Hide Impossible Activity Records From Leader Boards

[TrueDisciple]

Whether on purpose or by accident, many times there are activities on the big leader boards that are physically impossible. 

An example would be running faster than 3 minutes per mile. Or recording a bicycle ride while riding in a car. 

Sometimes this is accidental, because a person forgot to turn off the recording and then drove in their car. Other times, it's on purpose when a cheater tries to game the system. 

This needs to be done for segments, as well as for overall activities. There might be an activity with a normal overall speed, but one or more segments were recorded at an impossible speed. Or there might be an activity with no segments, but it shows up on the leader boards as an impossibly fast activity. 

There are users who periodically report these super-fast activities, but this takes up a lot of time for users. And also for the support staff who review the reported activities. 

This could be automated. If a user uploads an activity that is physically impossible, such as being a lot faster than world record speed for that sport, the app would say something like: "Something is wrong with this activity. It will not be counted for segments or other leader boards."

This would really improve the user experience when looking at leader boards and segments, and reduce user frustration when cheaters attempt to win all the leader boards. 

[reiche]

One of the valuable features of Strava are the segments, but if Stava don't keep the results tidy, then they are worthless. And I guess there are plenty examples of that, given the fault rate of them I am following. 

I have read the above, but I do not agree on the 5% increase limit before auto flagging, it is too vague. How often is the world record beaten? And how often is it beaten by as much as 5%? Looking at history of world record for men running 100m or 400m, there are no examples of that. And then how many of Stravas 95 mill. users will ever even be close to beating a world record legitimately? A very few, maybe 100.

I would like to propose som changes that will improve the validity on segments leaderboards. I'm addressing problems with the leaderboards of short segments like the following of between 130m and 140m (I see both distances in various places) of running:

https://www.strava.com/segments/7928452

Notice that the current record holder finished that in just 3 seconds, which seems very unlikely, as if it was completed using the same pace as for the current world record of 100m running, it would be done 12,45 seconds (if we assume 130m).

Looking at the first 6 positions on the leaderboard, they are all below this. Very unlikely!

And then looking at nr. 7,8,9 (and more) they have some gps coords that have a non continious sample rate.

So all in all, the top-10 on this segment is not worth a penny.

I would like to propose that records on segments goes through some validity checks.

1. If it is within 10% of a world record, flag it automatically and ask the user if they really think it is valid.
2. With a very small number of samples of gps coords on a segment, it can not be considered precise enough for the leaderboard.
3. Also if the speed between 2 consecutive gps samples, exceeds the world record comparable to the total segment length, then it is not valid.

User can have done a great job on their exercise, but without proven facts, it is not eligible to exist in the top-100.

Please note that I'm not even close to the top-10 in the examples above, so I am not trying to bring myself in the the lead. But I would like to compare my own times to legit segment times, without have to go through the list of suspicous top runners. And it is really not fair, to let bad GPS data let them exist in top-10 forever. 

Above were 3 free suggestions. If you need help in defining a complete model, for fair segment activity inclusion, I'm willing to take that up as a contract. 

Kind regards,
Michael

[zecanard]

I’ve just spent 6 days flagging a bunch of runs recorded in vehicles. There was one woman in particular who purposefully starts recording 1–2 miles before starting her run (while still driving), then stops recording long after driving off. On every single run. I reported her for suspicious behavior, but she still had her CRs several days later.

By the way, it took 6 days because I kept getting throttled. If the onus is going to be on us to flag cheaters, you need to remove, or significantly increase, the flagging limit.

In another instance a few years ago, there was one guy who went skating 10+ miles every evening, tagging his blatantly titled “Evening Skates” as Runs. It took more than a week to remove all of his CRs.

We also had a 5k that took place on an airport runway. A lot of Strava users found it hilarious to start recording while sitting in an actual airplane, and take a bunch of sub-1:00/mi CRs.

More recently, I’ve reported a number of cheaters sitting at the top of Club Leaderboards with more hours per week than there is in an entire week; however, last I checked, they topped yet another week with absurd hours.

This issue has been complained about for the better part of a decade on the old forums, and it’s crazy that we still have to spend so much time removing physically impossible activities from the leaderboards. You cannot build a service around a ranking system, have such a large user base, and not take cheating seriously.

[DannyDindia]

Example of rides made in vehicles that Strava couldn't flag automatically. (I marked them eventually myself):

https://www.strava.com/activities/8901156088/overview

https://www.strava.com/activities/8934470264/analysis

https://www.strava.com/activities/8617918015/analysis (average speed 55km/h)

https://www.strava.com/activities/8892451226 clearly forgot to turn off - elapsed time 335h

https://www.strava.com/activities/8975487012/analysis

https://www.strava.com/activities/8583089039/analysis forgot to turn off

https://www.strava.com/activities/8773658062/analysis middle in vehicle - no cadence+power while speed skyrockets

No more flaggings that I can do today....

Shouldn't be too difficult to flag these automatically.

[pajzs89]

Can we see examples when the automatic flaging worked? Whether old or new I really do not see any changes.

[holstein13]

Sometimes I get discouraged when I see what others have achieved on their bikes:

[KSawchuk]

Challenges have the same issue. Someone running 25,000 miles in 47 hours equates to 500mph.