With the London Marathon last week several reports came out showing that sports watches and fitness trackers were not as accurate as hoped. For most enthusiasts, this is quite well known and companies even sell devices to improve accuracy, for example, the Garmin Foot pod will override GPS information to create a more accurate result.
The issue lies with the GPS built into devices which relies on signals from GPS satellites. Currently, GPS accuracy for GPS-enabled smartphones are typically accurate to within a 4.9 m (16 ft.) radius under the open sky, and this accuracy then degrades when you are near buildings, bridges or trees. So in the case of the London marathon, where there is significant interference.
Dependant on the device you are using the accuracy can be improved by using data from multiple sources. With some more expensive devices, they include a GLONASS receiver which is the Russian version of GPS and expands the available number of satellites available for accurate measurement.
As already mentioned, some accessories can increase accuracy too, Garmin has their Foot Pod, but also some of the high-end watches work with a special chest HRMs which then provide various run dynamics such as cadence and ground contact time.
With some of the fitness trackers tested, recorded distances were above the actual distance, in the case of Pebble, it recorded a 17% higher distance than the actual marathon. The Pebble doesn’t have GPS and basis its estimation on steps, and therefore you should never regard it as reliable.
Which? carried out the tests to measure the accuracy and required users to walk on a calibrated treadmill at 4.8km/h for 10 minutes and run at 9.0-10km/h for 10 minutes.
However, they are not very clear on the tests done in an outdoor situation, nor on the distances done. The results the recently published used the London Marathon as an example, but it does not appear the tests were done in this scenario. So you can take the accuracy of their accuracy measurements with a pinch of salt.
In terms of GPS accuracy of my Garmin 235, the past 2 half marathons I have done in Blackpool have logged as 13.06-miles giving an inaccuracy of 0.31% this would make this more accurate than all the ones listed below.
As part of their testing, they also looked at calories burned, unsurprisingly this area is highly inaccurate and they claim in one instance a tracker reported 105% over the actual results. In my case, my Garmin nearly always reports 4K+ for calories, where I would estimate my maintenance to be 3.2-3.5K based on my high activity levels which can be up to a 20% difference.
Overall, I think the Which? report is highly inaccurate itself, however, it is good for highlighting the fact that fitness trackers are not perfect.
|Misfit Ray (£29.99)||-32% 17.8 miles|
|Garmin Forerunner 35 (£125)||-32% 17.8 miles|
|Samsung Gear Fit Pro 2 (£209)||-14% 22.55 miles|
|Tom Tom Spark 3 (£97)||-4% 25.17 miles|
|Fitbit Ace (£80)||-3% 25.43 miles|
|Polar A370 (£170)||-4% 25.17 miles|
|Garmin Vivoactive HR (£144)||-2% 25.7 miles|
|Huawei Watch 2 Sport (£240)||1% 26.5 miles|
|Apple Watch Series 1 (£204)||1% 26.5 miles|