cancel
Showing results for 
Search instead for 
Did you mean: 

Minimum Segment length ?

Ian
Elbrus

I noticed that someone recently asked that the minimum length of mountain bike Segments be decreased and the post was quickly archived by an admin on the basis that reverting Strava changes could not be discussed.

I have some sympathy for the poster - I was also somewhat frustrated by the recent increase in minimum length of bike Segments from 300m to 500m.  In my area there has been some good competition over Segments about 400 to 450m in length but such Segments can no longer be created or edited (the change happened as I was working on a new Segment of about 450m - it now finishes closer to a T junction than I would like but editing is no longer possible).

I don't remember seeing any requests from Strava users for such a change in the previous Strava Support forums so I wonder if someone at Strava could give some background to their reasons for making the change and their methodologies for determining minimum Segment lengths for each Activity type ?  Clearly measurement accuracy is a factor on shorter Segments, but even on a 300m bike Segment the accuracy should be within a couple of % which seems OK.

42 REPLIES 42

Frosty
Mt. Kenya

Strava really need to consider that a number of MTB tracks in forest zones are shorter and even under the old 300m limit. We have several great tracks that are technical DH or climbs all under 300m which we could not make segments for previously. Sure make road segments longer but MTB need shorter segment lengths! Come on @Strava listen to your users and get this sorted. one rule for all is outdated thinking! Your app is fast being left behind and at risk of becoming irrelevant.

Silentvoyager
Kilimanjaro

I think the main problem here is that Strava is trying to make the segment matching to still work for road cycling activities that use Garmin smart recording. With the smart recording the data points can be a few seconds apart. When cycling at speed, 10-15 meters/sec is easily possible. With the smart recording that might lead to datapoints being 50-100 meters apart.

Because of the above consideration the radius for matching the segment start and finish is quite large. I remember reading somewhere that it is around 50-75 meters. Unfortunately it seems that the same radius applies to all segments regardless of the activity type. I've seen that in some running segments where someone does hill repeats and turns around at the top of the hill too early, and they still get a top position on the leaderboard even though they are clearly too far from the segment finish. When running uphill 50 meters is a long distance.

Now with the radius that large, if the segment is too short, you may see ridiculously wrong matchings of segments. For example, on a segment that is 300 meters long, someone, if they are extremely lucky or if they cheat and know what they are doing, may end up covering only half of the distance and still match the segment. And that has nothing to do with the GPS accuracy. That could be achieved with a perfectly accurate GPS just by entering and exiting the segment at the right points, for example, by turning around just on the edge of the matching radius.

So instead of fixing the problem above Strava increases the minimal segment distance limit. That still allows someone to match a shorter distance and get an unfairly shorter time, but now that error in the segment matching won't be as egregious. For example, they may end up matching the distance that is just 20% shorter instead of 50% shorter.

One solution that Strava should strongly consider is using interpolation. That should allow to narrow down the matching radius from 50-75 meters to something more reasonable like 10-20 meters. Even if someone's device records points every few seconds, with the interpolation that could be lowered down to 1 second and match positions that are much closer to the segment's true start and finish. That would allow to shorten the minimum segment length while still matching more accurate segment times.

BB
Mt. Kenya

Super disappointed with Strava on this one. We have lots of very steep gravel climbs in Western PA that are less than 500 meters long. Too many were already eliminated for segment creation by the last arbitrary and unrequested longer segment requirement. I also don't believe the real reason for this change is that shorter segments are too hard to time accurately. How does determining when a start and finish line are crossed get easier if the start and finish lines are closer together?

I notice that in my local area a significant proportion of the new Segments created in the last few months have a length of 0.31 miles, the current minimum allowed by Strava. This suggests that Segment creators have been constrained by the somewhat arbitrary limitations and would probably have preferred to have their Segments a little shorter. I really struggle to see the downside of making the minimum, say, 400m instead of 500m.

gnordli
Mt. Kenya

There needs to be some consideration for the time it takes to complete the segment instead of just length.  A road bike may be able to do a 500m segment in 40 seconds or so.   Whereas a 250m mountain biking segment could easily take over a 60 seconds to complete, depending on terrain.  

My suggestion is people can "submit" a segment for approval, and if let's say 50 different riders have ridden it and the top 10% avg time is greater than 45 seconds, then it should be legit.  

I just threw some numbers out there.  Strava has access to all of the data, so you can compute would looks like a legit segment for mountain bikers, before it gets approved.

There are too many mountain bike trails less than 500m -- Strava needs to figure this out. 

 

Great point. Mayb only allow ppl with a decent GPS device to create segments too. 

eschlange
Shkhara

The new minimum segment length of 500 is especially silly when it comes to virtual segments (on platforms like Zwift, for example). On those segments, GPS is recorded consistently - Zwift records every 1s, for example. So shorter segments should be allowed, since accuracy is ensured across all users.

I wrote a thorough post about this on Zwift Insider which explains several reasons why Strava's current minimum length requirement is silly in the context of Zwift. I do hope Strava will reconsider.

https://zwiftinsider.com/strava-minimum-segment-length/

T3mppu
Mt. Kenya

I have contacted strava support on this and they told to bring it here so community would get more visibility on this.

I too had estimated the limits to be rised from 300m into 500m and support confirmed:
We have a limit in place to make sure that newly created segments would result in more accurate leaderboards and reduce segment noise overall. There are minimum distance requirements for ride segments (500m) as well as other sport types (250m).

So 500m for all cycling, 250m for running.

 

As I see, the shorter mtb segments would usually be in local forrest, where it is really difficult to get the 500m or it would be making dangerous segments if you need to cross a road or walkway to get to the 500m.

Also the speeds in such places are a lot less than on the road, so there would be more gps points recorded than on road segments that also are 500m limit.

So it would be really nice to get back the 300m (or even 250m) limit.

 

For the segment noise part, it would be also nice to be able to flag runs that clearly are not following the segment.

What a joke. This is all about less maintenance, more profit. 

Strava if you add some sort of feature to let us delete poorly created segments then we can do your dirty work for you.

Ian
Elbrus

I've been prompted by Strava to identify 1 of the first 3 replies as an "accepted solution", however, I'd prefer to wait until my question has been answered and I'm still hoping to get some feedback from Strava about the reasons for their changes (which don't seem to have been led by customer requests)