Gradients on Strava

[grellboy]

I live in Norfolk where to go climbing we quite literally have to make mountains out of molehills - I currently have two KOMs which my missus likes to mock ("Mountains!? They're barley even pancakes!"). But anyway, having found the steepest "mountain" around here I checked the segment on Strava and it says 0.2 19ft 1.2%!!! 1.2!!? Running over a crisp packet must more than 1.2%! Has anyone else found this - or am I simply clueless as to what a hill actually looks like?

 

[jefmcg]

KOM/QOM is just the term strava uses for fastest on a segment. If you were a runner, it would be "course record". There are plenty of KOM for downhills, and pancake flats. Most of them, I'd guess.

 

[Pale Rider]

The gradients on Strava baffle me.

If you track along this profile of Ditching Beacon it shows gradients in excess of 30 percent.

Which is about twice what it really is, but perhaps I'm missing something.

https://www.strava.com/segments/770191

 

[Elybazza61]

It does seem a bit random;even here in flatlandria there's some fairly steep inclines(albeit short) that don't get a mention but there are some which you would barely call a ramp gets catagorised

 

[T4tomo]

it might be momentarily on the inside bend of that first corner, but agree there is no sustained 30% gradient up there, plus strava is well known for having lots of minor inaccuracies in lots of aspects

 

[grellboy]

The gradients on Strava baffle me.

If you track along this profile of Ditching Beacon it shows gradients in excess of 30 percent.

Which is about twice what it really is, but perhaps I'm missing something.

https://www.strava.com/segments/770191

I've ridden ditchling 25 years or so ago (I say ride,I mean pushed the bike up!) but having just looked at the headline figures on your post I would say 9.1% avg gradient is probably about right. Would love to have another crack at it one day, although KOM might still just be beyond me!

 

[Pale Rider]

it might be momentarily on the inside bend of that first corner, but agree there is no sustained 30% gradient up there, plus strava is well known for having lots of minor inaccuracies in lots of aspects

I think the bloke who recorded the data must have been pulling a wheelie all the way up.

 

[grellboy]

It does seem a bit random;even here in flatlandria there's some fairly steep inclines(albeit short) that don't get a mention but there are some which you would barely call a ramp gets catagorised

Albeit. Do you remember the first time you realised that albeit was just one word and not three? Amazing!

 

[Mile195]

Albeit. Do you remember the first time you realised that albeit was just one word and not three? Amazing!

Not as amazing as the day I heard that Welsh speakers call a microwave a "pongty ping". That truly changed my life...

 

[Supersuperleeds]

There's a climb round here were the road signs tell you it is 10%, Strava gives it 5%

 

[Apollonius]

Does seem to be ridiculously inconsistent. Computers, I suppose. Round our way there is am allegedly Cat 4 climb that seems barely uphill which is less than 5km from one of the classic climbs up onto the Chase which was a categorized climb in the Tour of Britain a few years ago. Rated at about 6% by Strava and not categorized, but to my mind, the toughest climb for miles. Ah well, it's only a game.

 

[Shut Up Legs]

Ah well, it's only a game.

Now you've done it! The Stravilati will now hire assassins to make you (briefly) regret you ever said that.

It's not just Strava, though: my Garmin Edge 800 shows odd gradient readings when I'm riding, because the update rate is too slow, and it doesn't respond well to the small gradient changes every road has. Although the elevation data is normally pretty consistent, so the saved trip files would show reliable gradients (depending on whatever software you load it into).

 

[kiriyama]

I dont know the real ins and outs but i guess that Strava figures out gradients by overlaying your gps trail over pre existing terrain data (from Google or whoever) Non of this is 100% accurate and throws out some weird gradients. Especially on corners. There's a few VERY steep shprt climbs neer me (30%+) that only come out at about 5% of strava. I guess someone or some map making algorithm messed up the hight contours of the map. Cant be too bothered that they messed up a couple of hills in my small part of Yorkshire when they have mapped the whole world

 

[jefmcg]

from Google or whoever

For the record, strava use OSM (except, oddly the crop ride page is still google maps)

 

[Milkfloat]

For the record, strava use OSM (except, oddly the crop ride page is still google maps)

OSM has very little elevation data, so Strava will source it elsewhere. One of the curiosities of measuring ascent and descent is it depends on the accuracy you want to use. It is similar to the impossible to answer question 'how long is the coastline of the UK'.