Strava catorgorised climbs! Complete nonsense!!!

elnino23

So I ride in cobh cork al lot and we have some seriously steep climbs of decent distance and I've climbed them all with strava on and none of them are catorgorised!?! The one that is is the little stone bridge by the entrance to fota zoo which strava reckons is 28% and cat 4? And its about 300m long and flat for 250 of that! No way is that cat 4!!
The hill across the middle of the island which is long and steep is not catogorised and strava reckons its only 6%! Seriously does strava have the wrong topological info or what?? Climbing east hill is a serious climb as well and is at least cat 4.

Junior

And what would you like us to do about it for you ?

MajesticDonkey

Some of the segments on Strava are glitched. You can usually tell by looking at the VAM column in the leaderboard - if it's anything around 3000-4000, then the elevation on the segment is glitched. I can't find the segment you're talking about near Fota, so don't really know, but I'm fairly certain there isn't a 28% climb near it.

Apart from that, the segments are accurate. It helps to know how gradients work: if the climb you're on is 1000m long and you climb a total of 100m in elevation, the gradient is 10% - which is very steep. If it helps, the top section of Patrick's Hill in the city is around 25%. The category of the climb depends on both the distance and gradient, and it takes quite a bit to even make a cat4.

Darkglasses

Categorisation of climbs is far from an exact science even outside of Strava. You should expect it to be pretty subjective. Categorisation of climbs is even pretty inconsistent between different professional grand tours and races.

elnino23

MajesticDonkey wrote: »

Some of the segments on Strava are glitched. You can usually tell by looking at the VAM column in the leaderboard - if it's anything around 3000-4000, then the elevation on the segment is glitched. I can't find the segment you're talking about near Fota, so don't really know, but I'm fairly certain there isn't a 28% climb near it.

Apart from that, the segments are accurate. It helps to know how gradients work: if the climb you're on is 1000m long and you climb a total of 100m in elevation, the gradient is 10% - which is very steep. If it helps, the top section of Patrick's Hill in the city is around 25%. The category of the climb depends on both the distance and gradient, and it takes quite a bit to even make a cat4.

Cork road climb is the name of that segment and your right its def not 28% I don't think it's even 5%. My question I guess was what you answered that is there a well known glitch! There obviously is, This must show the climbing data is inaccurate in strava does it? The climb at cuskinny for instance that's pretty steep the segment is called "county hill climb champs" that's listed as 6% surely that's more than 6 is it? Does the garmin connect site give cat climbs? I guess it will give better data anyway

Chips Lovell

There's a categorised climb around Cork St in Dublin as well. GPS/elevation glitch. Hide the segment. Problem solved.

Cookie_Monster

If whoever creates the segment has bad GPS data then it sticks. Even if your GPS data is correct it'll still use the recorded segment data to show it rather than learning that it's bollox and correcting it. You can flag it but I've yet to see a flagged segment updated and fixed.

I've two 2nd places on cat 4 climbs which are in fact slightly downhill. Still they're 2nd places all the same:D

Lumen

This is one step away from compulsive hand washing.

nilhg

Flagging segments with bad elevation data doesn't achieve much, you can open a ticket and it'll be fixed in a day or so.

https://strava.zendesk.com/anonymous_requests/new

Just select segments/bad data and give them the direct link to the segment.

Alot of segments have average gradients which seem low mainly because when the segment is created flat or downhill bits get left at the start and finish, the Kildare version of The Wall is a good example of this, the main drag is over 11% but the segment average is 6.7%.

http://app.strava.com/segments/715134

mirv

Had a look at cobh cork there on the segment explorer.

There might be the odd steep bump or so, there's nothing that would be considered a long climb by the usual measure, especially seeing as the highest point by road on that island is at only 74m and most of the climbs tend to be less than a kilometer long at relatively easy gradients considering the short distance - they seem to be around 6-8%.

Not sure what you can do about categorising anything in cobh, short of growing some real mountains on the island.

ashleey

Lumen wrote: »

This is one step away from compulsive hand washing.

I have notified Strava of bad segments near me that were created by someone who had a miss recording GPS device. Strava have corrected them. I spotted them by using Veloviewer and their tips about absurdly high VAM records.

I felt unclean after for days.

MajesticDonkey

elnino23 wrote: »

Cork road climb is the name of that segment and your right its def not 28% I don't think it's even 5%. My question I guess was what you answered that is there a well known glitch! There obviously is, This must show the climbing data is inaccurate in strava does it? The climb at cuskinny for instance that's pretty steep the segment is called "county hill climb champs" that's listed as 6% surely that's more than 6 is it? Does the garmin connect site give cat climbs? I guess it will give better data anyway

No, because, as mentioned, it's whoever first records the segment data that the elevation data is taken from. Probably someone that used a phone with a crappy GPS sensor in it.

Do you actually realise how steep 6% is? Alpe d'Huez averages between 7-8%. Anything more than that is very steep. I think the gradient is an exponential value (but not sure) - as in, the higher you do in gradient percentage, the quicker the hill gets steeper.

mcgratheoin

MajesticDonkey wrote: »

I think the gradient is an exponential value (but not sure) - as in, the higher you do in gradient percentage, the quicker the hill gets steeper.

Percentage is just an easily understood way of measuring steepness, it's the elevation gain as a percentage of the distance traveled.

Mathematically speaking, the percentage (expressed as a decimal) is equivalent to the slope or the tangent of the angle. A hill of 45 degrees has a slope of 1 and a percentage of 100%.

Here's a graph showing how this value varies with respect to the angle

MajesticDonkey

mcgratheoin wrote: »

Percentage is just an easily understood way of measuring steepness, it's the elevation gain as a percentage of the distance traveled.

Mathematically speaking, the percentage (expressed as a decimal) is equivalent to the slope or the tangent of the angle. A hill of 45 degrees has a slope of 1 and a percentage of 100%.

A mathematical percentage maybe, but not percentage gradient. Like I said above, percentage gradient is (metres climbed) / (distance in metres), not necessarily related to the angle in degrees (or maybe I just haven't woken up yet).

EDIT: yep, I haven't woken up yet.

mirv

I'll chance my arm at semantics and would like to point out that the gradient of an incline is the vertical gain divided by the change in horizontal distance, rather than the distance 'travelled'.

mcgratheoin

mirv wrote: »

I'll chance my arm at semantics and would like to point out that the gradient of an incline is the vertical gain divided by the change in horizontal distance, rather than the distance 'travelled'.

dquirke1

mirv wrote: »

I'll chance my arm at semantics and would like to point out that the gradient of an incline is the vertical gain divided by the change in horizontal distance, rather than the distance 'travelled'.

Well spotted

In reality though, for anything remotely cycleable they're pretty much the same...

10% by one method gives 10.05% by the other...:p

hans aus dtschl

Categorised climbs in Cobh? I didn't think there was anything longer than a few hundred metres there... Little sprint hills