Segment Leaderboards

This is the post for all feedback and feature requests related to Segment Leaderboards.

 

How to format your new idea and/or feedback:

Feedback that is actionable and constructive is most effective. When sharing, please consider the following when framing your thoughts:

  • If you have a new idea, it's helpful for us to understand why that feature is important for you and the value it will bring. For example: I want __. It will allow me to do ___. The reason that is important is because of ___. An example of when I needed that was ___.
  • Specific examples and details that illustrate the problem/confusion you are experiencing. For example: When I do ___, then ___ happens and that is not ok because ___.
  • Screenshots that help visually illustrate the problem. For example: Here is a picture illustrating my point ___ from above. See how ___ is doing ___?
  • An explanation of why this is important. For example: This is important to me because ___.
  • If you propose a solution, an explanation of how that solution would help you and why that is important. For example: I think a ___ would be great because then I could see ___. This is important to me because I have to ___.

 

If someone has already suggested the same idea as you, you can help us quantify the demand for the feature or improvement by upvoting the original comment. To upvote, click the up arrow on the right-hand side of the comment. 

 

Do you need technical assistance or wish to report a bug?

Keep in mind that this space is reserved for feedback. It is not the optimal place to solicit technical assistance or report bugs. To receive assistance with, or to ask a question about our product, please first search our Help Center for the related article. If you do not find what you are looking for, you can submit a ticket with our support team by following the instructions here. To report a bug, first check our Known Issue section. If you do not see the bug you’re experiencing, you can submit this to our team by following the instructions here.

Follow
21

Comments

409 comments
  • It brings you on this page:

    https://support.strava.com/hc/en-us/articles/360061896712-Excluded-Segment-Efforts

     

    Where they says it's not possible to contest. :-/

    Really, I think Strava is really more and more frustrating their user and sooner or later, they'll leave...

    1
    Comment actions Permalink
  • Tapio Kanko - For your 2nd item, e-bikes already do have their own segments and leaderboards and KOM stats.  You just need to set the activity type of your ride to "e-bike" and it is taken care of.  No reason to change your privacy settings.

    1
    Comment actions Permalink
  • I agree with Carlos Vigo and Henry Hagert. Please Strava, consolidate bikes and ebikes segments with 2 different leaderboards.

    1
    Comment actions Permalink
  • Hi Isa Johnson thank you very much for all the suggestions. In regards to the duplicate activities that are being recorded with two or more devices, we do have a system in place that identifies and prevents a duplicate activity from being uploaded, however, this system isn't perfect and there are ways to go about achieving duplicate recordings. I will gladly pass this feedback as well as your other suggestions along to our product team.

    We value your input. Your comments are key in helping us explore ideas that are not part of our product roadmap and can help us prioritize our future work. 

    One thing to remember: just because a particular suggestion is popular or we reply to a specific comment doesn't mean we will take action. Our forums are just one component of our decision-making process. Please understand that we do not respond to all comments, and we may or may not decide to use your feedback. 

    1
    Comment actions Permalink
  • Isa Johnson Jason Lamoreaux

    I like the idea of a result that shows consistent efforts, e.g. from the 3 best efforts. I also think it is very important for motivation purposes to have a leaderboard that is more dynamic, .e.g. over the past 2 years. It is not yet a big problem, but as time goes on the regular top 10 results are going to become increasingly static and meaningless to the typical user.

    However we should always have the simple all time leaderboard as well because we don't want to forget these results and there is a certain elegance to a raw top 10.

    1
    Comment actions Permalink
  • Diego Coenga

    Absolutely, it baffles me that despite Strava insisting that they have algorithms to detect this that it still happens so regularly.

    It is absolutely transparent to the human eye and as a software developer I really don't understand why this is so hard to detect automatically. The speed alone should be obvious enough, the change in average speed is generally not subtle. The location is another dead giveaway, they return to the start point, spend 10 minutes stationary and then shoot off.

