How true is strava?

Marcryan206

Hi all

I know this properly has been talked about before! I did 22miles today and on my app it said fastest speed was 37.5mph web I look at the graph for speed it doesn't show when I did it? As I only went down one good hill which I got stuck behind a lorry goin down can't see how I made that speed?

Have look see if I'm just being dumb?

http://app.strava.com/activities/35311653

Many thanks guys

Scotty-Gee

Were you running it as an app on your phone?

If so its only as good at the GPS is getting.

An an offshore engineer who moved oil rigs, I always chuckle at people telling me how good there sat nav is- we used to pay £300 a day to get 1m accurate DGPS signals.

I suspect your speed was a GPS jump and not valid. There is also a strange linear decrease in speed- either you braked very smoothly or another strange gremlin.

Marcryan206

Yes was running on iPhone. Guess it must be those GPS Gremlins did think I'd been that fast!!. And very smooth breaking :P

declan1

Strava doesn't show you the top speed in the fully zoomed-out view. Go to performance, click the speed checkbox, and select around an inch of that graph around the highest peak. That will take you to a zoomed-in view, where you can see the 37.5MPH bit

charlie_potatoes

You hit that speed at the bottom of the hill on Titchfield Hill (WestBound) segment.
Click on that segment and check the graph as it's shown there

cougie

I'd think that was a blip. It's ok for average speeds but not accurate for max speed.

Marcryan206

Cheers guys! Guessing if that lorry didn't slow me down might have got a little bit faster! Still time to improve!

lotus49

I went out for a run yesterday and, according to Strava, I briefly hit 71kph. The average speed was correct but I suspect that I could only hit that top speed if I rode my bike off a cliff.

roger_merriman

lotus49 wrote:

I went out for a run yesterday and, according to Strava, I briefly hit 71kph. The average speed was correct but I suspect that I could only hit that top speed if I rode my bike off a cliff.

Big hill? I've got close to that down a long fairly steep hill but that's the exception the proves the rule, GPS's even the pricey top of the range ones, do spit out some just plain wrong figures, underpasses in built up areas are good for some highly improbable speeds!

TOM14S

I don't see speed spikes but I do have the mag speed pick up with my Garmin. Never sure how it calculates the speed though between what the mag speed sensor says vs the GPS, I would assume it prioritizes the sensor. Which then leads me on to topic wondering about stravas accuracy, if I put a much larger wheel circumference into the settings so the Garmin thinks I'm going faster than I'm actually going, how strava would cope with that. Would it take my adjusted av speed or time between two GPS points? I don't know how the Garmin would handle the measured distance speed vs. GPS speed distance discrepancy and plot speed distance altitude and route if it all doesn't much up. I might experiment at some point.

smoggysteve

Mentioned this on another thread. A smartphone cannot lock onto as many satellites for a true position as say a car sat nav. It relies on using mobile phone masts to triangulate itself. If you go into a bad comms area and the signal drops, if will effect your position. Mobile phones although capable of it were not designed to be used for tracking distance and speed accurately. The point was to give you a static position mainly for things like geo tagging

NITR8s

Ive had a top speed on Strava once of 52 mph I think. I do remember pushing it down a long hill and keeping up with traffic, but I am fairly certain if I went that fast I would have peed my pants.

As for 38mph, I nearly aways hit a top speed of about 30+mph without trying just cycling into work. I would proberbly say its round about right.

djm501

If you want good speeds to register use mapmyride - I once got over 15,000 mph registered on that. That was a good downhill! ;-)

paul_mck

lotus49 wrote:

I went out for a run yesterday and, according to Strava, I briefly hit 71kph. The average speed was correct but I suspect that I could only hit that top speed if I rode my bike off a cliff.

have hit 76kph before, and nealy peed my pants lol

GPS can be a bit buggy even on my Garmin (cycling through trees etc). Phones are much worse.

Bustacapp

SmoggySteve wrote:

Mentioned this on another thread. A smartphone cannot lock onto as many satellites for a true position as say a car sat nav. It relies on using mobile phone masts to triangulate itself. If you go into a bad comms area and the signal drops, if will effect your position. Mobile phones although capable of it were not designed to be used for tracking distance and speed accurately. The point was to give you a static position mainly for things like geo tagging

I have a GPS app on my phone that uses no data connection whatsoever. I check my speed on my bike computer at certain points and Strava is relatively accurate with that. The speeds are out by say 0.5mph but I think that's because the bike computer is not calibrated correctly. My bike computer always says I've done a bit more mileage than strava does.

smoggysteve

Bustacapp wrote:

SmoggySteve wrote:

Mentioned this on another thread. A smartphone cannot lock onto as many satellites for a true position as say a car sat nav. It relies on using mobile phone masts to triangulate itself. If you go into a bad comms area and the signal drops, if will effect your position. Mobile phones although capable of it were not designed to be used for tracking distance and speed accurately. The point was to give you a static position mainly for things like geo tagging

I have a GPS app on my phone that uses no data connection whatsoever. I check my speed on my bike computer at certain points and Strava is relatively accurate with that. The speeds are out by say 0.5mph but I think that's because the bike computer is not calibrated correctly. My bike computer always says I've done a bit more mileage than strava does.

