cancel
Showing results for 
Search instead for 
Did you mean: 
AdamVD
Pico de Orizaba
Status: Gathering Kudos

Currently the shortest segment that can be made is 500m. This is fine, but for some short climbs, this is too short. Such climbs could be 300m but take more than 30s for the KOM time. Instead of having a blanket rule that any segment distance less than 500m "isn't long enough for accurate segment times", use the gradient to estimate what the KOM time will be so that the segment length can be reduced but still be accurate because GPS data will be more accurate on distance when riders are going slower.

For example:

gradient <= 0%, minimum 500m

0% < gradient < 3%, minimum 400m

3% < gradient < 6%, minimum 300m

6% < gradient < 10%, minimum 250m

10%+, minimum 200m

 

8 Comments
zecanard
Kilimanjaro

A higher gradient means a shorter horizontal distance for an equivalent distance traveled, meaning segment matches that are even more sensitive to GPS inaccuracies. Meaning leaderboard times that are even more dependent on extremely precise matching of activities to segments.

Strava is notoriously imprecise when it comes to matching segments, sometimes matching activities that stop several hundred feet from the end of a segment. On a steep incline, which, as you note, takes longer to climb than a flatter surface, shaving off a hundred feet would make an even bigger difference in leaderboard times.

AdamVD
Pico de Orizaba

Thanks @zecanard , I think your right about how Strava can be inaccurate for bird's eye view of a slope. I don't think the gradients are significant enough to make much difference; plus I'm sure the "distance" is measured from a bird's eye view anyway so climbs are actually slightly longer than how long Strava says they are. The main point is, all cyclists will go uphill slower so, for uphill segments there will always be, on average, less distance between GPS points, making uphill GPS more accurate when it comes to measuring speed. For steeper climbs over 5%, the difference between 1kph is significant. Strava used to have a minimum of 350m and I want to know why they increased it to 500m

Ian
Elbrus

I'd still like to hear some explanation from Strava about the reason for increasing the minimum Segment length from 300m to 500m.  I'm not aware of any users asking for the change and several people have asked for the change to be reversed, without success.  It has been suggested that having more shorter Segments would increase the cost of maintaining data.  Could someone quantify the cost impact ?  (what percentage subscription increase would be needed to cover it ?) 

Status changed to: Gathering Kudos
Soren
Denali

Thanks for submitting your idea. It has been reviewed by our moderation team and is now open to voting.

MakeStravaGreat
Pico de Orizaba

I think it would be alot easier dropping the minimum segment length to 100m. Theres soo many trails i ride that are even less than 300m. theres old segments that length that work fine. i dont see the issue

AdamVD
Pico de Orizaba

Looking at this, it seems the segment length could be reduced for all segments if Strava replaced old segments (recorded with bad GPS data) with new segments using good GPS data. This will greatly improve the leaderboard accuracy. Look at this segment:

 - https://www.strava.com/segments/4526578

The segment is recorded with inaccurate GPS data for both elevation, distance and route. If this were replaced with any recent GPS data, the segment leaderboard would be more accurate. For example, I am 2nd right now - but, looking at the analysis, I am ahead on the road at every point and moving faster; however, I haven't got the KOM because of how speed is mapped to the segment GPS points. The algorithm needs to be improved.

StilPhone
Mt. Kenya

I agree with shortening the mimnum segments length - I accept it may not be accurate, but let me make that decision for myself. As a subscriber I feel we should be allowed to have a segment of 100m - we always used to. 

How hard would it be to attach a warning to segment creation that segments under x00m may be inaccurate? Its not like anyone is going to sue Strava for their segment being inaccurate!!!

helg
Shkhara

I believe the main reason of minimal segment length is GPS accuracy. Long riders are not happy that the whole ride is flagged because of GPS glitch on a short segment in the middle. Once shorter segments are allowed, could we have a personal setting to opt-out of short segment leaderboards - to prevent the flagging?