Running Tech photo by Filip Mroz

Are You A Topper Upper?

Why are GPS devices not 100% accurate? Alan Newman investigates...

runABC reporter and athletics coach Alan Newman likes to run 'naked' – that is to say, he's not GPS-enabled (or obsessed) like most of his running colleagues seem to be! He uses a simple wristwatch that has basic chronograph and countdown timer functions, but distance for Alan is generally a matter of guesswork coupled with his own experience.

You see, in Alan's world, the only time a distance needs to be precisely measured is in a race, and everything else can be 'near enough'. And yet, even when trained volunteers have spent hours using great care and skill to measure and certify course accuracy for a race, you will often hear runners comparing results on their tech of choice and muttering about it being long – or even worse, short!

Alan measured his local 5K parkrun course for the Run Director using a surveyor's wheel. It took over four hours, with double-checks of the tricky bits, and the result was 5,010m (with 10m added for good measure). Comments are occasionally heard that the course is short or long, and sometimes both short and long on the same day, but it's always the same distance – 5,010m – so how can so many runners record slightly different results?

It seems that GPS devices for runners are not quite as accurate as we tend to believe. To use the Global Positioning System (GPS) your watch needs to lock onto four GPS satellites – three to determine a precise position using latitude, longitude and altitude plus one to adjust for any tiny error in the receiver's clock. Four satellites will equal accurate position and time. 

We know GPS can be phenomenally accurate. It's used by scientists to detect minuscule movements of tectonic plates and monitor signs that might indicate potential volcanic eruptions or snow avalanches and to record iceberg melt rates. So why isn't GPS completely accurate when measuring a run?

The answer is the GPS devices used in geophysics are relatively large, heavy, hugely expensive, and usually attached to an immovable object, or at least one that only moves at an incredibly slow rate. The GPS on your wrist, however, is comparatively tiny, light, inexpensive, has a fraction of the computing power and is strapped to an arm that is swinging vigorously back and forth, providing a moving target for those all-important four satellites.

Typically, GPS-enabled running watches are around 99% accurate in perfect conditions. That means they could be at least 100m out in a 10K race or more than 400m in a marathon. So, the next time you stagger about like a headless chicken 'rounding up' a training run, or you feel like complaining that a 10K course is 100m short or long, spare a thought for the battle with cloud cover, trees, tall buildings, bridges and other satellite obstructions your hard-working GPS has just fought ... and relax a little.

Alternatively, try running 'naked' like Alan and enjoy your new sense of freedom...!

Photo by Filip Mroz on Unsplash

Previous & Next Runtalk