How to analize gpx file

Hi there,

another newbie asking very basic stuff. I use strava in my phone to record my sessions. I do not care too much about my top speeds (even though I like to check them out). Today I had a top speed of 28 knots foiling. I was using a new free race sail with my formula board and IQfoil with 1000 wing. Today was the first time using it in a nice/steady wind so I was able to push it a little bit but I was not expecting to have that top speed (plus, I don't think the 1000 can get up to 28, at least not if I am the rider). I decided to upload the GPX file to mapmyrun because it gives you average speeds every 160 m and it show 23.8 knots the max average. I then uploaded the file to ka72 (first time using that website) and it shows my max 2 second peak at 30 knots. I never crashed/catapulted, just dropped the sail few times to make adjustments, but still very suspicious.

any suggestions on how to check if those stats are legit?

Thanks, sailquik, for the reply. I tried to download GPSResults but my Mac says GPSResults needs to be updated.

When you refer to the run being suspicious, you are talking about all those 25 plus knots speeds I can see in the first picture surrounding the 32.5 knot one, right? I would take the 28 knots if they are legit or more likely legit :)

You are certainly right in being suspicious - this is an artifact. Going 3 knots faster than your "typical" speed on the foil feels like you're doubling your speed. Going 12 knots faster would have felt like you're about to break the sound barrier.

The GPX file you uploaded only contains location data, but no speed data. Even with a good GPS, location data are typically accurate to +- 3 m. That means the distance between two neighboring points can off by 6 meters, which translates to the calculated speed being off by 12 knots. But phones don't usually have good GPS antennas, so the error tends to be larger - and it can always be larger for some points. For point 2385 in the screen shot above, the error is 20 meters, which gives you a 1-second speed of 57 knots. Around point 2392, the position also "jumps" by a similar amount, but this time, it's distributed over about 8-10 points, which are each 6-8 knots too high. Both of these behaviors (big jump in one point, and a "distributed" jump) are quite normal in position-based speeds.

Phones differ quite a lot in how accurate their speeds are, but in general, the top speeds and 2 second speeds are quite likely to be several knots to high. The 5 x 10 second average gives you a much better idea of your real top speed, since it averages out the artifacts a lot. In this session, the 20.5 knot 5x10 average is perhaps a knot or two too high, judging by the 2nd to 5th fastest 10 second runs. Much better than the 10-12 knot error in the max and 2 second speeds!