Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

DSTOFEL

macrumors 65816
Original poster
Feb 11, 2011
1,090
843
I've got the 42mm SS Apple watch and want to use it for tracking my outdoor runs (i.e. distance, calories, etc...). Since I do not plan to carry my phone with me, I'm trying to calibrate first by carrying my phone with me on a few runs....per this article on the topic from Apple:

https://support.apple.com/en-us/HT204516

Here is the issue I'm having: I went to a regulation 440 yard track today to run 2 1/2 miles to calibrate the watch. I carried my 6 Plus with me on the run. The problem is that my my watch measured 2.71 miles when I only ran 2.5 miles. Now, my wife also has an Apple watch and we ran together today (both of us carried our phones) Her watch registered the 2.5 mile run as 2.54 (very close), but mine was way off!

I tried this once before on a different track, but ran 3 miles (carrying my phone) and the watch registered 3.25. This is only about 92% accuracy rate carrying my phone (i.e. with GPS).

I've got both location services and motion calibration & distance turned on (on my phone), both of my test runs were well over 20 minutes and the tracks where I've done my test runs are both in wide open areas (clear access to the sky...so GPS connection should not be an issue).

Just curious if anyone else has had this issue?

One question I have: When I carry my phone, I don't have any particular app open (i.e. apple watch app, activity app, etc...) and the phone is in sleep mode (i.e. display goes off). Does GPS work when the phone is in sleep mode? Could this be the problem?
 
I would like to know this too. I ran this week at the Y on an indoor track where 8.5 laps equals a mile. I ran 3 miles. The Apple Watch activity app thought I ran 2.8 and the Nike Running app thought I ran about 3.7... both running at the same time. I've not run outdoors since I got the watch.
 
I would like to know this too. I ran this week at the Y on an indoor track where 8.5 laps equals a mile. I ran 3 miles. The Apple Watch activity app thought I ran 2.8 and the Nike Running app thought I ran about 3.7... both running at the same time. I've not run outdoors since I got the watch.
Interesting! Your 2.8 register on a 3 mile run is about 93.5% accurate....a bit better than my results. My 2.71 register on a 2.5 mile run is about 91.5% accurate. Just curious....did you run with your phone and if so was your phone in sleep mode while you ran? I'm assuming that GPS works while the phone is in sleep mode....but not sure!
 
In my experience, the more you run with both iPhone + Watch the more accurate it becomes. I've run about 10 six-mile runs with both since getting my watch, and now at the end of the run both Runkeeper on my phone and Apple Watch Workout.app consistently say the same distance (6.12 miles, in my case).
 
In my experience, the more you run with both iPhone + Watch the more accurate it becomes. I've run about 10 six-mile runs with both since getting my watch, and now at the end of the run both Runkeeper on my phone and Apple Watch Workout.app consistently say the same distance (6.12 miles, in my case).
I've noticed the same thing outside however when I'm on a treadmill with just the Watch I'm not near as accurate but it's close enough for me.
 
I've noticed the same thing outside however when I'm on a treadmill with just the Watch I'm not near as accurate but it's close enough for me.
Same here as well , though I calibrated mine about 6 times both running and walking on the high school track and now it is the same with or without the watch and my measures runs are almost exactly accurate.

Can't count the treadmill as after calibrating, the treadmill and my watch are about 15% off, but I can't trust the treadmill since it is years old now and has no way of calibrating in real life gps.
 
In my experience, the more you run with both iPhone + Watch the more accurate it becomes. I've run about 10 six-mile runs with both since getting my watch, and now at the end of the run both Runkeeper on my phone and Apple Watch Workout.app consistently say the same distance (6.12 miles, in my case).
Thanks for the feedback. On the runs you've done while carrying your iPhone, is your phone in sleep mode (i.e. display off) or is it on? On the runs I've done with my iPhone, the phone is in sleep mode. I "think" that GPS still functions while the phone is in sleep mode, but not 100% sure. I'm trying to rule this in/out as a possible culprit.
 
Thanks for the feedback. On the runs you've done while carrying your iPhone, is your phone in sleep mode (i.e. display off) or is it on? On the runs I've done with my iPhone, the phone is in sleep mode. I "think" that GPS still functions while the phone is in sleep mode, but not 100% sure. I'm trying to rule this in/out as a possible culprit.

Definitely screen locked. I usually run with both Runkeeper and Workout.app running on my AW while Runkeeper solely runs on the iPhone. I'll periodically unlock the iPhone to see where i am on the runkeeper map.
 
Definitely screen locked. I usually run with both Runkeeper and Workout.app running on my AW while Runkeeper solely runs on the iPhone. I'll periodically unlock the iPhone to see where i am on the runkeeper map.
Great thanks!
 
I wonder if running on a track throws off the GPS accuracy. If the watch is taking occasional GPS samples, then a track could give it false calculations of how far you traveled. I would run on a regular road with the phone and watch rather than a track if you want better calibration.
 
