Is Strava nonsense?

MrJrider

So, over in France at the moment and enjoying some nice climbs, max around 10-12% only Strava has it down as 15-17% what gives??

MrChuck

I don't think it's Strava particularly but GPS in general doesn't seem to be all that accurate when it comes to altitudes and therefore gradients.

ShutUpLegs

Can we now replace 'I blame Wiggle' with 'I blame Strava' :?:

Raffles

ShutUpLegs wrote:

Can we now replace 'I blame Wiggle' with 'I blame Strava' :?:

if you cant say something constructive, please dont post, as your immature trolling in every thread you post in is becoming a serious nuisance.

springtide9

ShutUpLegs wrote:

Can we now replace 'I blame Wiggle' with 'I blame Strava' :?:

+1

It does seem that some people have a very poor understanding of the technology. To repeat for the second time this week, Strava is a computer program, so it's "GIGI"

http://en.wikipedia.org/wiki/Garbage_in,_garbage_out
It is used primarily to call attention to the fact that computers will unquestioningly process the most nonsensical of input data ("garbage in") and produce nonsensical output ("garbage out"). It was most popular in the early days of computing...

So if your GPS records garbage, you will get garbage! I still think there are times that it's still Wiggle's fault mind you.

Note that this forum has a search facility. If you search for 'Strava', you'll get more detailed explanations in other similar posts.
You could also read the Strava FAQ's (as in, read the ... manual) and get very detailed information how Strava processes the data and calculates the various metrics.
e.g. https://strava.zendesk.com/entries/2096 ... r-activity

Mr Dog

Strava is great fun. It motivates me to get out and ride. I would imagine its only as good as the data provided by your gps system. Lets go paint the world orange!

simon_masterson

I think at it's worst, it's a typical internet dick measuring contest. Some of the arguments I see around here are truly pathetic, and 'Strava evidence or it didn't happen' is laughable.

However, whatever it is that causes the inaccuracies (be it hardware, software or between keyboard and chair), the fact is that it still isn't a fair or regulated contest, so I find it hard to take seriously the 'achievements' that some people flaunt so proudly. If some people's GPS units are producing 'accurate' data and some others 'inaccurate', that only makes it worse.

But more to the point, it's a bad influence. Public roads are public unless closed for your purposes, and should be used as such. Particularly on steep, winding hills, I don't want to share the road with anyone who is more interested in beating strangers on the internet that may or may not be in cars or on motorcycles, than the road.

Gizmodo

Raffles wrote:

ShutUpLegs wrote:

Can we now replace 'I blame Wiggle' with 'I blame Strava' :?:

if you cant say something constructive, please dont post, as your immature trolling in every thread you post in is becoming a serious nuisance.

It would seem that some people have had a sense of humour bypass.

The altitude data from GPS recording is very inaccurate. Garmin and others use Barometric alitimeters to try and improve this but even they are prone to changes in the weather. I've gone into a cafe at 1 altitude and come out of it 100 feet lower.

When you upload a route to Strava you have 2 options - to rely on the altitude data recorded by your device (such as a Garmin) or to let Strava use public available sources to get altitude data from. Neither are particularly accurate outside the USA. https://strava.zendesk.com/entries/21068956-checking-your-elevation-correction

As long as you know that and you take elevation data with a pinch of salt, Strava is great fun.

Personally I blame Wiggle.

springtide9

Simon Masterson wrote:

I think at it's worst, it's a typical internet dick measuring contest. Some of the arguments I see around here are truly pathetic, and 'Strava evidence or it didn't happen' is laughable.

However, whatever it is that causes the inaccuracies (be it hardware, software or between keyboard and chair), the fact is that it still isn't a fair or regulated contest, so I find it hard to take seriously the 'achievements' that some people flaunt so proudly. If some people's GPS units are producing 'accurate' data and some others 'inaccurate', that only makes it worse.

But more to the point, it's a bad influence. Public roads are public unless closed for your purposes, and should be used as such. Particularly on steep, winding hills, I don't want to share the road with anyone who is more interested in beating strangers on the internet that may or may not be in cars or on motorcycles, than the road.

