I ran some analysis after downloading the tracks. GPSResults filters out the really high 60 knot runs around 16:38 for the GW60 but at the same time the GT31 was showing only 0.09 knots with a +/- of 64 knots, GPSResults rejected this as well. At both times the sat count dropped to 6/7 and I know why the data was so poor after further investigation. With you position and the surrounding mountains the sats that were just above the horizon produced poor signals which caused the anomoly, same thing happened with Bori's track. The GT31 will show the sat positions on GPSResults. For some reason KA72 didn't filter out the GW60's erronous tracks. When I run the data from both through GPSResults the filtered data is very similiar, KA72 caused the problem letting through the high peak...Dylan may need to tweek the filters.

Finally someone beat my top speed of 52 knots. Oops, that was km/h.

I checked out your track, too. There are indeed two segments with artificially high speeds, as you suspected. The first one has a big lapse in recording, with a 16.6 second long data point. The GPS lost reception for quite a while there. I have seen this happen a couple of times with my watch, too (although this did not result in very high speeds in my case).

The second "top fake speed" has all data points at 5 Hz, but the +- numbers are above 4.5, close to the maximum. The SDOP numbers are also extremely high for the first data point. This should be filtered out by any software. I cannot test it with GPS Action Replay, though, due to a GPSAR bug.

I have to get in contact with Yann (the GPSAR author) about the bug, anyway, and have to sent Manfred Fuchs some of my test files. I'll probably just start a "filtering" discussion with the two of them and also include Dylan from ka72.com.

I wrote up a little summary here: http://boardsurfr.blogspot.com/2017/05/60-knots-new-speedrecord.html

Thank you guys for checking this out. Roo, if the mountains can be the issue then I understand why it happens at Washoe. Peter, I'll check your summary tomorrow.  Also, I'll try to upload a full file of my session on 05/25 from GW-60 to see the difference.

Denis, did you do a lot of swimming during your session? There's an large amount of missing data points in your track, usually when your speed is close to zero. I have found a similar 2 second speed artifact in GW-60 traces from Melville Beach in Australia, where mountain are definitely not an issue. He got a 4 knot speed boost over his real top speed in the middle of what looks like a 4-minute swim, though. Details at http://boardsurfr.blogspot.com/2017/05/overstating-speed-by-4-knots.html

Hi Peter,

Sorry, it took me so long.. Yes, I spent some time in a water during both sessions (05/24 and 05/25), sometimes I had to jump in to clean the fin off 2-3' long tube grass/weeds that float on the surface, bends over the fin and never come off. I was suspicious about water too, do you think water can cause the glitch when it gets in contact with those cable connectors on the back side of the watch or it's only because of lost GPS signal under the water. I'll try to put the watch in a waterproof bag together with GT-31 next time.

Also, I couldn't open SBP files from those sessions with GPSResults, it could only open GPX file from 05/24 and it was also showing max speed over 70kts.

And I was able to upload full log from GW-60 on KA72 and added those numbers to my session on 05/25.

Denis, water kills the GPS reception quite well. Just a few cm are enough. Has nothing to do with the cable connectors. Since your body is mostly water, covering a GPS with parts of your body (for example when using an undergrip) also reduces the GPS signal the unit receives.

I'm not sure why you could not open the files in GPS Results. I had no problems with opening your files I downloaded from ka72.com. Make sure you are using the newest version! If you encounter the problem again, send me an email with the file.

The GPX format files do not have the information that's needed for the filters (SDoP and HDoP), so if that's what you feed GPSResults, you'll get all the artifacts you can get.

The Australians seem to be quite happy with the 50 degree weed speed fin from Black Project. It should shed stuff that normal weed fins don't shed. But since you're still using a pointer, any old weed fin may help..

Then I don't see the point of the wrist watch if the best placement for it on your head (helmet), biceps or somewhere else but not on your wrist.. I have GPSResults V6.105, haven't used it much lately. I see V6.120 available, so mine definitely needs an update. 

I hope those weeds just temporary thing and not going to ruin this season at Washoe. Try to use my Select HotRod when we have enough wind, it has about 20-25 degree angle and it's only 31 or 33cm.

The point in the wrist watch ... it's just so much easier, and everyone likes it better. Looks less geeky. The GW-60 already seems to sell better than the GW-52 ever did, even though it's expensive. But then, maybe people are used to spending a lot of money on watches. I think for Locosys, the decision to make a wrist watch was correct (and probably even based on market research).

Personally, I love the watch. I find it a lot more convenient than using an armband. It's also much easier to convince Nina to put it on that to convince her to wear an armband. I sometimes have the freestylers here put a loaner GPS on to get distance numbers, and with the wrist watch, I never have to worry about them loosing the GPS (they have lost armbands).

Usually, wearing the GPS on the wrist gives good data, even with an undergrip. Just not as good as other ways to wear it. I think you may indeed have worse reception already from the mountains around you. Artifacts like yours are not really common - I've seen them in about 1% of files from other sailors (excluding Boro, but that's a slightly different problem).