Koonstheman

There is lots of debate on this general topic out there, but I'm new to Strava and I thought I'd share/ask.

When I went to the Garmin Edge 500, about 10-12 years ago, I saw that I lost as much as 5% of distance from all my rides, which I had hitherto measured with a wired computer and the wheel magnet, with the wheel circumference programmed. Just a few weeks ago, I uploaded recent versions of those rides, ran the Strava correction feature, and I got my 5% back. I used my route map on Strava to compare each of my segments to the distance registered on the Garmin, and I've reached no theory about why one gains fractions of miles on some and not others. The lines on the Strava map do reproduce my routes faithfully.

So, because of the old days with the Cateyes, I instinctively leaned towards believing the Strava numbers, but is my understanding correct that all Strava does is take any specific segment and average the measurements of a bunch of GPS-based applications that people have uploaded? I wonder if they have improved their sample or methodology compared to that which I have seen mentioned on the internet , but several years ago.

Their site also mentions error inherent with climbing. Has anyone noticed any negative impact of that on the numbers, either on Strava or their Garmin?

caloso

Are your routes relatively lumpy? If it assumes a flat line, that could be the difference.

https://support.strava.com/hc/en-us/...-is-Calculated

Koonstheman

Thanks for replying. Not unusually so - 40-50 feet of average incline per mile, and I also don't see the steeper segments resulting in a wider discrepancy. At first, I assumed the vagaries of GPS (tree cover, etc), but then if Strava is just aggregating uploads, I'm not sure why it'd be different. I'm also not entirely sure I'm interpreting the alleged Strava methodology correctly; it seems to assume that someone else must have uploaded data from that very segment. What no one has? Or maybe it is just getting GPS readings from other sources - not uploaded by riders/

Iride01 

Is that GPS distance you are using on your Garmin 500 or do you have a wheel sensor? If a wheel sensor, then calibration can be all of your issue.

Personally I don't know why I should worry about whether I went 40 miles or only 38 miles.

Your cars odometer only has to be accurate to 7% by law. Though most are way more accurate than that.

Steve B.

I measured accuracy of my original Garmin 810 vs. a Cateye wireless and found it short about 1 mile in 50. That's on road rides, fairly straight roads, not terrible tree coverage and no wheel sensor. That was accurate enough for me and I've never bothered checking. I often create routes in RWGPS and find my 1030 is within a 1/10th in 20, so good enough.