For goodness sake, lighten up - I don't think anyone sensible thinks Strava is some kind of race - its a bit of fun and adds a bit of motivation to a ride.
I'm not sure how strong a rider you are, but you must be pretty fast if you consider the majority of the segments unsafe. Most segments are uphill and most have a 60mph speed limit.

adm1

Gizmodo wrote:

When you upload a route to Strava you have 2 options - to rely on the altitude data recorded by your device (such as a Garmin) or to let Strava use public available sources to get altitude data from. Neither are particularly accurate outside the USA. https://strava.zendesk.com/entries/21068956-checking-your-elevation-correction

As long as you know that and you take elevation data with a pinch of salt, Strava is great fun.

I think you may have just answered a question that was niggling me....

I ride with my iPhone (not known for most accurate GPS) and the Wahoo Fitness app recording GPS data. When I get back home, I upload the data to both Strava and MapMyRide - however when I compare the same ride, from exactly the same GPS data on both of the sites, the Strava record almost always shows the total elevation gain to be anywhere from 1.5 to 2x that shown on MapMyRide.

Was just about to post to see if anybody knew why this was - but maybe that's it!Perhaps it is using public altitude data...

Bordersroadie

springtide9 wrote:

Simon Masterson wrote:

I think at it's worst, it's a typical internet dick measuring contest.

blah blah blah. . . .

blah. . . .

For goodness sake, lighten up - I don't think anyone sensible thinks Strava is some kind of race - its a bit of fun and adds a bit of motivation to a ride.
I'm not sure how strong a rider you are, but you must be pretty fast if you consider the majority of the segments unsafe. Most segments are uphill and most have a 60mph speed limit.

I agree. Strava is great fun and since discovering it my mostly solo training has an added dimension. Me and my Edge 500 look forward to rides and the motivational thing is real - I'll gladly go out on days that otherwise I may not bother, because (a) I want to get that KOM or move up from 4th place or whatever and (b) I don't want my buddies to see that I'm slacking.

My commute has been lengthened to take in a couple of nearby segments and over the last year my fitness level has improved moreso than it would have done had I not been "Strava-powered", without a doubt.

Sure, it's not exactly science and my best KOM, a 3 mile local hill, has been retained for ages only because I did it on a rare strong north easterly wind, not because I'm the strongest local guy, but that's part of the fun of it. Who cares about the inaccuracies if it gets people out on their bikes and having a bit of good-spirited competition.

The danger element doesn't apply in the Scottish Borders as we have no traffic up here, hee hee!

slowbike

Yes strava is nonsense - but is fun isn't it.

thegodplato

Strava is simply an added element to see how good/bad we each ride but I always take into account wind, etc. Like whats been said before the GPS can be quite inaccurrate as the other weekend I went out with my club and 3 of us recorded 4500, 4900 and 6500 feet worth of climbing for exactly the same ride. All on Garmins too!

buckmulligan

Strava is a pack of sh!te.

Bordersroadie

BuckMulligan wrote:

Strava is a pack of sh!te.

Care to elaborate?

GuyWhitlock1993

adm1 wrote:

Gizmodo wrote:

When you upload a route to Strava you have 2 options - to rely on the altitude data recorded by your device (such as a Garmin) or to let Strava use public available sources to get altitude data from. Neither are particularly accurate outside the USA. https://strava.zendesk.com/entries/21068956-checking-your-elevation-correction

As long as you know that and you take elevation data with a pinch of salt, Strava is great fun.

I think you may have just answered a question that was niggling me....

I ride with my iPhone (not known for most accurate GPS) and the Wahoo Fitness app recording GPS data. When I get back home, I upload the data to both Strava and MapMyRide - however when I compare the same ride, from exactly the same GPS data on both of the sites, the Strava record almost always shows the total elevation gain to be anywhere from 1.5 to 2x that shown on MapMyRide.

Was just about to post to see if anybody knew why this was - but maybe that's it!Perhaps it is using public altitude data...

If I were you I'd be inclined to lean towards what Strava says. I myself use MapMyRide, and I did a 35 mile route last weekend, MapMyRide said it was 679ft, whereas ridewithgps.com said it was 1690 ft of climbing, so again its a double or more that of MapMyRide...

