“My GPS watch said I ran 10.09km – why was the course long? Will you be adjusting our finish times accordingly?”
Believe it or not, emails containing sentences like this regularly land into the inboxes or race directors or race timers around the country. The purpose of this short blog is to highlight how GPS data from watches can differ, in some cases significantly, from actual officially measured distances,
HOW ARE COURSES MEASURED?
The “Calibrated Bicycle” Method (World Standard) / Jones counter method.
Step-by-step:
-
Calibrate the bike
-
The measurer rides over a precisely measured straight section (often 300–500 m).
-
They count the clicks on the Jones Counter.
-
This establishes the number of counts per kilometer.
-
-
Ride the marathon route
-
The measurer rides the shortest possible route a runner could legally take (called the “tangent line”).
-
-
Add a safety margin
-
A small extra distance ( 0.1%) is added to prevent the course from ever being short.
-
For a marathon, that’s roughly 42 extra meters.
-
-
Re-check calibration
-
The bike is recalibrated afterward to ensure no measurement drift occurred.
-
Why not use GPS:
GPS devices are not accurate enough for certification because:
-
Signal can bounce off buildings (especially in big cities)
-
Small errors add up over 42 km
-
Satellite accuracy varies
That’s why your watch might show 26.5 miles — but the official course is still exactly 26.2 miles.
GPS gets less accurate on paths and off-road for a few technical reasons — mostly related to signal quality, environment, and mapping assumptions.
When you’re on roads:
-
Many apps “snap” your location to mapped roads.
-
Small GPS errors get corrected automatically.
Off-road/footpaths:
-
Trails may not be precisely mapped.
-
There’s no snapping correction.
-
Your track zig-zags more.
-
Small position errors accumulate → inflated distance totals.
EXAMPLE:
- In both scenarios above, you can see that the course has been measured on the shortest racing line.
- The runner (yellow line) has ran over the distance in both scenarios, running closer to the middle of the road than on the racing line
- The GPS data on both is very eccentric. This tends to happen in built up areas or off road. On sharper bends, GPS, does not perform well either, so on a course with hairpins or 90 degree turns, even two runners going side by side could have 200m of a difference over a 10km road race. 1 watch could be 9.9km, the other 10.1km, despite them running together.
DONT BELIEVE US? TRY IT YOURSELF:
If you want to test our theory, head to a 400m athletics track, and run 4 laps of it in lane 1. When you get to the end, you will have run exactly 1,600m. But check your watch and see what it says. And check out the running line on your watch data when you connect it to garmin / strava etc. It will look something like this below…and is highly unlikely to be exactly 1,600m either!
(Note, image below shows a session on a 400m track, the blue is some jog recovery but you can see that despite the athlete staying in lane 1 the whole time, the GPS data suggests an erratic line throughout)
SOME EXTRA READING (courtesy of john Quigley who has measured Olympic marathon courses): https://www.corkathletics.org/news/2565-how-far-did-you-run-in-the-cork-city-marathon.html