    If there is a concern over false positives then just give users the ability to un-flag their rides if the algorithm is a little over zealous.

    1
    Comment actions Permalink
  • Ashley Clarke

    They do already have this. I know someone who got flagged as being on a e-bike when they weren't.

    And yet somehow it seems to make little difference. Perhaps just make this algorithm a bit more aggressive. I think my friend was quite flattered at the flag, I certainly wouldn't take offence if Strava thought I was climbing steep hills at eBike speeds.

    Also take into account athletes previous activities. If they have a history of riding ebikes then it would be more likely, if they have on previous occasions been falsely flagged then give them a bit more leeway in future.

    1
    Comment actions Permalink
  • Diego Coenga unfortunately you are too nice, a lot of them simply do not care, most ignore the flag and it gets marked as resolved. I once pointed out a guy who was doing 125kmph up the motorway after his MTB session. He ignored the flag and then just got abusive when I suggested he crop off the end of his ride. The irony being he could have cropped it quicker than the time spent sending abusive, long winded messages. Instead he went on a rant for two paragraphs about having nothing better to do. The irony was lost on him so I just disengaged. Until Strava either accept the idea of local moderation or auto flagging rides until they are cropped, some people just don't care.

    1
    Comment actions Permalink
  • I agree with Richard Walker a 365 rolling target if preferable. It would mean you can have a yearly champion on 31/12, but not spam January with fake achievements. I like the idea above of a best 3 or 5 rides on a segment over a one year period. I would call this being the 'Mayor' of a segment. Obscure segments with few riders would not have this prize.

    Local Legends is a bit of a bust. I hear no chatter about it, and there is no incentive to target these awards. If you stop riding the same segment for even a little time the award drops. 90 days is no time at all, and it basically encourages you to ride the same route again and again. This is somewhat antithetical to the spirit of Strava, which should be encouraging more riding in new areas. Adventure/territory based games would be much more interesting than Local Legends 'ride along the main road 45 times,' as it is at the moment.

    Also it would be nice if these forum threads were newest comments first to spur debate.

    1
    Comment actions Permalink
  • Richard Walker Erik Bakke Graeme McDonald

    A 365 day rolling period is clearly far superior to a strict yearly period. Going out on January 1st is just so weird atm.

    You might not throw the Guinness book of records away each year, but you tend to do this in sport all the time. You don't say someone isn't the Olympic champion because they aren't the fastest of all time. You don't say someone didn't win the TdF because there were better riders in previous years.

    It is even more important in Strava because such a lot of the historical data is total garbage.

    But like with the current "This Year" stats these are not the primary leaderboard, this is still the all time data.

    1
    Comment actions Permalink
  • ^ or perhaps they think it is not worth cluttering up the app's UI with something that only a few people would want or use ? 

    running with your dog? ...real high priority for it's own activity...NOT !

    1
    Comment actions Permalink
  • Randy Hermann You can add &page=99 or other numbers to the url to jump directly to a page. Example:

    https://www.strava.com/segments/1432325?filter=overall&page=123

    1
    Comment actions Permalink
  • Why is it because it is a sport you do not take part in it is considered trolling? If you do not know what canicross is then just do a search and try and educate yourself.

    As a community there are many groups that take part, but not represented properly, and all we are asking for is some way that Strava recognises an activity type so it doesn't annoy other runners. At parkrun, for example, canicrossers were breaking course records all over the place, and now assisted runs can be marked as so.

    Ben Robinson, a UK canicrosser, holds the world 5k record in a time of 12:24 over cross-country, you can read about his achievements in a wide range of press such as Runners World and even breaking out into non-sporting magazines like GQ.

    The UK governing body, BSSF, has been lobbying Strava for a few years to get cani-sport recognised - yes we do actually have a governing body. UK athletes go to to Europe and the USA to compete in championships.