Can't wait until payday to pick me up the cheapest garmin I can find.

Paul 8v

I think Strava is great, have a pi$$ing contest with your mates to see who's fastest, measure your times on sections to improve your speed et, map your ride and all for free? Mustn't grumble really...

robbo2011

GuyWhitlock1993 wrote:

If I were you I'd be inclined to lean towards what Strava says. I myself use MapMyRide, and I did a 35 mile route last weekend, MapMyRide said it was 679ft, whereas ridewithgps.com said it was 1690 ft of climbing, so again its a double or more that of MapMyRide...

I wouldn't. Around here, mapmyride and my Edge 800 agree reasonably closely. ridewithgps massively overstates the elevation gain, usually by 50-100%.

adm1

robbo2011 wrote:

GuyWhitlock1993 wrote:

If I were you I'd be inclined to lean towards what Strava says. I myself use MapMyRide, and I did a 35 mile route last weekend, MapMyRide said it was 679ft, whereas ridewithgps.com said it was 1690 ft of climbing, so again its a double or more that of MapMyRide...

I wouldn't. Around here, mapmyride and my Edge 800 agree reasonably closely. ridewithgps massively overstates the elevation gain, usually by 50-100%.

Hmmm.....you'd think that the mapping databases these sites all use should have come from OS or similar and be reasonably consistent.

Another strange Strava thing: some of my frequent routes have segments that show as Cat 4 climbs on Strava, but in reality I wouldnt class as a climb at all. One lane that I go down probably only has an elevation of 40 feet or so over a mile, but Strava pegs it as a Cat 4 at 470 feet elevation!

How on earth does this happen? Could it be that the person who created the segment in the first place had a wildly inaccurate GPS and Strava just used data from that? There doesn't appear to be any way to edit it either...presumably that has to be done by the original creator.

MattC59

Raffles wrote:

ShutUpLegs wrote:

Can we now replace 'I blame Wiggle' with 'I blame Strava' :?:

if you cant say something constructive, please dont post, as your immature trolling in every thread you post in is becoming a serious nuisance.

Oooooh....... Get you ! Handbags !
Personally, I blame Wiggle for your humour bypass.
:roll:

springtide9

adm1 wrote:

robbo2011 wrote:

GuyWhitlock1993 wrote:

If I were you I'd be inclined to lean towards what Strava says. I myself use MapMyRide, and I did a 35 mile route last weekend, MapMyRide said it was 679ft, whereas ridewithgps.com said it was 1690 ft of climbing, so again its a double or more that of MapMyRide...

I wouldn't. Around here, mapmyride and my Edge 800 agree reasonably closely. ridewithgps massively overstates the elevation gain, usually by 50-100%.

Hmmm.....you'd think that the mapping databases these sites all use should have come from OS or similar and be reasonably consistent.

Another strange Strava thing: some of my frequent routes have segments that show as Cat 4 climbs on Strava, but in reality I wouldnt class as a climb at all. One lane that I go down probably only has an elevation of 40 feet or so over a mile, but Strava pegs it as a Cat 4 at 470 feet elevation!

How on earth does this happen? Could it be that the person who created the segment in the first place had a wildly inaccurate GPS and Strava just used data from that? There doesn't appear to be any way to edit it either...presumably that has to be done by the original creator.

Have you read the manual? (as in, FAQ) I even posted a link to the right section :roll:

Strava by default uses the data from a GPS barometer if available, as it's usually the most accurate. Barometer's work by air pressure, and air pressure is dependant on weather. GPS height data is usually rubbish.

The data within a segment will be the data from the rider who originally created it. You can 'flag' the segment if you believe the data is inaccurate. So to repeat, with height data, you are really reliant on the quality of the data available and/or created from the ride.