I didn't say it uses phone signals (Nothing to do with data) exclusively, I said it uses them also. It cannot lock onto as many satellites as you can with a car satnav or a Garmin etc. Also, your geographical position will effect your signal as will altitude. Sometimes the beacon can be switched off, so as your device is trying to ping a sat for a time stamp, it gets sent spoof messages that throw it off and give you dodgy information. (I've worked with Sat comms in the army for 12 years, trust me on that)

Bustacapp

SmoggySteve wrote:

I didn't say it uses phone signals (Nothing to do with data) exclusively, I said it uses them also. It cannot lock onto as many satellites as you can with a car satnav or a Garmin etc. Also, your geographical position will effect your signal as will altitude. Sometimes the beacon can be switched off, so as your device is trying to ping a sat for a time stamp, it gets sent spoof messages that throw it off and give you dodgy information. (I've worked with Sat comms in the army for 12 years, trust me on that)

All I know is my app says it reads phones location once every second or so and does not use data connection.

I do know that using phone masts on other apps I've had in the past to pinpoint my location has been way off, like a mile or more from where I actually am.

joe.90

i know people with garmins and iphones etc, and to be far they always give very similar readings +/- 0.1 mph on averages

schweiz

The iPhone will use A-GPS which is Assisted GPS, the assistance coming from the location provided by the GSM network to locate the phone and obtain a faster postion fix. My phone gets a fix far quicker than my Edge705 (30 seconds compared with 1-2 minutes dependent on when I last used the 705 and if I'm starting from the same location that I stopped previously.

The iPhone will have the capability to track 12 satellites* which is as many as a Garmin Edge 705 can track. This is because the GPS chip inside will be on off the shelf chip than anyone can buy and they're mostly 12 channel. Saying that, the Garmin only ever shows 7-8 satellites and you can pretty much forget 2 or 3 of those as they are on the horizon and not improving accuracy to any great degree.

The weather, buildings and terrain can all mask GPS signals so the quality of signal in terms of both the number of satellites being used and signal attenuation can vary dramatically.

The difference between a phone and a 'proper' GPS will be antenna positioning and size. Look at the antenna area on the top of a 705 and then think where an Apple engineer has tucked an antenna away to keep the size and shape of an iPhone

As stated previously, if you want a really accurate GPS signal then Differential GPS (DGPS) is the answer but it's not cheap!

*unless the bods at Apple have restricted the I/O to save power or some such.

Critch

I just stood in my conservatory (opaque, polycarbonate roof) with heavy cloud cover above, both of which I am told affects GPS signal. I do not have network assistance for GPS enabled on my HTC smartphone and I was able to ping 11 satellites steady with an accuracy of 4 yards. There was 1 space for 1 more satellite to be connected but as I said I am inside my house with heavy cloud cover....

How many satellites do you need for reasonable accuracy? Not trying to be intentionally nitpicky but I also use my phone as my main SatNav (Sygic) and its brilliant, never once let me down and records my speed accurately. When I get to the door of the premises I am going to it tells me I have arrived and it tracks accurately. I think some may be under-estimating the ability of modern smartphones but I will concede to their expertise.

To answer the OP, I would suggest that most of the data is accurate for the purpose it was designed [at the risk of contradicting myself] until it glitches and that can depend on many things but nothing will be perfect all of the time with so many variables. Personally I use Endomondo and a couple of weeks ago I hit dead on 37mph on my Cateye computer whilst out riding. When I got home and checked Endomondo it said max speed ....... 37mph. Good enough for me. Its glitched once on me in all the time I have used it and said I did nearly 3000mph. Just have to use ones noddle, but for the most part they a worthwhile tool to use and accurate in the main.

Downhill and perhaps also drafting a lorry a little 37mph is not unrealistic at all but you should be able to gauge that speed and feel that its right.

Marcryan206

I have heard of phones being a little unreliable but after reading what a few of you have said I do believe it to be rather better than I thought. I do get what ever one is saying regards some times you get the odd crazy data come through. Of course if you want better data and are working professional on the bike then you would spend the money on better GPS but for what we are using it for I think it is rather good enough. Of course Strava has its enemies but for me I like to see how I'm getting on in progress and so on.

schweiz

GPS needs 3 satellites to provide a 2D fix and at least 4 satellites for 3D (i.e. height). If you've got 11 satellites and many of those are within 45° from vertical then you'll get the resolution that you describe. 4-5 yards/metres is about normal. However, data glitches can give strange results. This lunch I was nearing the top of a climb and as I stood on the pedals my body blocked some of the satellite signals and I got multiple auto start/stops on my 705 and the reported speed was anything between 0 and 30-odd km/h. I live in a mountainous area so my horizon is quite limited hence only receiving 7-8 satellites. If I were to be in downtown New York, I'd probably struggle to get even 3. I seem to remember a Top Gear episode when they had to drive through Manhatten and the in-car navigation systems packed up due to lack of satellite cover.

paul_mck

how long did it take to fix position in your house?

doing it quick enough, and while moving, is the challenge.