Select to expand quotemathew said..
What does the error-data say ?
Error-data, what error-data? Are you implying it is an error! surely not!!
Select to expand quotemikey100 said.. Error-data, what error-data? Are you implying it is an error! surely not!!
I think it's a serious question, it would be good to see how relevant the error data is. That's the +/- values. It would be reassuring to see them very high
I suppose that max speed might come up if you attached the watch to the top of the mast.
File has been forwarded to Sailquick...
Seriously, just posted the pic as a 'lighter moment on Seabreeze' to highlight how obvious an error can be. A bit tired of the argument n negativity in the 'Speed' threads of late.
The big max speed number was a 'spike' of the type that used to be pretty common in Track Point devices way back in the Stone Age of the GPSTC, but they are quite uncommon these days with the better Doppler devices.
The spike occurred when the sailor was stationary, either launching, landing or resting so it didn't affect the end results.
The filters in the analysis software did the job they are designed to do and ignored the spike point. 
It was the acceleration filter that picked it up of course as it was a zero to 107 to zero in 0.2 seconds.
Some other points in the track were filtered in the normal way with the Doppler Error values as is normal when there is a bad signal or the GPS is submerged.
It was a pretty normal session track apart from the display max spike. With a screenless logger you would not have even known about it. 
As an extra note. If sailors come across unusual stuff like this, please send them to me or Decrepit. We are always interested to learn more about non-normal data and we can sometimes get interesting insights from these things.
I suppose from Sailquik's explanation above, that I can no longer claim to have done over 100kts on my windsurfer 
... but on a good point, I can be more confident in knowing that the speeds posted in GPSTC, using the approved devices which have been proven to produce highly authentic data, and have a checkable error data file, are in fact correct.
