Pace Accuracy

Mstrlucky74 wrote:

Thanks. How accurate have most found this to be?

Before buying my Spark3 I was using my Samsung G6 phone with MapMyRun to track my runs and for a while I ran with both -- I found that the Spark3 reported significantly shorter distances (and consequently slower avg pace times), e.g. Spark3 would say I ran 9.7 miles when the SmartPhone would say I ran 10.

Given that the Spark3 is device dedicated for running I assume that it was the more accurate of the two but short of actually running on a pre-measured course I'm not sure how to prove it.

xray79 wrote:

Distance measured by any GPS device is different per device on the same run/route. That is because of the fact that the GPS has an inaccuracy of a few %.

Understood but the question is which device provides more accurate GPS, the watch or a SmartPhone? In my experience the Spark3 consistently gives a shorter distance traveled than my Samsung smartphone. If both are enabled and calling out mile splits I can see the Spark3 drift further and further behind with each mile passed. I'm wondering if the sampling rate is different, hopefully the Spark3 is more accurate.

tfarabaugh wrote:

A variance in distance between devices is totally normal. The nature of consumer grade GPS is that each data point collected is only accurate to within 15m-30m of your actual location. We give it a lot more credit than it deserves, it is not exact, it is an estimate. The watch collects the data points and then tries to figure out a route using them (considering that some points may overlap or duplicate) and a corresponding speed. The watch then smooths the data to come up with the most reasonable route taken according to the algorithm it is programmed with. This is why every watch, even two of the same model, will come up with different distances and speeds, as they are collecting different GPS data. Plus if one device has an altimeter and one does not (your phone may but the watch does not) the difference in vertical distance will also impact the totals. Two GPS points collected may be 50m apart from one another on the horizontal plain, but if they are on a steep hill might actually be 100m in actual distance (think of the length of the base of a steep triangle versus the length of the slanted side). Each device deals with vertical distance differently and some do it better than others.

I think both the watch and the phone are doing the same thing in terms of GPS (maybe different sample rates though) so I'm not sure why one would be consistently shorter than the other.

You bring up a good point about elevation - I don't know if the smartphone is recording that real-time or its just the software (MapMyRun) that is adding that information in after the fact. TomTom's MySports web site also shows elevation for the route I ran so I am betting that is just used to let you know how much elevation you covered, not distance.

I just did some Googling about the same issue and the consensus seems to be that even if are running a very hilly course the amount of extra distance the hills adds is so insignificant that it is more of a rounding error.

Shamelessly stolen from http://www.letsrun.com/forum/flat_read.php?thread=2192407

"An 8% slope would be 8 (feet/meters/yard/cubits...) of rise for every 100 (feet/meters/yards/cubits...) of run.

Doing simple Pythagorean solution of a^2 + b^2 = c^2 where a=100, b=8, gives us:
10,000 + 64 = c^2 (where c is your distance actually run up the slope).

Solving for c gives us 100.3, or a difference of 0.3%

This means for every mile (5280 feet) measured on the flat when actually running up a steep (8%) slope, you actually ran 5297 feet, or a stunning 17 feet longer. 10 mile run? That's 170 feet difference. You sure you can map anything on mapmyrun to that level of accuracy?"