It looks like the ground speed in thermals is related to the sampling rate in the IGC file.
The higher the sampling rate, the higher the speed in thermals.
May be a correction factor is used to correct the speed if e.g. a circle exist only of 3 point because the circle going through 3 point is longer as a triangle. But if the same correction is used if a circle exist of e.g. 10 point, the speed will be to high? Is this true?

We've thought about this a while ago and it is a good idea for fligts that have a sample rate below 8-10 seconds (less than half-circle). It is pretty useless, even misleading for logs with 12 or more second sample rates.

Actually, this is related to Nyquist sampling theory. The sample rate needs to be at least 2X the frequency you want to measure. So you need at least two sample points per circle to prevent aliasing. But cutting it this close would require a large number of circles to get usable results. Better to set the sample rate much shorter than 1/2 the circling time.

Well. As much as I have seen gliders flying around (and flown myself for last 7 years), they don't jump around like crazy and take sharp corners as seen in some flight logs. Knowing this, it shall be possible to calculate intermediate points between the flight-log positions. That would make reply and track-drawing much smoother. Even with my logs where every second is logged.

I use a sample rate of postion 1 per second . Seeyou reports ground speeds of up to 75 knots when thermaling despite wind speeds being below 10 knots, my indicated airspeed is about 45 knots. My club is using seeyou version 20. I guess there is somthing wrong with the way seeyou calculates speeds when thermaling.

Please send an example of such flight log to support at naviter dot com. Thanks.