Quiet day in work.....
Wondering if any strava experts can answer this - I've just been doing a comparison on one of my favourite trails of the mine vs the top ten times and it looks like when they posted there times the segment was shorter (on the strava map - the trail hasn't changed for years).
If you do the compare on the map view on all the quick times that have been posted - the segment started 20meters further down the trail so they start 5 - 10 secs up on me before I even start the segment - was wondering if anyone else has noticed this before doesn't make any sense as the distance as far as I can see in the comparison view is the same - but on the map view it quite clearly missing the first 20 metres. Surely the segment on stravas map overlay should always be the same ??
If the segment is edited, the whole leaderboard gets recompiled. I know because I did it recently to a segment where everyone was stopping right on the end of the segment and times were pretty variable. I moved it back 30m or so.
Almost certainly bad GPS data. Many phones, for instance, are useless for keeping a lock while moving.
If you look at the 'ride' that all those fast riders achieved their quick times on - are they full ride or have they been cropped? Did they use an alternate entrance? Are they definitely riding the same segment as you? Did they hit START 20m in to the trail?
Is it a climb? If not, it is more like 5 seconds rather than 10 unless you are slower than me (walking is faster than me).
I reckon it is bad gps data from a non-resident unit I.e. Phone.
However, unsure how Strava is getting them all starting further down the trail than you are.
Do you have a private segment on that stretch?
As above the biggest issue with strava accuracy is that it only takes the data point nearest to the start and finish of each section. If they are 10-20m in to the trail so be it and Strava does not interpolate between data points to provide more accurate times. This is what you are seeing, the segment hasn't changed it's just that the nearest data point was 20m down the trail.
It's also why on very short sections people can do times on Strava which just aren't possible.
If you look at the 'ride' that all those fast riders achieved their quick times on - are they full ride or have they been cropped? Did they use an alternate entrance? Are they definitely riding the same segment as you? Did they hit START 20m in to the trail?
Yes. that's another option.
I had someone take a KOM of mine the other week where they went down the hill, did a u-turn about 20m from the start of a segment and then went back up. Strava decided they were close enough to the start to track the segment. Looking at the compare took that meant they were 6s up on me at time=0. 🙄
Still, I went and smashed it the next day. 😈
There's two parallel start downhill trails where I ride about 20m apart and taking a different route down a hill.
I get a strava time on both of them if I ride either one.
I tend to view leaderboards as vaguely interesting rather than definitive.
I can show you the comparison if you look at the 2 maps the top link is my ride
https://www.strava.com/activities/977055512/segments/23947372291
this link below is the same segment looks unedited - but the leader clearly starts 20 - 30m into the same trail missing out the dog leg at the top that you see in my run - (as do most of the top ten on the leaderboard)
https://www.strava.com/activities/125633881#2822595508
if its true about the start point - this doesn't explain why the map outline is different - surely this should always be the same ?
I see this kind of shit all the time. Usually an iphone :P...just move on and forget about it, its only a virtual bit of information in cyberland.
I learned that Strava was crap last year when I was working on my technique on the High Voltage section of Follow the Dog, most of the main trails at cannock have about 4 segments each and on one occasion I was a fair bit down on the Down to Decision Point segment but got a PR on High Voltage. High voltage finishes after DTDP, but the segment time was shorter.
Don't take it too seriously now
There's two parallel start downhill trails where I ride about 20m apart and taking a different route down a hill.
There is a segment near me named for the steep climb up some steps.
When i first looked at it I did wonder exactly how people were managing to ride it so quickly. I have no illusions about my skill/fitness or lack thereof but the difference in times was rather surprising.
Looking at the top times though all of them were on the road about 10m or so to the left with a rather less steep incline.
I was worried this thread would get the don't take it too seriously type of reply, trust me I don't at all. Im not a strava obsessive compared to some riders I know, but I thought (I was interested to see on one of my regular trails where faster riders are making up time and maybe session those sections of trail and try and work on techniques etc)
But what I have highlighted above means there's a pretty big flaw in the system - makes the strava comparison tool useless!
It helps you see if they're faster than you because they're faster, or if they've just got a lucky blip.
Everyone about me on the Strava tables around cannock are lucky, as I am a riding god.
I'm under no illusions ...I know who the "genuinely" fast riders are locally and who I have no hope of catching... and the ones who have put that much time into me on that segment aren't that fast - or are on strava EPO 😉 ! or the trail has changed - or strava is broken - as highlighted above!
Lots of info here on why it may not match some rides to a segment or poorly match if it does. [url= https://support.strava.com/hc/en-us/articles/216918227-Optimizing-Segment-Creation-how-to-create-good-segments ]https://support.strava.com/hc/en-us/articles/216918227-Optimizing-Segment-Creation-how-to-create-good-segments[/url]
If you genuinely think they didn't ride the whole segment you can challenge it with Strava and have them removed.
also:
[i]Occasionally a ride or run will be falsely matched to a segment or receive an inaccurate segment time. Sometimes this is due to the fact that our segment matching algorithm must be flexible enough to account for GPS drift and can be a little loose with the matching. Poor GPS data can also contribute to inaccurate segment times. [/i]
@waswas - I think your strava article pretty much confirms it one of the paragraphs read
"Segment matching uses the GPS points in the data closest to the start and endpoints of the segment, and as this can vary with each activity, timing on a segment can vary slightly because of this. At the present time, we don't interpolate or extrapolate GPS data to normalize for the exact start and end positions of the segment."
although in this case the start points that some of the top ten are picking up are way down the trail after some tight flat corners..... that's pretty poor in my opinion if the activity doesn't match on the same start point then it shouldn't register - that would make it more consistent.
Look at the KOM's earlier effort:
https://www.strava.com/activities/125633881/segments/2822595456
It's pretty much the same track as yours, so the segment is the same.
It's GPS / Strava randomness on the KOM run - he got lucky 🙂
If you're in woodland, the GPS tracks can be quite a way off your true position, too, due to weak/shadowed satellite signals.