    So if assisted categories like eBikes can be recognised, then why no assisted sports like canicross getting recognised? It might not be on your priority list because you do not do it, but for a great many others it is.

    1
    Comment actions Permalink
  • Take no offense I was just saying it’s not something I have heard about. I’m glad you are so passionate for it. I also believe yeah.. strava should create that as an activity type. If they have CrossFit, velomobile, and stand up paddler. They should definitely have canicross. Would be a great leaderboard!

    1
    Comment actions Permalink
  • I apologize for flagging same issue that has been raised in previous comments. It saddens to see that issue with unrealistic results in leaderboards isn't detected and sorted out. I keep noticing ebikes or other obviously motorized vehicles recorded as traditional rides. Leaderboard is full of them and no way any traditional cyclists will ever get there as we don't climb with 60km/h speed. It is too much work to flag each activity one by one; and this is example of one segment only. Detecting such abnormal records should be pretty straightforward isn't it? Makes the leaderboard meaningless 

    1
    Comment actions Permalink
  • Jason Lamoreaux Strava have to draw the line somewhere, there are dozens if not hundreds of fringe "sports" such as canicrossing, (or Eric Bakke's ice yachting :-)  ) whose participants probably think their sport is big, but when you think that Strava has well over 75 million accounts and adding many more each month, canicrossers would be about 0.003% according to what Google says is the number of active participants and assuming every one of them signed up for a Strava account (unlikely). That number is insignificant and just not commercially important. Strava is, after all, a commercial operation.

    And then there is the animal rights/PR angle ....canicrossers might say that no dogs are overworked but it would only take one rogue participant to step over the line in pursuit of a KOM and Strava could have bad publicity and even a lawsuit. They dropped the live downhill cycling segments because somebody had a crash chasing a downhill KOM and tried to sue Strava for encouraging dangerous practices. Strava would probably want to stay well away from any possible bad animal welfare publicity.

    1
    Comment actions Permalink
  • Yulia Mironenko There are segments in Strava that, by some reason, get unrealistic times for most of the riders. They are not cheating and their rides must not be flagged. 

    Look at this, for example. If you open the activities of the top10 riders you see that their speeds in other segments are "normal", but in this particular one they get 100 kmph!!

    1
    Comment actions Permalink
  • Auto flagging doesn't always work but I would love to be able to flag a segment-effort as unrealistic or dangerous.

    Like this segment: https://www.strava.com/segments/2070803?filter=overall . This section of road has a bike path next to it and the road itself is prohibited for cyclists. The road is faster than the bike path next to it so most fast efforts are on the road and not on the bikepath.

    Autoflagging doesn't help because there are some efforts which were done during a competition, so they should be allowed. Others should be flagged. The big problem (for strava) is that people go 'KOM-hunting' on dangerous roads which could end up in Strave being blamed for accidents.

    1
    Comment actions Permalink
  • FEATURE REQUEST:

    All the athlete to select the age-range he/she is interested in.

    Examples:  

    • If I'm 64.5, I might want to see how my performance stacks up against those 65-70.
    • If I'm 85, I might not want to compare myself to everybody over 75  (80 means more)

    How about creating filters that the Athlete can control when looking at the leaderboards.

    • Filter by age range (athlete can select min and select max, rather than fixed ranges)
    • Filter by sex
    • Filter by weight (athlete selects min/max, rather than fixed ranges)

     

    Also, I do agree with the previous posters, that suggest it would be nice to see the age of the KOM.

    1
    Comment actions Permalink
  • Although I would like to have a flexible query tool to filter leaderboards, I think just having a “Last Three Years” (or 4 or 5 years) selection option would resolve some of the complaints I have read. It would not affect the “All Time” leaderboard, but would give a better idea of how active athletes are performing than the “Current Year” selection, which is fairly useless until later in a year. If pro-level riders race on a segment, their times would eventually drop off a “Last Three Years” selection, without requiring complicated filtering.

