Accuracy of GPS data


SHORT VERSION: A lot of people ask, 'How good is GPS data?' For a SailTrax DIY system there are basically two answers:

      One, it is good enough for talking strategy and tactics after a day of sailing, and seeing how things like windshifts hit the fleet.

      Two, do not use it for calling a boat over the start line. If the goal is to call boats over the line then be prepared to pay A LOT more for a different system.

What follows discusses both of those two answers and provides examples of some very easy tests a user can do to see the performance of their GPS receivers. [Side bar comment: Formally the reference here should be to a Global Navigation Satellite System, GNSS, as GPS (the Global Positioning System; U.S. owned and operated) is just one version of the various GNSS accessible by a commericial receiver used for determining location. For simplicity and familiarity, the term GPS is used here rather than GNSS.]


LONGER VERSION:

Both the 'good enough' and the 'do not use it to call boats over the start line' answers boil down to the same issue: using readily available commercial GPS receivers. Yes, there are some very nice and very expensive geoposition systems which can nail down a position to within centimeters. But again, those are expensive and usually require additional hardware in various forms.

The SailTrax DIY system was conceived and designed to be easy, inexpensive, and good enough for after action discussions about strategies and tactics. Commercial grade GPS receivers generally claim 10 to 30 foot resolution, which is also roughly 3 to 10 meters. In practice the receivers do a little better than that, and who can blame the manufactuters for 'slightly underclaim, slightly over perform'? Here are two very easy tests to see how well your GPS receivers perform:

TEST ONE - the dynamic test

Load your GPS app of choice on your phone and your spouse's or friend's phone, then together go for a drive, or a walk, or maybe even a sail on the same boat. Afterwards run the GPX to KML conversion service, then plot those two KML files together. It will be seen the two paths are for the most part very close to each other. There may be periods when the tracks show the receivers move to opposite sides of each other, even though that is not what happened when the data was recorded. Such movement is that 10 to 30 foot variance in action.

TEST TWO - the static test

Start recording GPS data and then just leave the GPS unit in one place for an hour to two. Afterwards run the GPX to KML conversion service, then plot the KML file. The data will very likely shown the GPS receiver appears to have wandered around a small bit, even though it was sitting in one spot. This wandering will not be far; it will almost always be well within that 10 to 30 foot variance. [Side bar comment: The issues demonstrated in this static test tend to go away once the receivers start moving; when the receiver is moving the characteristics seen in the dynamic test prevail.] Here are the results from doing a static test with eight different GPS receivers all at once during development of the SailTrax system:


Figure N1 - The position data recorded by eight GPS receivers set down stationary on a grass field for two hours. Three receivers were Garmin Foretrex 401's, two were Foretrex 601's, and three were Moto g Play 2024 cell phones running the GPSLogger app. The displayed 20 foot radius circle provides a sense of scale. [Image is shown in half size. Click image to see it full size]


Figure N2 - Same data as Figure N1, and laid out so the change over time can be seen. [Image is shown in quarter size. Click image to see it full size]


The apps and the harware mentioned on this page are for informational purposes only; no guarantee or endorsement is implied in any way about those products or their manufacturers.


Copyright 2025 SailTrax. All Rights Reserved.