Garmin versus Strava...

[Faust29]

I've seen this discussion before, but maybe one of the tech gurus can put it into English...

I got a late start today, as I sat sipping espresso waiting for the wind to die down. I managed a nice loop from home to the East Ridge in Caspers... The Bell Trail is a lot harder and faster than the last time I rode it.

The total mileage was just under 30 miles. My Garmin says that the elevation gain was 4,055 feet. After syncing the Garmin data with Strava, Strava says that the elevation gain was 2,909. Normally, on a long ride, they report different values. This ride, however, always seems to generate the largest discrepancy. I'm guessing it has to do with the "rolling hill" nature of the Bell Trail...

Any ideas? And, which is more accurate?

 

[Cougar]

Garmin should be more accurate, as it actually has scientific thingies (barometers?) inside that actually measures the elevation differences. I know that the iPhone (running/recording the Strava app) is fairly inaccurate because it doesn't have that fancy tech. So what Strava does is overlays your route onto elevation maps and makes corrections as it deems necessary. I didn't think it did that with Garmin submitted data, but I only recently got my Garmin so I'm not as familiar with the idiosyncrasies like that.

 

[scottay]

..... I know that the iPhone (running/recording the Strava app) is fairly inaccurate because it doesn't have that fancy tech......

I think Strava should have separate top ten lists for any times taken off a phone......
.
.
.

 

[knucklebuster]

What Cougar said. The Edge 500 and up have a barometric altimeter (not sure if the lower level ones do). If you upload to Strava it "shouldn't" change the altitude gain from what your Garmin says. You can hit the "elevation correction" button on Strava, then Strava will recalculate based on their topo data, like it does with phone data. I've found when I've done that, sometimes it's more elev, sometimes it's less than what the Garmin says. Oh yeah, if you have weather moving in (high or low pressure) it can affect the barometer on the Garmin and make it seem off.

I do notice Stava adjusts the mileage automatically on Garmin uploads. My take, the Garmin GPS measures distance as if on a flat earth. Of course from our HS trig classes we remember the hypotenuse is a little longer than the base of the square triangle, so when you go up and down hills, we're hypotenusing . So usually the Strava mileage is a little more than the Garmin (in my case anyway). If you use the optional wheel sensor for the Garmin, it's the most accurate for distance. :ugeek:

 

[Faust29]

I'm running the 510, so that puts it in the "more accurate" camp. I didn't realize it had a barometer... From the sounds of it, I'm better off just looking at the Garmin data.


Sent from my iPhone using Tapatalk

 

[RS VR6]

Other thing that can affect the way a Garmin works is the amount of data that's stored on it. If you have a ton of rides stored on it...that can affect it too. In had a 500 that I never deleted rides on. The miles recorded was off compared to the 510. It was always short. After I deleted the rides...the mileage was accurate again.

I have my rides deleted after each upload.

When it comes to elevation... I just enable the elevation correction.

 

[Faust29]

Other thing that can affect the way a Garmin works is the amount of data that's stored on it. If you have a ton of rides stored on it...that can affect it too. In had a 500 that I never deleted rides on. The miles recorded was off compared to the 510. It was always short. After I deleted the rides...the mileage was accurate again.

I have my rides deleted after each upload.

When it comes to elevation... I just enable the elevation correction.

As I look at my unit, and see that I haven't emptied it since last spring! Oops...

 

[herzalot]

Of course from our HS trig classes we remember the hypotenuse is a little longer than the base of the square triangle, so when you go up and down hills, we're hypotenusing

"It was my understanding there was to be no math..."

Chevy Chase as Gerald Ford. You old bastards will remember that, but the two younger people on this forum have no clue.

 

[Faust29]

Of course from our HS trig classes we remember the hypotenuse is a little longer than the base of the square triangle, so when you go up and down hills, we're hypotenusing

"It was my understanding there was to be no math..."

Chevy Chase as Gerald Ford. You old bastards will remember that, but the two younger people on this forum have no clue.

We lost the youngsters with "high school trig". Trigonometry sounded like a disease... They rolled it all into Algebra 1 and 2, and did away with the name!

 

[UPSed]

We lost the youngsters with "high school trig". Trigonometry sounded like a disease... They rolled it all into Algebra 1 and 2, and did away with the name!

That's pretty sad. I had Algebra 1-2 in the 7th and 8th grade. I hope it is hit least as hard as Trig.

 

[MnMDan]

I haven't noticed any difference between my Garmin 510 Elevations and Strava Elevations. Only when the Garmin goes wacky with weather/rain/snow due to poor GPS elevation data do I use Strava to correct the elevation...and then its' pretty close for what it should be on a road ride, and within 15% for trails...and the Strava Phone Elevation tracking is poor at best...

 

[mtnbikej]

I upload my rides to 3 places....Strava, Garmin Connect and Geoladders.

