dougfitz said:
The only accepted method of measuring road courses for IAAF events is the calibrated bicycle method, which in turn relies on a steel tape measure. Neither the surveyors wheel nor a GPS is used for accurate course measurement.
but that will measure the distance travelled by the bicycle, which will not be the same as any given walker, as the walker will not follow the exact same line. On events like you mention, participants will keep roughly to the straight line, as they want to be competitive. Road distances are generally measured for vehicles, and cars too will generally keep to the same lane and so travel roughly the same distance. But this is not the case with something like walkers on the Camino. For example, on rural roads with little traffic, people wander about all over the place, turn around to chat to someone, wander to the other side to look at the view, etc etc. Over short distances, the difference will be trivial, but it will mount up over a walk of several weeks. Different people following the same route are unlikely to walk the same distance.
dougfitz said:
SATNAV devices (GPS and GLONASS being the only two global systems available at present) can be made very accurate, but these are quite expensive survey instruments, not the hand held or vehicle mounted devices used by the general public, or the even less accurate GPS functions on smartphones of various manufacturer.
GPS devices do not measure the distance you covered, they record your position at given intervals, draw a line between each, and measure that. The accuracy will depend not only on how accurate the readings are, but also on how often a reading is taken. There are plenty of GPS tracks out there which clearly only take a reading every 5 or 10 mins, which is hopelessly inaccurate. Plus GPS often can't take continuous readings - you'd struggle to get a meaningful measurement of St Adrian's Tunnel, for example
dougfitz said:
Over the past two weekends, I have participated in check walks for the Canberra International Walking Weekend, and carried both a Garmin Etrex 30 and a Nokia smartphone running the Sportstracker application on all four of the walks. Over 20km, the Garmin was within 400m or about 2% of the length measured by bicycle, while the smartphone was just over 2km over - about 10% greater. There were similar results on the 10 km walks.
If I extrapolate that to 200km, I get 4km and 20km out respectively, and say 800km from the Pyrenees to Santiago, I get 16km and 80km! I find that rather hard to believe. I suspect the difference comes more from the way the distance between readings is calculated. Try dumping the tracks to your computer and then using a standard geometric formula, like Vincenty's, to calculate the total distance. (Though even Vincenty's won't be the same as your bicycle, as it doesn't take altitude into account.)
At the end of the day, does it really matter? There is no such thing as the 'correct' distance between places. For me, the distances given in guidebooks (and on my maps) are there to help you plan your day(s). Anyone who thinks they are "utterly precise" is deluding his/her self