Reply

Created by Windsurfunstu > 9 months ago, 15 Jun 2015

Is your alpha calculated on both jibes at each end of run? or just the one end (your fastest) then 500m back to your start point? oooor am I completely off track?

Anyone?

It was named after the Greek letter Alpha but you don't need to cross over. Not sure how the distance is specified? Maybe if you reset the proximity to a negative number it would require you to cross over, a true Alpha.

Total distance not more than 500m, a 180deg turn and not more than 50m separation between start and finish points.

Now realspeed, KA72 and gpsresults, calculate to this criteria but GPSarPRO also throws in a minimum distance of some sort.

So if you meet the the above specs but total distance is only 300m or so, the first 3 programs will give it to you but GPSarpro wont.

If you do cross over it will be way off optimum because you'll have a big upwind component.

The fastest possible alphas are made easier if there's a 50m gate and 250m gybe point marked. You can come in fast downwind to the in gate marker, square up as you pass it and carry your speed into the first leg. Gybe around the gybe marker, Then come out as fast as you can passing as close as you can to the out gate marker.

The gybe marker should probably be closer to 230m to allow for the length of the gybe.

Select to expand quote

Start - sail 250m

gybe

sail back to within 50m of start point

Well almost, but if you water start 250m away from the gybe, you won't get a good average, you need to be going as fast as possible when you enter the start gate.

Select to expand quote

So since the link on GPSTC regarding alpha's isn't working for me, ....

Can anyone clarify if the alpha speed is calculated as simply the "average of the speeds from each track point" (assuming a Doppler GPS unit)?

Select to expand quote

So since the link on GPSTC regarding alpha's isn't working for me, ....

Can anyone clarify if the alpha speed is calculated as simply the "average of the speeds from each track point" (assuming a Doppler GPS unit)?

I assume so, the only other posibility I can think of is distance over time, and distance isn't accurately calculated, so the average of each doppler speed calculation, is probably the way it's done. (I'm not sure about your reference to "trackpoints")

Cheers guys

That's got to help!

Counting can help with distance estimation, if you know your speed, (with speed chat in your ear you do), then practice counting at 1sec intervals, at 25kts 230m is 18sec.

Select to expand quote

I assume so .... (I'm not sure about your reference to "trackpoints")

Thanks, I assume so too, it would make the calculation straightforward and accurate.

By track points I simply mean the data points (1 Hz data in the GT-31).

Another question is how is distance calculated for checking the 500m max criteria? Is it the "sum of the distances between data points"?

Possibly a idea to have an officially recognised and consistent way to calculate Doppler-based Alpha?

Select to expand quote

Cheers Decrepit, .........You're an amazing source of info. Its sailors like you and many others that share so much on the seabreeze forums that the windsurfing community have a lot to be grateful for. I really appreciate it, thank you!

Hear Hear!

Select to expand quote

>>>

Another question is how is distance calculated for checking the 500m max criteria? Is it the "sum of the distances between data points"?

Possibly a idea to have an officially recognised and consistent way to calculate Doppler-based Alpha?

Yes, I think total distance is the sum of the doppler distances, but I think most of the software uses positional data to calculate the seperation.

It certainly would be a good idea if there were standard calculation methods. Alphas especially can have a lot of variation between different software, some do really short alphas down to 50m or so total distance, others don't recognise anything shorter than 400m, then there's the difference between positional and doppler calculations for the sepperation. I alows your PB alpha another doesn't.

RealSpeed and GPS Results use the same method for Alpha Calculation AFAIK. The results can sometimes vary slightly depending on the 'filter' parameters you set.*Very* occasionally, I can get a legitimate Alpha in RealSpeed that just misses the proximity circle in GPS-Results. Not sure why, but it may be to do with internal calculations and handling of trackpoint precision. Since Mal Wright (the author of RealSpeed) invented the Alpha, I trust his calcs!

Mal has the original description and rules of the Alpha in his website: www.intellimass.com/index.htm

The link is broken now but I have asked him if he can fix it.

Select to expand quote

Another question is how is distance calculated for checking the 500m max criteria? Is it the "sum of the distances between data points"?

Possibly a idea to have an officially recognised and consistent way to calculate Doppler-based Alpha?

With trackpoints data, the distance travelled is based on working out the distance between each point using a standard algorithm. You then add up the distances, and if it's less than or equal to 500m (and more than some arbitrary minimum that varies between software) then you check to see if your start point and end point are less than 50m apart (using the same formula) and if so, then you have a potential Alpha.

With doppler data, the distance travelled is stored with each point so you don't need to recalculate it, but you still need to work out how far apart the first and last points are.

In theory, the doppler data should be more accurate. It is generally a little further than the trackpoint data (since the lines travelled are curved, not straight) which means you should be covering more ground, which should mean a better Alpha. However, different devices have different reactions when you suddenly change directions... It's complicated.

Because GPS devices are really good at working when you travel in a straight line, and a bit shakier when you start turning corners, there are lots of blips and issues that can happen during an Alpha that will wind up with it being eliminated. The boundary conditions for these eliminations vary from software to software. This is why you often get different Alpha results on different software.

Dylan.