    1
    Comment actions Permalink
  • Please bring back "today" "this week" and "this month" leaderboards. First, they are a great way to see if a trail is open and running as the spring melt hits. Second, especially the "today" leaderboard is great for comparing among a group of friends or to see how you stacked up on that day's conditions. Third, trails change. Thanks.

    1
    Comment actions Permalink
  • Jan Nielsen - maybe I'm lucky, but in my area there are lots of Segments (I've completed more than 5,000) and I think less than 1% have fake results.  There are occasionally examples where some data looks suspicious but these can often be resolved fairly easily - for example, a couple of days ago an inexperienced athlete who previously cycled at 8-10 mph set 10 KOMs in one ride at sustained speeds of 35 mph.  I politely asked if it was a genuine ride and he quickly replied to say he was using Strava to measure his commute by bus (and he deleted the Activity).  I also sometimes see rides at 15 mph that end with a couple of miles at 40 mph - clearly because the athlete forgot to stop Strava at the end of the ride - again these can usually be resolved with a polite message to request the athlete crop the activity.

    1
    Comment actions Permalink
  • Jan Nielsen - If you see results in the leaderboards that you feel are "fake" for whatever reason, just flag them and they will be removed.  I do it all the time when I see obvious e-bikes or just someone who forgot to turn off their GPS at the end of a ride and took some KOMs on the drive home.  It is pretty simple from the desktop platform.

    1
    Comment actions Permalink
  • William O.

    Maybe I'm lucky but in my area I've seen very few obvious problems with Segments down to 300 metres.  Occasionally there are some performances where the average speed is higher than the max but these are a very small minority.  I've never seen an anomaly of more than 3 or 4 seconds so I don't know how someone could gain 9 seconds at both ends of a segment.  There is quite a bit of friendly back-and-forth local competition which is both fun and motivating (surely both are good things ?).  Activities that obviously involve a vehicle are not tolerated by the local community and are quickly flagged - generally the leaderboard integrity seems to be good (due to collective effort).  

    Yesterday I noticed a spurious KOM on a Segment I had created - the person involved had joined the Segment about 10% after the start - in the past I would have edited the Segment to either shorten it (to negate the advantage of joining after the start) or lengthening it slightly to disqualify the spurious leader.  As the Segment is only about 450 metres I can no longer edit it to deal with this distortion, which is a little frustrating.

     

    1
    Comment actions Permalink
  • Leonardo, 
    I think we all agree that there is a problem of athletes mis-categorizing the event.

    We should start by doing both of these:

    1. Default athletes to an activity type called "Activity" (meaning "non-specific") in their Profile.  This generic "Activity" type of activity does not participate in Segments.
    2. Automatically detect anomalies as a function of (average speed, peak speed, acceleration, slope), and perhaps GPS accuracy (intervals, significant speed shifts on same slope, etc).

     

    The first (#1, a user default) is pretty straight forward.  New users get an activity type of "Activity" rather than RUN, WALK, BIKE, E-BIKE, etc.  This will virtually eliminate the person who gets the App, and simply click the "START" button to log an activity, from corrupting the segments used by others.  I like that we currently assume that the activity is the same as the most recent activity.

    One problem I see, is that there is a long alphabetized list of activity types.  Athletes see "BIKE" at the top, and choose it without searching for E_BIKE.  We need to make sure thinks like "E_BIKE and BIKE appear next to each other.  Maybe "BIKE-Electric/Moped" and "BIKE-Traditional" (don't make it so easy for the lazy to choose the wrong category)

     

    With respect to Automated detection of mis-categorized activities (e.g. somebody logging a motorcycle ride).  I think the goal here would be to "auto-flag" and if the athlete doesn't respond or fix it in a week, the activity-type is switched to "Activity" (meaning non-specific), and thus does not participate in any segments.

    NOTE:  Currently, if an event is flagged (sometimes just a data error on that segment), the ENTIRE EVENT is removed from Strava.  This hardship seems a bit extreme, when only one segment has an issue.  I would like to be able to flag just a single segment, in cases where the GPS/data corruption is limited.

