Just how accurate are GPS tracking devices?

[Aravis]

There's been endless discussion on this site and elsewhere on why the recording of elevation data varies so much. An area where there is work to be done.

Recording distance is the bread and butter task for cycle GPS, and it would be concerning if different units produced markedly different results. Is a difference of 0.2 miles in 50 tolerable? Perhaps. It's certainly proportionately more than I've seen since I've been running an Edge 200 and a 520 in parallel. The biggest difference was in my most recent ride - see below.

However, assuming the actual figure (if such a thing exists) for the OP's ride was indeed just over 50 miles, it hard to believe this would be measured more accurately by a wheel revolution counter, where it's impossible to be certain that the calibration is 100% accurate.

The data for my most recent ride is as follows:

Garmin Edge 200: 109.87 miles
Garmin Edge 520: 109.69 miles
BikeHike: 109.79 miles
RwGPS: 109.9 miles

That is formidable consistency. If a wheel counter produced a result of 110.5 miles, would you assume that the answers above were all wrong?

 

[snorri]

Confucius had some words of wisdom regarding clockwork watches..........Man who wears two watches never knows the right time.
Perhaps this could be updated to man with two satellite positioning systems.

 

[Aravis]

Confucius had some words of wisdom regarding clockwork watches..........Man who wears two watches never knows the right time.
Perhaps this could be updated to man with two satellite positioning systems.

Doesn't sound very wise to me. I thought the wisdom of keeping backups was pretty well understood nowadays.

Man who wears one watch doesn't know whether he knows the right time or not. Man who wears two has a much better chance.

 

[Shut Up Legs]

There will always be inaccuracies, enter an area of poor reception, or have a better GPSmodule in one unit and this will be evident.

The trick is really to use these units as a comparison

Do the same ride twice, and if the speed is faster, then you will have been faster, if your heart rate is higher than it will have been higher

Simply because all the inaccuracies and errors are consistent for that unit

Except when the weather affects the Garmin's barometric altimeter, which occurs on occasion. Although, I've found that the Garmin 705 and 800 (I've never owned any other model) measure elevation changes fairly well most of the time. It's only on days with sudden and major pressure changes that their altimeters can be unreliable.

 

[the snail]

Doesn't sound very wise to me. I thought the wisdom of keeping backups was pretty well understood nowadays.

Man who wears one watch doesn't know whether he knows the right time or not. Man who wears two has a much better chance.

A three watch system with a weighted voting algorithm would be better though.

 

[Markymark]

A three watch system with a weighted voting algorithm would be better though.

You could always go to Norfolk where you're guaranteed of knowing the exact time no matter what you're doing.....1957.

 

[postman]

I don't need to be told i have just ridden crap.i can feel it.