I have long suspected this and I’m sure others have to, but have you ever noticed the high proportion of Strava KOM and course records that are held by people who have recorded their ride using the Strava Android app. Furthermore it seems to be more prevalent when it comes to short downhill segment. Then the next question begs, why are the Android riders so good and yet the IPhone riders barely rate?
So in an effort to satisfy my curiosity I decided to take a short ride with both my Garmin 500 and Android app running. What follows are the results of a number of segment in my local area.
The rides with the lock symbol are the rides recorded on the Android Device.
Rides marked with a # indicate the current record is held a ride recorded on an Android device
Firstly Toshka’s Track # and I’ll admit this one caught me by surprise, despite the downhill section the Android app was slower, go figure.
Memorial Dirt Sprint # 5 seconds over 300 metres, that’s a big discrepancy
Ainslie Wall Downhill # This one was a draw
Down we go to Antill # Another 14 second win the Android app
I have more examples but I think you get the point.
Now I love Strava and I think it is a great way to get motivated to ride harder but I would also like to see a level playing field. When I have some more time I intend to pull apart some of the GPX files for these rides to see if there is anything strange occurring, I suspect the android device is not recording the data fast enough or accurately enough so it is possible the first data point captured for a segment is already a couple of seconds after the start and likewise the last data point captured is couple of seconds before the finish.
Besides making it impossible to beat some of these records, a real concern for Strava would be the posting the of erroneous data which could lead to a dangerous situation where riders start chasing down fake times in an effort to beat a record which really shouldn’t exist. Now I know riders need to take care when pursuing course records but I also believe the Strava Android App needs to record these times accurately. Having erroneous lower course records surely is increasing the risk by making riders think it is possible to ride faster than it really is.
So folks be safe out there, because that course record you are chasing, might not be possible.
I would also like to say, this only my opinion on this and I am happy to be corrected, so if anyone can explain why this happening I would like to hear from you.
Not the first time I’ve read this, or come across it chasing short KOMs. Someone has mentioned it was the sampling rate of the smartphone app being different? A plausible theory I guess if Strava do some kind of guesswork/compensation based on a lower recording frequency.
Best thing I’ve ever done is to turn off KOM notifications. Too many car drivers, group rides, and no surprises – smartphone using KOMs.
Strava – Chase _your_ best this season. Try better it next season.
Hi, I conducted a similar experiment last year and mainly found that the accuracy of iPhone recording was so poor that Strava could fail to match the segments up. My article is here: http://www.scarletfire.co.uk/2012/07/iphone-gps-fail/
You might also want to check out this veloviewer article which explains it very well, from a technical perspective:http://veloviewer.com/blog/alternative-leaderboard/
Most likely a device sampling/accuracy problem, not a Strava problem per se.