Strava is also a worldwide product, and not just for users in the UK. I would suspect the the OS haven't mapped Russia or parts of Asia, etc, so their 'solution' needs to work worldwide.
The bottom line is that it's not a very easy thing to fix, and the databases all cost money to access, hence why not one of the various online mapping tools/sites has the perfect solution.

adm1

springtide9 wrote:

Have you read the manual? (as in, FAQ) I even posted a link to the right section :roll:

No of course not. I'm male.

springtide9 wrote:

Strava by default uses the data from a GPS barometer if available, as it's usually the most accurate. Barometer's work by air pressure, and air pressure is dependant on weather. GPS height data is rubbish.

The data within a segment will be the data from the rider who originally created it. You can 'flag' the segment if you believe the data is inaccurate. So to repeat, with height data, you are really reliant on the quality of the data available and/or created from the ride.

Righty-ho...thanks for that. I suppose I'd better read the effin' manual then and get on with flagging some of the BS segments....

slowbike

adm1 wrote:

Righty-ho...thanks for that. I suppose I'd better read the effin' manual then and get on with flagging some of the BS segments....

No!

As you said, you're male .... males don't read the effin' manual ... they shout a moan about something that doesn't work properly, then hit it with a hammer and if it still doesn't work they go down the pub to moan some more ...

I don't suggest you try and hit strava with a hammer though - unless you want a new monitor?

racingcondor

springtide9 wrote:

Strava by default uses the data from a GPS barometer if available, as it's usually the most accurate. Barometer's work by air pressure, and air pressure is dependant on weather. GPS height data is usually rubbish.

I had a crit a couple of years back where the weather was changing and as a result the 'altitude' dropped 200ft over the course of an hour. Barometric pressure may be more accurate than GPS altitude but accurate it isn't!

Paul 8v

racingcondor wrote:

springtide9 wrote:

Strava by default uses the data from a GPS barometer if available, as it's usually the most accurate. Barometer's work by air pressure, and air pressure is dependant on weather. GPS height data is usually rubbish.

I had a crit a couple of years back where the weather was changing and as a result the 'altitude' dropped 200ft over the course of an hour. Barometric pressure may be more accurate than GPS altitude but accurate it isn't!

That's some big time subsidence right there :-)

Bordersroadie

Aren't we barking up the wrong tree here?

Surely the point of Strava (the fun bit for those who aren't a bit too much on the serious side) is about who can get up hill x in a faster time than y.

The elevation data is irrelevant to this (I presume it's my GPS map reference point and time of day that govern this), so who cares if it's inaccurate as caring about it somehow misses the fundamental point of Strava?

slowbike

Bordersroadie wrote:

Aren't we barking up the wrong tree here?

Surely the point of Strava (the fun bit for those who aren't a bit too much on the serious side) is about who can get up hill x in a faster time than y.

The elevation data is irrelevant to this (I presume it's my GPS map reference point and time of day that govern this), so who cares if it's inaccurate as caring about it somehow misses the fundamental point of Strava?

To a point - but if the easy climb is categorised as a monster CAT1 climb it can be a bit irritating ...
Got one segment that keeps appearing here - it's an anomaly - although there is a little incline it is in no way a climb yet Strava records it as a CAT4 climb! Just down to annoyance really - so I delete it every time it appears ...

Does Strava automatically create segments if it thinks they are a categorised climb? Cos I sure as hell don't create it, yet I'm the only one recorded as having ridden it and I can delete it ...

Bordersroadie

Slowbike wrote:

Does Strava automatically create segments if it thinks they are a categorised climb? Cos I sure as hell don't create it, yet I'm the only one recorded as having ridden it and I can delete it ...

Funny you should ask that, as I recently noticed a new (1.5mile 5% ave) segment appeared in my regular hilly 32 mile loop (which has some real hills recorded as segments by someone else) and it was attributed to the list of segments created by me. As you said with yours, I certainly didn't create it although amusingly I seem to be the KOM on it so maybe I'll just leave it there for now. . .

adm1

Slowbike wrote:

To a point - but if the easy climb is categorised as a monster CAT1 climb it can be a bit irritating ...
Got one segment that keeps appearing here - it's an anomaly - although there is a little incline it is in no way a climb yet Strava records it as a CAT4 climb! Just down to annoyance really - so I delete it every time it appears ...