  • Like
Reactions: Night Spring
Thanks for the feedback. On the runs you've done while carrying your iPhone, is your phone in sleep mode (i.e. display off) or is it on? On the runs I've done with my iPhone, the phone is in sleep mode. I "think" that GPS still functions while the phone is in sleep mode, but not 100% sure. I'm trying to rule this in/out as a possible culprit.
Doesn't matter what your iPhone is doing as long as it is connected to the :apple:Watch. For instance I usually have Music, Nike+ and Garmin Live track running.

Do you know and measure your cadence and are you keeping a constant stride length? If either one vary much (hills and speed changes) the :apple:Watch (and any pedometer device) has a harder time estimating distance. :apple:Watch doesn't use the GPS to measure your distance. It uses GPS to dynamically calibrate your cadence/stride length. If these vary much accuracy will go down.

For instance I carry my iPhone 99.9% of the time. However I do an extreme trail running about 2 to 4 times a month. This throws my calibration in to a 'funk' (often off 20%) for a couple of runs. Also if road running with extreme hills effects it too.

Also on a treadmill (hate it too) I seem to increase my stride length and lower my cadence (not sure why because I'm 175-180 on flat road). My :apple:Watch always comes up 5%-10% short.
 
Last edited:
I wonder if running on a track throws off the GPS accuracy. If the watch is taking occasional GPS samples, then a track could give it false calculations of how far you traveled. I would run on a regular road with the phone and watch rather than a track if you want better calibration.
Thanks...this is a great point. A track very well could be throwing things off if the GPS is not taking constant samples. I'll clear the calibration and try on a flat surface next.
 
Doesn't matter what your iPhone is doing as long as it is connected to the :apple:Watch. For instance I usually have Music, Nike+ and Garmin Live track running.

Do you know and measure your cadence and are you keeping a constant stride length? If either one vary much (hills and speed changes) the :apple:Watch (and any pedometer device) has a harder time estimating distance. :apple:Watch doesn't use the GPS to measure your distance. It uses GPS to dynamically calibrate your cadence/stride length. If these vary much accuracy will go down.

For instance I carry my iPhone 99.9% of the time. However I do an extreme trail running about 2 to 4 times a month. This throws my calibration in to a 'funk' (often off 20%) for a couple of runs. Also if road running with extreme hills effects it too.

Also on a treadmill (hate it too) I seem to increase my stride length and lower my cadence (not sure why because I'm 175-180 on flat road). My :apple:Watch always comes up 5%-10% short.
Thanks for all of the information. Yes, on the Calibration runs I've done, I'm intentionally keeping a very steady/even stride length/speed. That's why I was perplexed by how far off the watch was (i.e. distance tracked) when just running a nice steady pace around a track. I thought it was odd that my wife's watch recorded the 2.5 mile run at 2.54 (fairly close) and mine was way off at 2.71). Another poster made a good point in that running on a track (i.e. circle) might be throwing off the tracking as the GPS isn't constantly tracking, but is instead taking "periodic" samples of location. I'm going to clear the calibration and try on a long straight surface next to see if this helps. I'm only trying to get it "reasonably" close (i.e. 98% accuracy would be good enough for me).

However, you mention that the watch doesn't use the GPS to measure distance, but instead uses GPS to dynamically calibrate your cadence/stride length. I wasn't aware of this!
 
I wonder if running on a track throws off the GPS accuracy. If the watch is taking occasional GPS samples, then a track could give it false calculations of how far you traveled. I would run on a regular road with the phone and watch rather than a track if you want better calibration.

Check out this link below to some info regarding GPS accuracy. Basic advice is to not use a track for calibration as the GPS has a very tough time on small oval distances. Better to find a straight distance for calibration purposes.

http://fellrnr.com/wiki/GPS_Accuracy
 
I would also guess a running track is not as good of a place to calibrate than straight road running - the loops may be too small for calibration/GPS to be truly effective
 
I would also guess a running track is not as good of a place to calibrate than straight road running - the loops may be too small for calibration/GPS to be truly effective


Exactly - I used to have a Safari bookmark for another GPS test site where they used a high school running track and compared various GPS watches and phones. It was absolutely amazing when the GPS route was overlayed on the track and you could see the GPS sometimes taking shortcuts on the end of the ovals and sometimes going too far outside of the end of the ovals.

A running track is, however, a good place to calibrate a foot pod if you want to use one.
 
Exactly - I used to have a Safari bookmark for another GPS test site where they used a high school running track and compared various GPS watches and phones. It was absolutely amazing when the GPS route was overlayed on the track and you could see the GPS sometimes taking shortcuts on the end of the ovals and sometimes going too far outside of the end of the ovals.

A running track is, however, a good place to calibrate a foot pod if you want to use one.
Forgot I had it Bookmarked too. Bet this is it.

http://www.singletracks.com/blog/gps/gps-distance-accuracy-test-smartphone-apps-vs-dedicated-gps/
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.