Strava and Garmin Connect are usually spot on when it comes to mileage.....and normally less than 100 ft when it comes to accent to each other. Both like to adjust up on the accent numbers.

Geoladders on the other hand is always .5-1.5 miles less, and can bee 500-800 ft less than what the Garmin says.

For example my ride today:

Garmin GPS: 21.09 miles 1685' accent

Strava: 21 miles 1999'
Garmin Connect: 21.03 miles 1937'
Geoladders: 20.6 miles 1607'

 

[Faust29]

So... Basically I'm thinking about something that probably doesn't have an answer. There will always be some variance... Small at times, maybe larger at others.

I like the Garmjn connect site, and usually only look at Strava to see how I compare, and to see what the fast guys can do.


Sent from my iPhone using Tapatalk

 

[HBkites]

I uploaded Yesterday's ride to Strava. I use a Garmin 810, and it upload automatically (trough my phone) every time I save a ride.
The Garmin data showed approximately 2100 ft. When I pressed the elevation correction on the website, it changed it to 3900ft.
I don't have an issue with small correction but 45% off is forcing me to question the logic of using one (or both).

 

[Cougar]

I think something might be wrong with the correct elevation button in Strava

 

[Rob S.]

Accuracy is not just a matter of the device you use to record your data. In your case, you're using two different gps recorders, and two different software programs to analyze the data. The results for your total elevation gain are dependent upon the sampling frequency of your gps device, as well as the sampling frequency of the software that interprets that data.

I use an iPhone for serious route finding, as well as for goofing around with Strava. I find the accuracy to be quite good. A friend also uses iPhone/Strava, so we usually compare data. Last week we rode with about 1600 feet elevation gain, and our numbers were EXACTLY the same. Yesterday there were three of us with 3,000 feet of climbing. The maximum discrepancy was 110 feet. That's pretty good accuracy, I think.

Where I see huge discrepancies as you describe is when I use different software to analyze the gps data. I can take the gpx file and run it through Motion-X, Strava, and Utrack, and come up with substantially different numbers. Figuring out the accuracy is a complex calculus problem that requires more brain cells than I have remaining. But stick with me and try to visualize the elevation graph that you'd see for Strava, Geoladders, etc.

For ease of discussion, let's talk about a 10 mile, point-to-point, predominantly uphill ride. The graph you see on geoladders is a curvy, jagged line that shows the ups and downs of your route. Take the elevation at the ten mile mark, and subtract the elevation at the zero mile mark, and you have your elevation gain. Now do the math again, using the elevations at ten, eight, six, four, two, and zero miles. You get a different answer, but both numbers are correct!

Obviously, you don't want to use the first example, with two data points, because that doesn't take into account all the little ups and downs along the trail. Likewise, you don't want to measure altitude change over every grain of sand, because that will provide data that is equally useless.

 

[OTHRider]

The maximum discrepancy was 110 feet. That's pretty good accuracy, I think.

Or, the three iPhones were very precise. All three phones could be between 1,000 and 1,110 feet off. Us shooters argue about this all the time.

Accuracy is how close a measured value is to the actual (true) value. Precision is how close the measured values are to each other.

Low accuracy, High precision

High accuracy, low precision

High accuracy and high precision

Good god, this is almost as bad as correcting spelling. I'm doomed. [/nerd rant off]

 

[scottay]

.. ..... All three phones could be between 1,000 and 1,110 feet off. ..... ...

Exactly. I've read the older the phone, the worse the data also..
.

 

[Rob S.]

Us shooters argue about this all the time.

Some shooters need to argue, Duke. Others just empty our magazines rapid-fire into that tiny circle.

Just minutes ago I got off the phone with Kimber. My Solo Carry is going back to the mothership for the second time.

 

[OTHRider]

Touche. Bummer about the Solo........

 

[dstepper]

I have been playing with GPS units and difference topo software for many years. The only thing I found that was accurate for calculating elevation gain was a climbers altimeter I used to use before I lost it. The expensive Garmin I also lost on a ride had a very accurate altimeter in it would also calculated the grade percentage in real time...I stopped using Garmins on every ride because as I lost too many of them. The Garmin I use now etrex30 does not sync to any of the online sports websites. But the National Geographic topo mapping chip it uses is excellent, in shows water, cabins, mines and other landmarks. I use GPS for exploring not for keeping track of my rides. Tried using apps for the phone but since I ride often out of cell service and would forget to load maps before I left it did not work well for me. Then I would carry a extra battery pack because my phone battery would go dead before I finished the ride. Good GPS much easier.

Dean

 

[Faust29]

Thank you all for the different perspectives... I knew I was asking the right crowd!

I think in this case, it just boils down to examining the same data set with two different software interpreters, as Rob S. mentions above. This particular route always presents the widest discrepancy, and I was curious if others had similar issues.