cancel
Showing results for 
Search instead for 
Did you mean: 

Strange Activity data ?

Ian
Pico de Orizaba

I have completed quite a lot of bike Segments in my local area (over 5,000) and I'm quite familiar with the Segment leaderboards.  If I attempt a new Segment I often look at the leaderboard and see familiar names of keen local cyclists.  If I see a name I don't know or a big performance gap between the top 2 on the leaderboard I may look at the Activity of the KOM-holder in case of anomalies - if I see obvious use of a vehicle (such an Activity mostly at 10 mph followed by stationary for 5 mins followed by the last mile at 50 mph) then I may comment on the Activity to request the athlete trim their data.  If there's no response I may flag the Activity as "in a vehicle" in an effort to maintain integrity of the leaderboards.     

Today I saw something strange - I tried a few new Segments and saw a new name who had the KOM for about 4 of them.  Their profile was "open" and I could see they had only done 2 Activities (on consecutive days in 2019).  For the relevant Activity with the KOMs the average speed was 19.9 mph and the max speed was 16.3 mph - I've seen average speed higher than max on individual Segments if the athlete didn't complete the whole Segment, but I've never seen this for a whole Activity (anyone have an explanation ?).  The other thing that is very strange is that in an Activity of about 40 minutes the speed was in a very narrow range of 19 to 21 mph - it didn't go out of this range for sharp bends, junctions, changes from smooth tarmac to rough gravel, uphill at 3% or downhill at 3%.  Has anyone else seen anything similar ?  My first thought is that they were on an E-bike limited to 20 mph but I don't want to flag their Activity unless it can be proved beyond doubt.  As they haven't been active on Strava for 3 years I don't expect any response to a message.  One of their "KOMs" is about 20% faster than the 2nd place (of about 300 people).

Any suggestions ?

1 REPLY 1

CreakyCrank
Elbrus

Hi Ian - I hear and feel your pain.  Sometimes, people go out "segment hunting", where they soft pedal then hammer for a segment... but certainly doesn't sound like that here.  Some of this could be explained with segments that may be short, or have poor GPS matching (that is explained in an article I found here), but again, in this case, it sounds like that isn't likely the case either, as it is for a long activity and spans multiple segments.  I'd have also suspected an e-bike at first (or car, motorcycle), but even then, there would likely be speed variations. 

It seems to me that the activity you are looking at may have been 'digitally altered', where somebody takes an activity and runs it through a tool to alter the speed values (if done poorly, then you would see things such as constant speed for up/down, HR that doesn't match the inclines, power data that doesn't match, etc.).  If the activity (activities) you are looking at has HR and/or power data, check to see if those values change with elevation too, that can give additional insights.  In any case, I would likely flag this, as Strava has the tools to investigate for data that seems to not make sense.

Hope this helps... happy riding!