Your GPS is lying to you, at least a little bit

6 min read
4

If you’ve ever participated in a race with your GPS watch, then you should know the frustration at how the GPS measurements on your device and the actual race distance measurements are often out of sync. I’ve been to many a running race where people complain that the board markers are wrong because their device says it is. The truth is though, that GPS tracking is not an exact science and people often forget that when relying wholly on their GPS to guide their ride/run. In running races, you often start quite a distance behind the starting line, unless you’re fast enough to be seeded at the front and so there is always some distance to add on, but even then you will still be out at the end of the day. Even point-to-point cycling races are likely to see a difference in mileage between actual race markers and their GPS tracking devices.

I will hopefully, by giving certain factors that affect GPS accuracy below, be able to clear up a lot of this confusion and break down exactly how GPS devices actually track your distance and speed and why they are unlikely to be as accurate as you might like them to be. At least for now. Some of these factors are straightforward, some less so, but perhaps it will help you to appreciate the complexities of GPS tracking rather than leave you cursing because according to your device you ran a kilometer further than you should’ve. Pictures courtesy of Singletracks.

The Error Radius

A GPS never knows exactly where it is, it’s simply just estimating its current position. GPS devices are constantly calculating their location against the earth’s latitude and longitude, based on their satellite positioning. This estimation is based on different mathematical formulae (more detailed examples of which can be found here). As there is some guesswork involved in the formula, there is a certain margin for error that needs to be taken into account – and this margin is known as an error radius. A GPS unit only knows its true position is located somewhere within its error radius. It might not be a big radius, but even if it miscalculates where you are by a metre or two, its likely going to add up over a long stretch. Satellite companies are getting better at calculating this more accurately, but the margin of error remains even on the most advanced of devices today.

Measuring Straight Lines

Perhaps the most obvious reason for a GPS and actual distance to be out of alignment is that of measuring a straight line. In truth, it’s incredibly difficult to run or cycle in a straight line and as such, you will either pick up a deviation from the normal line in your path adding to your mileage or perhaps if your variations are small enough, you are actually under-measured as the GPS tracks you in a straight line when you’ve more closely resembled a slightly drunken walker:

For runners especially, you don’t take ever turn on the inside curb or you might move side to side to get to the different water points. If the GPS picks up on these movements, it will add them to your distance travelled.

Measuring Curvy Lines

This applies more to Mountain Biking where the trails are often quite twisted and not always in a smooth line. The issue here is that GPS devices ping the satellites at set intervals and so when you’re going around a sharp bend at high speeds, it’s possible the GPS device misses these turns or sharp deviations in your length because it just wasn’t pinging the satellite at the given point in time. Most GPS devices these days ping quite frequently, though in high speeds there is a change of error with this. Ironically, this also means the faster you ride, the less accurate the device will become.

The below example showing the actual path and how the GPS device tracked it through the blue dots gives an example of this:

The Third Dimension

This is perhaps the hardest part for GPS devices to actually measure, elevation. A GPS device cannot pick up every nuance of the elevation you are travelling and as such the device is making us of Pythagoras (and you thought that subject was of no value back in school) to calculate the altitude it is picking up and trying to calculate the distance travelled between its polling. This could lead to some overestimation when travelling on an uphill, whereas an under-estimation on downhills. It’s also worth remaining you that the error radius mentioned in the first point, is also in play here.

It’s important to remember that when actual race distances are measured, they use actual measuring sticks to measure the distance, which removes this issue from the equation. Something your GPS doesn’t have the benefit of doing and why it could be a problem. Again, the faster your speed, the more inaccurate the calculations will be between pings, so this is another issue more prevalent with cycling than running.

Since GPS units are able to estimate elevation, designers can use that number (basically the Z coordinate), in conjunction with the X and Y readings (latitude, longitude) to calculate 3-dimensional distances between two points on a trail. Of course GPS units are notoriously bad at estimating elevation, but we shouldn’t assume this information is ignored in GPS distance calculations. Like latitude and longitude readings, elevation readings are estimated within a range, so they can lead to either overestimates or underestimates at any given time.

So, do GPS units always get distances wrong?

Essentially, what this means, is that for simple routes, like on the road and slower speeds, a GPS device is more likely to overestimate the distance, while on curvy, fast routes, it will underestimate. This is one of the reasons why runners will run into the problem of having run more than the race markers say they have (an excruciating problem towards the end of Comrades marathon) and mountain bikers feel like they are being measured a little less than the distance travelled.

At the end of the day, while your GPS is unlikely to get things right, its unlikely to be that far off. I’ve generally found for running, a GPS device is likely to be around 300-500m our for every marathon distance, which may seem like a lot – but knowing I don’t run in a straight line, it’s probably not far off. Just be aware of these differences and put them into consideration when you run.

If you want to properly test the accuracy of a GPS device, it’s probably best to test them in conjunction with a rolling measuring stick and then have multiple devices on hand to see which one differentiates the most. Even then though, the margins for error in the calculations are not consistent and you would likely need to do this on different routes to get an idea.

Me, I’ve just made peace with trusting the distance boards and running to them rather than my watch. The GPS device is still a useful guide however and is extremely useful in planning your pacing accordingly.

Last Updated: March 6, 2017

Craig Risi

A man of many talents, but no sense how to use them. I could be discovering the cure for aids or finding ways to achieve world peace, but I’d rather be watching movies and writing here instead.

  • HvR

    *Bites his lower lip not to correct all the technical inaccuracies*

    General jist is correct.

    • Gman

      Where are the technical inaccuracies? Just curious.

  • HvR

    Maybe want to add one more thing.

    Price, not all GPS tracking devices are created equal. And as per usual the cheaper one will be the less accurate one.

    Mostly in 2 ways, the electronic GPS module which receive the GPS satellite signals and calculates position from that signals. More expensive units have more channels they can listen on for satellites on a more regular basis, so time to first lock is faster, losing satellite signal has less influence on the calculated position error.

    Second is the distance calculation, GPS module just gives you a pin point location (as in the figure in the article). The distance calculation is embedded software that is designed and implemented by device manufacture. Cheap ones just draw straight lines every-time there is a known position; more expensive units will implement better calculations, estimators and other hardware like accelerometers to catch errors and correct them.

  • ReBoot

    I work with 3 different Gps systems. Each differently priced with different purposes.

    The Garmin Gps (up to R10 000) in your car, or as we call it “the handheld gps”. Accuracy on a good day with very good satellite geometry, about 4 meter. On a bad day up to 15m accuracy.

    The DGPS (differential gps) (up to R75 000 unit and R16 000 annual fee) where you pay a yearly fee and you get a “correction” signal send to your gps receiver. For the R16 000 fee your system is just under 1m accurate.

    Used mostly in agriculture

    Survey grade RTK (real time kinematic) GPS (up to R400 000). System works based on 2 gps units.

    Crude explanation: 1 unit will be placed on a known position, the satellites will calculate a position in real time and compare it with the known position. The difference (correction) will be sent via radio signal or internet connection to the 2nd unit. The 2nd unit will recalculate its coordinate using the correction. Accuracy anything from 10mm to 40mm.

    Used in Land Surveying, Engineering, Construction etc …

Check Also

Montblanc is making smartwatches look good with the Summit

If you're in the market for a smartwatch that isn't just rich with features but looks good…