Strava elevation gain

[IanG1]

Anyone give a reason while there can be large elevation discrepancies when uploading the same ride but using a different method of recording.
For instance I use an android phone and had a figure of 832 feet for a ride which was very close to the overall gain figure recorded on my Boardman cycle computer. However a friend on exactly the same ride recorded 1230 feet using his Garmin. Also why is it when I first upload a ride the elevation figure is around three times the actual figure for the ride but later comes down when the ride calories are put into the record page?

 

[jdtate101]

Because the Garmin uses a barometric sensor which is generally more accurate, but can be susceptible to pressure changes from bad weather and rain blocking the port. The phone just uses GPS altitude data which can vary a lot in accuracy. Overall I've found the Garmin's to be more accurate whereas the phones tend to overstate both speed and height gain. You can always ask Strava to correct the hight data if it's gone haywire, it usually takes about 20mins. IMHO Strava massively overeggs the calories burnt, so i'd just ignore that bit of data.

 

[IanG1]

Thanks for the reply, my boardman computer uses barometric pressure and this pretty much matched the gain I recorded on the phone?

 

[vickster]

IMHO Strava massively overeggs the calories burnt, so i'd just ignore that bit of data.

Not as ridiculously as a garmin does. I actually find iPhone Strava fairly accurate on calories based on my 30-40 per mile rule of thumb

 

[jdtate101]

Which is odd because the Garmin is always the lower of the two for me, by a factor of 25% at least.

 

[vickster]

That is odd. I am not a lightweight so maybe they've fubar'ed the algorithm for heavier /older folk /females? I would say or me, 25-50% more on the garmin!

 

[summerdays]

One of my segments it claims is a Climb Cat 4 - I hate hills I choose routes to avoid them if I can help it and that route is almost flat so I've no idea how it thinks there is 200 ft of climbing on it.

 

[helston90]

My Strava on my android phone is always quite a bit out when I first hit stop- sometimes 1000ft out, so I know to ignore it, look at it an hour or two later and it's the same (ish) as everyone else on the ride.

 

[Crackle]

One of my segments it claims is a Climb Cat 4 - I hate hills I choose routes to avoid them if I can help it and that route is almost flat so I've no idea how it thinks there is 200 ft of climbing on it.

It's the mapping data it uses. I've got four 25% climbs along the prom. The flat as a pancake prom. Still, my power output is suitably impressive because I can do all four 25% climbs at 25mph, cool.

 

[Rasmus]

It's the mapping data it uses. I've got four 25% climbs along the prom. The flat as a pancake prom. Still, my power output is suitably impressive because I can do all four 25% climbs at 25mph, cool.

Actually, these fake cat 4 climbs are more likely to be the cause of a glitch in elevation data from a barometric altimeter (ie garmin 500/800). Strava will automatically create a segment if your ride includes elevation data above the threshold (distance in m * average gradient in % >8000).

Apart from knowing the terrain in reality, you can recognize these segments by their generic names (eg "A841 climb", or even "M8 climb") due to auto-generation.

You can tell strava to "hide" the segment on your ride page, and it will be hidden automatically on future rides.

My garmin 200 never produces glitches like these, but strava does not provide the option of trusting the data. Meanwhile, barometric altimetry data is treated as 100% trustworthy, but that is clearly not the case.

 

[Crackle]

It might do but mine has no barometer.

 

[Rasmus]

Segment matching is based on location only, so once the segment is there it will show up on your ride, even if both your device and the maps know the "hill segment" is false.

 

[Crackle]

Segment matching is based on location only, so once the segment is there it will show up on your ride, even if both your device and the maps know the "hill segment" is false.

So the first person who created it, likely introduced the error?

 

[Rasmus]

So the first person who created it, likely introduced the error?

Correct, except the segment is not created by the user, it is done automatically by Strava.

For an example, see this ride from the CC trip around Arran last year. Mike's garmin glitched out, Strava made the segment and immediately gave him a KOM award, before filling up the leaderboard with all the other people that have ridden that piece of road.

 

[summerdays]

I'm keeping my QOM on the glitch bit ..... its likely to be the only Cat 4 one I ever get!!!