Series 4 GPS off at runs?

[xDKP]

I’ve been running with my Apple watches for years and love it.

I’ve noticed that my S4 has been measuring too many kilometres compared to the measured route.

Ran a local half marathon today and around the 20 km mark my watch was 0,35 km ahead.

It ends up making a pretty big difference - my watch said I ended at 5:27 / km and the official around 5:35/ km.

My goal was to hover around 5:30 as that’s the marathon pace I’m aiming at for the upcoming race so no major catastrophe. But if the watch is responsible for me going over the calculated 4 hours because of bad gps measuring I’m going to be quite disappointed.

 

[Significant1]

Workoutdoors has a setting to disable intelligent distances. It seems to use the gps distance, when this is disabled. What intelligent distance is, I am not sure, but I suspect that it uses the motionssensors in some way. Intelligent distance is default, if I remember correctly, and it might be what Apple's workout app uses.

 

[ftaok]

I’ve been running with my Apple watches for years and love it.

I’ve noticed that my S4 has been measuring too many kilometres compared to the measured route.

Ran a local half marathon today and around the 20 km mark my watch was 0,35 km ahead.

It ends up making a pretty big difference - my watch said I ended at 5:27 / km and the official around 5:35/ km.

My goal was to hover around 5:30 as that’s the marathon pace I’m aiming at for the upcoming race so no major catastrophe. But if the watch is responsible for me going over the calculated 4 hours because of bad gps measuring I’m going to be quite disappointed.

0.35 km over 20 km is less than a 2% difference. I think that's totally within the margin of error. Are your routes with the S4 noticeably more inaccurate compared to your previous AW's?

 

[PBz]

I’ve been running with my Apple watches for years and love it.

I’ve noticed that my S4 has been measuring too many kilometres compared to the measured route.

Ran a local half marathon today and around the 20 km mark my watch was 0,35 km ahead.

It ends up making a pretty big difference - my watch said I ended at 5:27 / km and the official around 5:35/ km.

My goal was to hover around 5:30 as that’s the marathon pace I’m aiming at for the upcoming race so no major catastrophe. But if the watch is responsible for me going over the calculated 4 hours because of bad gps measuring I’m going to be quite disappointed.

If you are ‘racing’ an event, I would not count on any GPS to be perfect. Even a Garmin with GLONASS and Galileo (AW has these as well) can be off by this margin. Have a running timer and compare it to the km markers. Many things can affect the accuracy of GPS, especially in an urban environment. That variance seems within margin of error.

One thing that could also be an issue is if you had to run ‘around’ many groups of people. I have only run one half marathon but I remember noting that if I ever run another one I will be mentally prepared for what felt like running 6 miles in the first 5 miles of the event.

GPS variance and ‘side’ running seems more than reasonable for this variance.

 

[xDKP]

If you are ‘racing’ an event, I would not count on any GPS to be perfect. Even a Garmin with GLONASS and Galileo (AW has these as well) can be off by this margin. Have a running timer and compare it to the km markers. Many things can affect the accuracy of GPS, especially in an urban environment. That variance seems within margin of error.

One thing that could also be an issue is if you had to run ‘around’ many groups of people. I have only run one half marathon but I remember noting that if I ever run another one I will be mentally prepared for what felt like running 6 miles in the first 5 miles of the event.

GPS variance and ‘side’ running seems more than reasonable for this variance.

Thanks for the response - I remember it to be a bit more accurate on some of the earlier models - but might be wrong.

Yes it's only 2% and I kind of think it's okay everything consideret if it's the same level as the competition but I really need to be on point on racing day to make sure it's not causing me to go over/under the target pace...

 

[Monkswhiskers]

2% is acceptable I would say. You are never going to run a perfect line as measured by the course officials, all those taking corners a bit wider/going around people over that distance add up. Even the Garmin top range watches aren't 100% accurate (Some Fenix owners I know have woeful gps accuracy!).
If you want correct live pacing then the Stryd footpod is the way to go, way more accurate than GPS.

 

[iPhysicist]

2% is great. As far as I know GPS does not guarantee a certain accuracy. Usually the accuracy is within 30feet of your actual position. You have to keep in mind that the overall accuracy is also determined by the number of satellites your watch gets a lock on. You can say that the GPS lock does usualy have an in-accuracy of 10feet radius depending on your surrounding (from my experience). If you are running in the woods - make it 30feet or worse. Same goes for cityruns with "urban canyons". There are a lot of factors that the watch or any other GPS tracker does have no impact on by itself. Its kind of "**** in - **** out".

Also the official timekeeping is always for ideal line so unless you run it you will get different results.

 

[xDKP]

Thanks gents - 2% seems to be well within acceptable distance ...

Have just tried to reset the watch calibration so will post if it makes any difference.

 

[Monkswhiskers]

Thanks gents - 2% seems to be well within acceptable distance ...

Have just tried to reset the watch calibration so will post if it makes any difference.

If you have recalibrated this might result in worse results initially. When I first bought mine it took a couple of weeks to settle down, was 20% out at first and I considered taking it back. Now it is very consistent with my usual route. I think the accelerometer learns your stride pattern and uses in conjunction when gps is weak.
Don't forget the main gripe with the techies is the hand off with the iPhone if it is close by at the start of a run, there is a period where it tries to find the phone for a couple of minutes so uses the accelerometer to calculate.

 

[xDKP]

If you have recalibrated this might result in worse results initially. When I first bought mine it took a couple of weeks to settle down, was 20% out at first and I considered taking it back. Now it is very consistent with my usual route. I think the accelerometer learns your stride pattern and uses in conjunction when gps is weak.
Don't forget the main gripe with the techies is the hand off with the iPhone if it is close by at the start of a run, there is a period where it tries to find the phone for a couple of minutes so uses the accelerometer to calculate.

Yeah the hand off problems have made some issues for mine as well. I drive to a parking area and meet up with some friends for a great run around a lake and through some forest - but we start close to the car so it can mess it up a bit I think

 

[FunkyMagicUK]

Yeah the hand off problems have made some issues for mine as well. I drive to a parking area and meet up with some friends for a great run around a lake and through some forest - but we start close to the car so it can mess it up a bit I think

Switch watch to airplane mode before you start your run solved this for me.

Also...for marathon (and half) pacing...use the race distance markers, doing the pace math against your elapsed time is a pretty good technique that distracts you and eats up a few miles of the distance.

I generally rely on watch as i find it incredibly accurate over all distances I’ve raced from 400m up to 41miles, but do a manual check at say the halfway point, or if I’m close to a target or PB time then I’ll work out required pace in my head as the last few distance markers are passed.

 

[QCassidy352]

Not only is 2% fine, but you also have to remember that most courses are measured using optimal pathing, or close to it. Most runners don't actually run the course that efficiently for a variety of reasons.

My point being that you may, in fact, have run very slightly longer than the measured distance by virtue of taking wide turns, crossing roads inefficiently, and other very small differences between how the course was measured and where your feet actually went.

 

[xDKP]

Switch watch to airplane mode before you start your run solved this for me.

Also...for marathon (and half) pacing...use the race distance markers, doing the pace math against your elapsed time is a pretty good technique that distracts you and eats up a few miles of the distance.

I generally rely on watch as i find it incredibly accurate over all distances I’ve raced from 400m up to 41miles, but do a manual check at say the halfway point, or if I’m close to a target or PB time then I’ll work out required pace in my head as the last few distance markers are passed.

Everything that can take your mind off the pain in the legs might be a good thing