Does Strava automatically create segments if it thinks they are a categorised climb? Cos I sure as hell don't create it, yet I'm the only one recorded as having ridden it and I can delete it ...

On the flip side, I just noticed that another of my regular climbs (Barhatch Lane in Cranleigh, Surrey) is categorised by Strava and MapMyRide as a Cat 4 climb. However, I was looking at the Tour of Britain routes and profiles this afternoon, only to see that it is categorised as a Cat 1 there..... :shock:

It's a bastard of a hill no doubt - 2 miles long and topping out at about 19-20% gradient. But Cat 1 ??? Makes me feel a whole lot better about my hill climbing 8)

springtide9

adm1 wrote:

Slowbike wrote:

To a point - but if the easy climb is categorised as a monster CAT1 climb it can be a bit irritating ...
Got one segment that keeps appearing here - it's an anomaly - although there is a little incline it is in no way a climb yet Strava records it as a CAT4 climb! Just down to annoyance really - so I delete it every time it appears ...

Does Strava automatically create segments if it thinks they are a categorised climb? Cos I sure as hell don't create it, yet I'm the only one recorded as having ridden it and I can delete it ...

On the flip side, I just noticed that another of my regular climbs (Barhatch Lane in Cranleigh, Surrey) is categorised by Strava and MapMyRide as a Cat 4 climb. However, I was looking at the Tour of Britain routes and profiles this afternoon, only to see that it is categorised as a Cat 1 there..... :shock:

It's a bastard of a hill no doubt - 2 miles long and topping out at about 19-20% gradient. But Cat 1 ??? Makes me feel a whole lot better about my hill climbing 8)

FYI

Strava Knowledge Base
https://strava.zendesk.com/categories/2 ... ledge-base

https://strava.zendesk.com/entries/2095 ... gment-data
Bad Segment Data
There are a few kinds of bad segment data:

- Bad elevation data (either inflated or inaccurate)
- Unrealistic times for the KOM/QOM and top positions in the Leaderboard
- Unfavorable start and endpoints
- Original segment data has errors or GPS drift
- Segment is Hazardous or has hazardous aspects

Bad elevation data exists if the data the segment was created from (the original ride or run) had bad elevation data. Sometimes, categorized climbs are auto-generated by Strava if the area has no prior Strava activity - this can lead to auto-generated climbs based on bad elevation data that is either unrealistic or incorrect. If you are the creator of this segment, you can delete the segment. If you are not the creator, the best thing to do is to "Hide" the segment on your ride or run. See Hide/Unhide Segments.

False times can be posted on the Segment Leaderboards if GPS errors occur, or people leave their devices on while driving a car. If you see unrealistic times posted for a Segment, where it looks like bad GPS data or data recorded while driving a car, you can flag the offending ride. For instructions on how to flag a ride, see False KOM/QOM.

Unfavorable start or endpoints can greatly affect whether you will match the segment. When the segment was created using data that does not represent the correct route, it is most likely due to GPS drift which occurs when the recorded coordinates drift away from the actual path taken. For tips on how to prevent unfavorable start and endpoints in segments, see Create a Segment. Other unfavorable start or endpoints include intersections and stop signs. Again, if you are the segment creator you can edit the segment to improve the start and endpoints. If you are not, you can create a duplicate segment that includes better data, then "Hide" the former segment.

If the original segment data has errors or GPS drift, it is problematic for matching to other user's rides, and may provide inaccurate data. It is best to try and recreate these segment using better, more accurate data.

If a segment has a hazardous feature, you can flag the segment as hazardous and remove all competition with the Segment Leaderboard. For more information, see Flagging a Segment as Hazardous.

binsted

springtide9 wrote:

Strava by default uses the data from a GPS barometer if available, as it's usually the most accurate. Barometer's work by air pressure, and air pressure is dependant on weather. GPS height data is usually rubbish.
The data within a segment will be the data from the rider who originally created it. You can 'flag' the segment if you believe the data is inaccurate. So to repeat, with height data, you are really reliant on the quality of the data available and/or created from the ride.

.

The overall height difference between start and finish of the sector should be the same whatever the barometric pressure.