[Closed] So, is Strava wrong or vertibrate books wrong

I'd say the book was out, by quite a distance on some of the routes. I was flicking through the White Peak book a couple of days ago and it had one of the routes listed at over 1500m for a 4ish climb ride around Ladybower (yes I know thats the Dark Peak 🙂 ). My guesstimate would have been about 1100m.

Edit: I did the same loop as you from the book in 2016, my Garmin 800 recorded 1755m!

Strava is usually wrong (about most stuff, it should really only be used as a guide) and there's a good chance the book was written using GPS data, which could also be wrong. So the question is probably 'which is least wrong'?

Trust.
No.
One.

I don’t think Strava calculates elevation but rather just accepts what it is given by the GPS device.

On group rides I always see a wide variation in elevation amongst the group.

Try putting the route into RideWithGPS or Komoot and seeing what the estimate there is for comparison, it'll give you another benchmark figure to compare the two readings you've got.

I think the book overstates some rides. Not sure if it's trying to make the rides out as harder than they are to put off the "less well prepared" rider or act as a cautionary tale to people taking it on...?

Strava themselves acknowledge the difficulties in recording it accurately.

Strava Elevation Info

If you are talking to other people about it, then it is definitely 2300.

https://support.strava.com/hc/en-us/articles/115001294564-Elevation-on-Strava-FAQs

Edit: beaten to it^.

the most accurate way is probably to plot the route out on ridewithGPS to get the elevation.

The most accurate way would be to plot it on a paper OS map and count the contours. Digital Elevation Models are notoriously lacking in granularity.

Is that Strava after asking strava to correct elevation on the web version? That is then anchored to a base map and seems to agree with plotting on a base map like OS.

If you log with a phone it automatically checks against the map. If you log with a GPS Strava just accepts the random number generated by the gps

there’s a good chance the book was written using GPS data

Or rather data captured from the barometric altimeter on a GPS device.

Sometimes they can be very accurate (tested against a known elevation change), other times they can be disappointingly inaccurate. My Garmin 520 was at least 200m out on a 800m-ish ascent ride yesterday. And it wasn't even raining (it's usual Kryptonite).

What Ampthill said, have you tried 'correcting' the elevation in Strava (desktop version)?

My Garmin Instinct has a tendency to under calculate ascent on some rides (normally trail riding where its up/down more rapidly) so i go into Strava on my desktop and, under where the elevation is shown for a ride, there's a correct elevation link. You can change it back too

The most accurate way would be to plot it on a paper OS map and count the contours. Digital Elevation Models are notoriously lacking in granularity.

Back in the days of paper-based maps as standard, there were all sorts of ways to attempt to correct for conversion of a 2D representation into 3D terrain, most of which are really time consuming and still not that accurate.

You get a trail that "rolls" up and down a contour, never quite showing as gaining 10m but you can still gain a few dozen metres over the length of the trail as it rises and dips 5-8m at a time.

Ride the route again and see what it says this time around!

"Correct elevation" will come out a bit under an accurate figure for similar reasons as Crazy Legs describes above, IME.

Still useful when your Garmin altimeter goes on strike though.

My wife and I both have Garmin watches, both will be fairly close to each other if we're out together, but if I'm to do the same activity a few days later in different barometric conditions, there can be quite a difference.

In general it's not bad, as I know my local mountain is 874m from it's start at sea level, my garmin will generally have me doing around 920m of ascent, which is pretty accurate give the wee ups and downs on the ascent.

Last nights ride had 556m of ascent on the garmin, I'll correct it on strava and see what it does.

Edit - corrected that, and a similar ride I did last week, with a slightly different route to the trig, and they're fairly close, 510m and 501m, so not bad.

Does it matter anyway?

With the exception of the Lakeland 200 and anything published as a route guide in MBR, most rides are planned around easier climbs. You could do a loop with 600m elevation made up of one big road/fire road climb and an off road descent, and it would be an easy ride, probably take an hour? You could do the ride in reverse and it take you all day.

Quality and quantity.

I recorded a short part of a windsurf session last week. Only a couple of blasts back and forward, less than a mile and Strava recorded 71 feet of climbing. I'm pretty sure there were no hills on that lake.

The most accurate way would be to plot it on a paper OS map and count the contours. Digital Elevation Models are notoriously lacking in granularity.

You get a trail that “rolls” up and down a contour, never quite showing as gaining 10m but you can still gain a few dozen metres over the length of the trail as it rises and dips 5-8m at a time.

Yeah, 10 m intervals isn't that granular! Stand next to a 10m diving board and imagine how much elevation change you'd miss over a 50 mile ride if that's your smallest measure.

Less of an issue if you're riding fireroads or constant gradient tracks,but a lot of people aren't.

Yeah, 10 m intervals isn’t that granular! Stand next to a 10m diving board and imagine how much elevation change you’d miss over a 50 mile ride if that’s your smallest measure!

Hate to flatten your ego, but probably roughly as many as the number of times you just step on the first rung (crossing the first contour) then chicken out and go back.

It's a bit like counting grid squares to measure distance, for every square you double back in and walk way more than 1km, there's likely one where you just cross the corner, so over the whole loop it's reasonably accurate.

😀 Honestly, my ego's not involved! I don't use strava much, when I have, I've seen half the climbing on my device as a friend had on his on a 70 miler, full SDW rides 30% under or over what it's supposed to be, so I don't attach much significance to it. And I'm very unlikely to deciding whether or not to do a ride based on its advertised climbing any time soon.

How accurate it is does really depend on the trail you're measuring. If a trail is heading up hill, but constantly undulates up and down 5m at a time, those undulations aren't going to show on the line of the trail on the map, or be captured in 10m contour spacing. If your trail is like this, you're going to be measuring significantly under if you're just counting contours.

The 70 miler above was the Reitti 2000 from Helsinki and back, no big hills, but it's always going up and down, lots of little pinches over granite outcrops, mostly well under 10m. I think I recorded 3,500 feet, my friend had 7,500ft, first hit on google says it's 9,200ft. Not really bothered either way, it was a lovely ride!

Edit: I can see that for a lot of trails, maybe most, the contour counting is going to be fine, I get the analogy with counting grid squares averaging out under/over counting.

Reitti 2000 from Helsinki and back

ZOMG I remember that!

Does it matter anyway?

In the bigger picture, no, of course not. But in MTB terms yes it does a bit. For exple, if I do a 80km ride with 2000m of climbing and get very tired.... I look at the LOTL in the guide at 4100 and think there is no way I can do it without really digging deep. The 4100m In the guide is twice as much as the 2000m measured on my last ride. So I don't bother.

But if the LOTL is actually just 3000m then it's not much different than what I've already done and so I should just go for it.

The above example is pretty unimportant, a better example would be taking a guided group out, or taking the family out. I can't compare what we've done with what we're thinking of doing from the book.

If the book's been put together by one person, you can probably expect that their metres are consistent from route to route, so however you might measure the climbing on your ride, the 4000m route has probably a bit less than double the climbing.

I’d put money on them both being wrong.

Hate to flatten your ego, but probably roughly as many as the number of times you just step on the first rung (crossing the first contour) then chicken out and go back.

But that's not how Strava or OS do it