roads for long climbs tend to be designed for a continuous grade. A small error on the GPS data can move you around and the topo elevation points are not very dense leading to false ups and downs on a continuous climb. Also if you are riding near a cliff (e.g. a canyon) then small positional errors can give huge topo elevation errors. They use smoothing to mitigate this, evidentally not too successfully on your ride.

    GPS users often think that they should get 100% accuracy on their data. Garmin have done very little to educate them to set reasonable expectations. Note that people have crossed oceans and travelled to the moon on less advanced technology than this!
    thanks for the explanation. I guess I should just stick to one website to keep it consistent

  11. #11
    chamois creme addict
    Reputation: Eric_H's Avatar
    Join Date
    Feb 2004
    Posts
    1,468
    Garmin Connect recently changed their elevation correction data and algorithm I believe. If I use this feature now I find I always get more elevation gain on routes than I should, comparing to the same routes ridden before. For example a regular route with ~750m of climbing gets "corrected" to ~1200m on Garmin Connect. Before this was not the case, using elevation correction would clean up the data but the overall gain/loss did not change much.

    To get the most accurate elevation data from your 500, make sure you leave it outside to "acclimatize" for about 20 min. For sure there is a temperature dependency with the barometer in the 500, and I find I have much more elevation drift in the winter if I take the Garmin from inside the house (19 C) and start riding outside (say 5 C) without letting it acclimatize. Also, turn it on and let it find the satelites before you take off. I have a known elevation at my house of 15 m and have entered this value as an elevation point in my Garmin, and I also have the current elevation on my display. When I first turn it sometimes reads -70 m, but once it finds the sattelites it will drift back to within 15 m, +/- 5 m. But if I press the "Start" button right after I turn it on, my elevation point entered in the Garmin sets the elevation to 15 m, but the unit is effectively still drifting. So it is much more accurate to let it "drift" into roughly the correct elevation before starting the ride.

  12. #12
    fu
    Reputation: ph0enix's Avatar
    Join Date
    Aug 2009
    Posts
    2,182
    Strava is known for reporting ridiculously high max. speeds.
    From what I understand, Garmins often overestimate the elevation. I recall reading something about leaving them turned on for a few minutes before riding so that they can settle and figure out their starting elevation correctly.
    My other chainring is a 39...
    Strava profile

Similar Threads

  1. Runkeeper/Strava to Garmin Connect possible???
    By RobbMaxx in forum General Cycling Discussion
    Replies: 0
    Last Post: 07-11-2012, 09:41 PM
  2. Replies: 38
    Last Post: 02-03-2012, 05:57 PM
  3. ? about transfering ride data or courses on garmin connect.
    By jaydub_u in forum General Cycling Discussion
    Replies: 4
    Last Post: 12-30-2011, 05:02 PM
  4. Discrepancies between Garmin and Strava?
    By Fireform in forum General Cycling Discussion
    Replies: 12
    Last Post: 10-24-2011, 02:31 PM
  5. Strava vs Sports Tracks vs Garmin Connect vs TrainingWithGPS...
    By nyvram in forum Components, Wrenching
    Replies: 6
    Last Post: 01-30-2011, 06:50 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

Hot Deals

Contest

Tour De France

Latest RoadBike Articles


Latest Videos

RoadbikeReview on Facebook