Apologies if this has been asked before (I did try and search previous threads) but does anybody have any data on whether a Garmin (in my case a 945) watch is more accurate when recording distance than an iPhone (13 Pro)?
The delta between the two devices seems to be getting bigger with time. I did a run yesterday and Garmin had it at 6.25 miles with 850 elevation gain while the iPhone was 7.05 miles with a 927 elevation gain. Obviously I would prefer to go with the iPhone 
but I’d like to use whichever is more accurate.
How is the iPhone carried? Arm band, pocket, belt? If only for not having to deal with different places people are carrying their phones, I’d venture that your 945 is more accurate give that Garmin can (realistically) only worry about it on peoples wrists. Which is also probably a better spot for GPS signals than right next to your core.
You could try looking at an overlap of both tracks against a satellite map to see if the Garmin is cutting corners or if the iPhone is meandering off of where you’d expect the path to be.
One way out of this if you’re not worried about during your run pacing listed on the device is to allow Strava to do the “correct distance” “correct elevation” thing. Then fix your distance/elevation/pace avg in Trainingpeaks or whatever after that.
But, I’d go with the Garmin.
Of note, I have always had distance/pace issues with any device if the path has lots of little jukes or turns over short distances. In other words, your stereotypical greenway going by a creek or something. Never have this issue if it’s more true street, road, sideway by street running where the turns are a lot further apart and obvious.
How is the iPhone carried? Arm band, pocket, belt? If only for not having to deal with different places people are carrying their phones, I’d venture that your 945 is more accurate give that Garmin can (realistically) only worry about it on peoples wrists. Which is also probably a better spot for GPS signals than right next to your core.
You could try looking at an overlap of both tracks against a satellite map to see if the Garmin is cutting corners or if the iPhone is meandering off of where you’d expect the path to be.
It’s in a carry case on my upper arm so should be pretty well connected throughout. Good suggestion about looking at the two tracks on the map - thanks!
One way out of this if you’re not worried about during your run pacing listed on the device is to allow Strava to do the “correct distance” “correct elevation” thing. Then fix your distance/elevation/pace avg in Trainingpeaks or whatever after that.
But, I’d go with the Garmin.
Of note, I have always had distance/pace issues with any device if the path has lots of little jukes or turns over short distances. In other words, your stereotypical greenway going by a creek or something. Never have this issue if it’s more true street, road, sideway by street running where the turns are a lot further apart and obvious.
You might have nailed it. My run yesterday was in the hills with lost of switchbacks and running along side creeks.
One way out of this if you’re not worried about during your run pacing listed on the device is to allow Strava to do the “correct distance” “correct elevation” thing. Then fix your distance/elevation/pace avg in Trainingpeaks or whatever after that.
But, I’d go with the Garmin.
Of note, I have always had distance/pace issues with any device if the path has lots of little jukes or turns over short distances. In other words, your stereotypical greenway going by a creek or something. Never have this issue if it’s more true street, road, sideway by street running where the turns are a lot further apart and obvious.
You might have nailed it. My run yesterday was in the hills with lost of switchbacks and running along side creeks.
For mountain bike folks, I’ve suggested before maybe using a wheel speed sensor to help with the whole “trees, plenty of sharp turns” thing. No equivalent for a run really other than altering the route, or updating the data using the correction feature of Strava.
Run with one of those rolling survey wheel thingies. Haha.
Have you tried this comparison on a route with a known distance? Maybe a 400 meter track?
In general, the best way to look at which GPS is most accurate, is simply looking at the map of the two tracks overlaid together. It’ll usually give you clues to where individual units go off where you went, thus being inaccurate.
DCR, I’ve seen your route tests of GPS for garmins, apple watches, etc. (the one that goes by the stadium, under the overpass….).
Possible to do one seeing how the iphone compares?
You should use mapmywalk.
GPS on a watch or phone isn’t very accurate.
Yeah, I’ll try and remember it tomorrow on my ride to record something with the iPhone 13 Pro I use (which, isn’t quite as good for testing as running, since it’s usually pretty easy to do well on road cycling).
Thanks.
FWIW, I record my bike rides using a fenix5 and have an older iphone (XR) in my jersey pocket. When I’ve dual recorded, the garmin has been more accurate.
Garmin. Garmin. Garmin. Garmin….
I did some comparisons a while ago. Multiple watch GPS devices measured my routine run within 0.02 miles. The standard deviation of my Garmin-measured runs was 0.02 miles (IIRC). Apple was around almost 0.1 miles different from my Garmin runs, and it’s standard deviation was around 0.2 (again IIRC). I don’t think Apple measures position as frequently and has the same level of noise filtering as a Garmin device.
I probably wouldn’t trust elevation accuracy so much on all devices but all I know is if I swim 1.5km along the beach then turn with my Garmin chiming every 0.5km. It chimes 3km exactly in line where I left my towel on the sand every time. My standard run loops I finish at the same point every time for distance and would be trusting a GPS making company over a phone company every time…
My garmin (fenix5) has the auto lap set for every mile, so on my running routes (typically 1/2 road, 1/2 tree-lined path) you get a pretty good idea of the repeatability. IME the garmin is consistent within +/-10m per mile, and probably within 25m for my typical 10k loop.
Agree on not trusting elevation (easy to see the error when you run a loop course and it says you ended at a different elevation than you started).
Along with all the other comments, the accuracy of your phone depends on the app using it. To make apps run faster and save battery life the location API is specifically designed to give you a ballpark location quickly. If it has hit the satellites recently then it will cache that location and the next time you ask for a location it will either return the old location or make an educated guess. For running/cycling, the app developer would specifically have to specify the highest level of precision and lowest levels of caching. Many developers don’t know to do this.
For that matter, my Garmin 945 has settings to allow it to reduce accuracy in order to extend battery life. However, the default seems to be the opposite from the phone. It appears that the Garmin defaults to high accuracy and you have to manually select lower accuracy.
Thank you everybody for the replies - It sounds like the Garmin is way more accurate - good to know!
Have you ever compared it to mapmywalk?