Site Loader

Well – the truth is it cannot possibly be. There’s a lot of misunderstandings, false statements and “myths” around how satellite positioning sytems work in general – especially regarding the tiny receivers we have in our watches or phones – mostly due to not understanding how these actually work. And I don’t mean understanding in depth every aspect of it. No, just scratching the surface reveals a lot. Please mind that this text concentrates on sportswatches (this applies to smartwatches to some extent as well). Dedicated handheld GPS devices is a completely different story.

First of all, I should have used a different acronym in the title. We commonly refer to GNSS (Global Navigation Satellite System) systems as GPS systems. It’s obviously wrong since GPS is only one of GNSS systems. Long time ago it was THE only one, true, but this is not the case anymore. Since 40 years ago to be exact 🙂 Your average, even budget, sportswatch or smartwatch is able to use at least two GNSS systems (GPS and Glonass usually) so it should be referred to as “GNSS watch” not “GPS watch”. Just sayin’ 🙂

How does the system work then? This is going to be somewhat simplified explanation. As indicated before – just scratching the surface. It should be more than enough to understand the principles and, more importantly, to see why what a lot of people keep repeating is simply wrong.

Every RNSS system is a network of satellites going around the globe. They are not “suspended” over a certain point on the earth. They go round and round all the time. For various reasons, one of them being not to crash into each other, every system is set at a different altitude. This means that satellites placed on a lower orbit need less time to complete full journey around the Earth than those on a higher orbit – they need to travel much shorter distance. There’s a popular myth saying that everytime you use your sportswatch in the same place (some even try the same time of day), the accuracy will increase overtime, because your device remembers where satellites are placed, and starting your workout everyday at the same time in the same spot means it will “see” the same satellites as the day before :-). If using only one system – it might be true if you count in that a GPS satellite needs a little less than 12 hours (11h58′) to complete full circle around the Earth. You will have the same GPS satellites in the same locations 4 minutes earlier every day. If you add a second system (pretty much all devices do this by default), which circles the earth at a different speed (GLONASS – 11h15′, GALILEO – 14h22′) things get insanely more complicated and reality is you will never be able to “see” all satellites in the same locations as any other day. This myth is not and was never true.

GNSS work based on trilateration (same as triangulation but uses distance instead angles). If you know your (in this case -the module in your sportswatch) distance to one point, you can be anywhere on a sphere which radius is defined by the distance. If you know your distance to two different points (knowing the distance between them as well), you can be anywhere on a “circle”. Much better than the whole sphere but still not good enough. Only if you know the distance to three different points you can define your location precisely. These three points form a triangle, so… you get it. However, for your receiver it’s not enough. The accuracy of time measurement in your ridiculously expensive watch is not anywhere near enough the accuracy of atomic clocks that satellites use, so it need a fourth satellite to get the time. So, though for TRI-lateration you need 3 satellites (and I hear this time and time again) your device needs to “lock” on 4 to be able to calculate your precise location.

You’ve probably seen a lot of videos testing one device against another. Usually on the same route, simultaneously. Sometimes even worn on the same wrist trying to create identical conditions for both. The logic behind this is perfectly ok, but the reality is it’s all inaccurate. Quite a lot actually, and here’s why.
The signal your device gets is affected by tons of things. It bounces off almost everything, it is blocked by buildings, mountains etc., there’s intereference with other radio signals, so even if you would wear two watches one on top of another they would get different data. A few milimiters make a difference. A few centimeters – it’s a different story. You cannot “feed” the same set of data to two devices this way so the “test” results are quite flawed. If one device is totally off all the time by a few meters, then yes, you may draw some general conclusion from such a test. Looking at the line up close does not make much sense though. Even if both devices read the signal every second, they will do so about half a second apart. When running, you will cover almost 2m in 0.5 seconds. The signal will be different, blocked by a building, a tree, a truck etc. You just cannot really compare both. And the general conclusion when comparing similar quality devices will be – they are pretty similar. I don’t want to say all these tests are useless. They are just not as decisive as they claim to be, and you should not base your decision to buy this or that device on such tests. The next thing is claimed precision of GNSS signal. First thing to take into consideration is that this is true in “ideal” circumstances. Good weather, good visibility and proper device. The antenna in your watch is way too small to achieve this. There’s also power consumption issue – regardless of how often you set your watch to check your location, the GNSS module is designed to use as little energy as possible and antenna consumes most. It simply cannot achieve the theoretical system no matter what the weather is ond how clearly you can see the sky. GPS offers in theory the best accuracy, down to 30cm (ca. 12″ – for other systems it’s 1-2m which is ca. 3-6ft), but again – when nothing interferes, with a big enough antenna and so on. You should assume your watch, based on GNSS signal alone, should achieve accuracy closer to 5m. When you look at the route it draws on the map it’s usually much more precise. And this is because it uses a myriad of other sensors to correct what it reads from GNSS. Multiaxis accelerometer, compass, altimeter, gyroscope – it’s all used to determine the speed and direction of your movement. The algorithm then predicts where you should be when the next set of data is downloaded from satellites and corrects it if the result is way off. This is how your watch gives you much more accurate location than it would if it relied solely on GNSS.

To accurately compare two watches you would have to somehow feed data from one GNSS module to both – otherwise it doesn’t give you any valuable knowledge. Given the fact that top models of different brands are very close in that regard, you would certainly get different, most possibly contradicting results if you run a usual “test” you see on Youtube multiple times. Of course testing one watch one day and then the other a week later or asking your running buddy to wear one while you wear the other makes even less sense – and I’ve seen this many times with quite decisive conclusions claimed. I’m sorry to say but it’s all pure entertainment.

I believe most watches you can buy currently give you more GNSS accuracy than you actually need. I know one always wants faster, better, more accurate data but we will not be able to reach ultimate accuracy anytime soon and besides that – you will not run faster or ride your bike further if your watch could record your route with, say, not more than 1cm error. Don’t believe Garmin/Suunto/Coros will be more precise than another Suunto/Coros/Garmin/anybrand. It will be one day than another day it will not. Devices in similar price range will usually give you very similar accuracy.