    1
    Comment actions Permalink
  • Yesterday I noticed a spurious KOM on a Segment I had created - the person involved had joined the Segment about 10% after the start - in the past I would have edited the Segment to either shorten it (to negate the advantage of joining after the start) or lengthening it slightly to disqualify the spurious leader.  As the Segment is only about 450 metres I can no longer edit it to deal with this distortion, which is a little frustrating.
    -Ian Burton

    I think it is possible to change the start/finish algorithm for segments.  Today, I think we look for the "first" point within a circle of uncertainty.  I think perhaps looking for the two points could help.  Something like the following...

    At the start of a segment-definition, define a point, exactly where the segment is said to start.  Then, determine from segment-definition data, the direction of the route as passes the start point.  From that direction and point, now define an "abeam" line.  Now, instead of a circle, you have a line (like a start/finish line).

    When an athletes encounters the segment on his route, look for the two date-points (the last point before crossing the start, and the first point after crossing the start).  Use this two point to find the TIME and DISTANCE between the two points.  Using this, you can use this average speed to interpolate the precise time the athlete cross the abeam start point.  [Note, if there is GPS jitter which makes multiple data-points back and forth across the line, then use only the last time (data-point) when the athlete was behind the line, and the subsequent data-point (by definition will be after the line) for the TIME/DIST points].   This algorithm could be done applied to existing data, and would likely change a few of the existing segment rankings.

    1
    Comment actions Permalink
  • William O.

    For what it's worth, I just looked at a Segment I created a little while ago - it's 447 metres long, I only attempted it once and it took me 55 seconds (about 18 mph).  I edited it just now to count the number of GPS points in it - there are 36, which suggested my GPS is measuring every 1.5 seconds (I just have a simple and old Android phone for GPS).  Hence with 36 data points the error in time/distance might be about 3% - not ideal for distinguishing very similar performances but it's enough to tell me that I'm some way behind the KOM holder (32 seconds) and hence would still give me a good sense of any improvement I might make in working towards a target. 

    I think having such Segments has some purpose and value in Strava and I disagree with the apparent (unannounced) change to increase the minimum length from 300 metres to 500 metres.

    I therefore implore Strava to revert to the 300 metre limit (and if the lack of precision offends some athletes they can always hide the shorter Segments 😀)   

    1
    Comment actions Permalink
  • "In the case of a GPS anomaly this can be addressed by splitting the ride and cropping out the offending data"
    -Ian

    Splitting the GPS and cropping out is great for those who are technically savvy.  But that's beyond what the average athlete is willing (or capable) of doing.  

    Flagging is great... but we should flag the problem... not the whole Activity (unless the activity-type is wrong).  I've seen GPS errors on Segments occur because you stop near the beginning or end of a segment.  That shouldn't lead to the removal of the whole activity.  If Athletes start losing their Activities and can't see all of their Hours and Miles, they are going abandon Strava to look for another tracker.

    Let's keep the solutions simple for the athletes.

    We can put the complexity into the software.

     

    1
    Comment actions Permalink
  • szh hke - I understand your concerns about over-proliferation of Segments. 

    Maybe I'm lucky, but where I live I don't see it as a major problem.  As an example, I went on a 2-hour bike ride from my house yesterday.  It included 50 Segments, ranging in length from 0.18 miles (current Strava minimum) to 2.15 miles.  I have 7 of the Segments hidden (5 for corrupted leader-boards due to poor Segment creation and 2 for duplication).  This is all fine for me.

    0
    Comment actions Permalink
  • Suzanne Garner - CR = Course Record (fastest overall time posted for that segment).  It is the equivalent of a KOM or QOM (King of the Mountain or Queen of the Mountain) for cycling.  PR = Personal Record.  

    0
    Comment actions Permalink

Please sign in to leave a comment.