I made a tutorial video that shows some ways to analyze alphas in GPS Speedreader. Some of the tools shown are only available in the most recent version (currently 1.3.8), downloadable from ecwindfest.org/GPS/GPSSpeedreader.html
Big thanks to Decrepit for suggestions and bug finding.
There's also a tutorial about how to define background maps in GPS Speedreader. While that's a completely unnecessary function since you can export the tracks to Google Earth, it can sometime be nice to see where you set your top speeds.
Thanks Peter, your alpha tutorial taught me a fe useful things, how to measure distance with crtl-alt-drag, (that's me with linux, normal people should only need alt-drag), So people ask me how long should they count before they gybe. I've just measured a 22kt alpha and it was 12.8s. As the tutorial demonstrates you can also measure time to get back up to speed after the gybe.
And I'm loving the background images, looking at tracks without one now seems like going out without a harness on. Something is definitely missing.
And of course being able to see if alphas are too wide, helps tune your technique. Also if your GW60 gives a much better alpha than anything else, checking alphas at 60m could tell you why. There can be 2m difference between the way the watch calculates separation and other software, so if GPSSpeader shows that good alpha at a tad over 50m that's the reason.
I just uploaded GPS Speedreader version 1.4.0. This version can now export tracks in SBP format, which is useful if you want to upload a trimmed or combined file to ka72.com or other web services. It also has a "Crop to selection" function for one-step trimming, and a few minor bug fixes. Downloads are at ecwindfest.org/GPS/GPSSpeedreader.html
All downloadable files have been checked for viruses and trojans at virustotal.com, and were called "clean" by the ~ 40 different anti-malware programs the site uses. Reports are at www.virustotal.com/gui/url/e2eddb51fd715c0e8d0ea62fd8066d3ed2e8a335ab83fb9b289fee31bc2b1e62/detection (Windows version), www.virustotal.com/gui/url/6a1bdf2006c5bbec3a5e4b695d7ac98413aae1144f2751bbab0009b3c9b4aa08/detection (Mac version), and www.virustotal.com/gui/url/4e204889f854a0fee36ebba1f587e9d1d43f2bb2bd662b8db18eab9d07a29224/detection (Linux version/jar file). If you get an alert from Windows Defender or your anti-virus program, it's a false positive, or your computer or router is infected.
Thanks Peter, I've just loaded and tested with an oao file.
As expected there's just the slightest of differences caused by rounding up a mm/s file of oao to cm/s format of sbp.
Select to expand quotedecrepit said.. rounding up a mm/s file of oao to cm/s format of sbp.
Sorry, this should be rounding up/down
just rounding up makes the saved file give some higher speeds. just rounding down gives some lower speeds
Hi Decrepit, I noticed that you focused on a "22 knot Alpha" in one of your comments above - As such I'm wondering if you happen to know what the minimum average speed on the inbound leg will usually need to be to ultimately produce a 22 knot Alpha?
What the official definition of the Alpha again?
100m one way, jibe, back 100m and being 50m from the starting point?
I read it somewhere but can't find the page now.
for gpsteamchallenge we use alpha 500 which is a track of 500 meters with a jibe somewhere and no more than 50m. distance between entry and exit points.
For the other question I checked a 22 alpha in flat water and was doing about 27knts in the entry section, then 10 knts minimum speed in the turn and about 23-24 going out. I guess better people would have faster speeds during the turn and exit.
Nup, Alpha definition is a MAX of 500m and no more than 50m separating start and finish. This can lead to some very strange looking alphas. Sloshin's 22kt alpha sounds fairly typical.
Here's one of mine, ave aprox 24 in 22 out with almost 29kt peak in the bear off and just over 10kt min
Thanks for your replies gentlemen, ..very interesting!
In your case Powersloshin, I'm wondering if you rushed your 27 knot average start leg off an even faster reach? ..or is it that you have enough body weight and vigour when sufficiently powered up to accomplish 27 knots square to the wind? ..which your moniker may also imply 
..can I ask your board width and volume?
For Decrepit, that increase from 24 to 29 knots at the start of your gybe appears mighty impressive to me, is it a consequence of a handling technique, maybe like yanking your back hand in hard as you can manage to initiate the gybe, or perhaps that you have gybed immediately on a spreading gust front? ..also can I ask your board width and volume? ..And your data appears to be founded on 19 satellites - is that a Motion GPS that was also using GLONASS and maybe Galileo satellites as well?
I'm a light weight, haven't been weighed for a long time, but doubt I weigh more than 65kg, have to sail uncomfortably overpowered to scrape into 30 knots off the wind on my 55 wide/80 litre slalom board, and I don't often manage to do it. ..The Alpha is actually by far my favourite element in the GPS Team Challenge - I have three Alpha's that are into 22 knots, as assessed by KA72 from data recorded by my GW60 watch, which I have consistently set to 5 Hertz, but on each of those occasions it was on my wrist. ..More recently I have been taping the watch to the top of my helmet, although I haven't been able to determine any greater accuracy via KA72, moreover I now have significant outgoings for Duct tape, nor is it a fashionable look!
Said three Alpha's were on different days in 2020, the last (22.5 knots) actually on Christmas day, I was pleased with that, it seemed like a Christmas present. Haven't been able to get into 22 knots in 2021 though, have to think being 65 years as well as kgs isn't helping. ..Anyway the nagging question in my mind is how real are they, I notice above Decrepit that you have made a detracting comment as to the accuracy of the GW60 for Alpha's. Should I be trying to figure out how to use that complicated looking programme above to review my last Christmas present, or trying to acquire a Motion GPS, or both?
Select to expand quotePhromsky said..>>>>
For Decrepit, that increase from 24 to 29 knots at the start of your gybe appears mighty impressive to me, is it a consequence of a handling technique, maybe like yanking your back hand in hard as you can manage to initiate the gybe, or perhaps that you have gybed immediately on a spreading gust front? ..also can I ask your board width and volume? ..And your data appears to be founded on 19 satellites - is that a Motion GPS that was also using GLONASS and maybe Galileo satellites as well?
Hi Phromsky, Most of my gybes are like that, my board is set up for downwind speed, (I use as small a fin as I can get away with, si I'm not that fast across the wind but accelerate well as soon as I point downwind, going into the gybe. The GPS is a DIY, using gps, glonass and galileo.
Here's a link to the session.
www.gpsteamchallenge.com.au/sailor_session/show?date=2021-01-27&team=2
And here's the best alpha that day I posted, the one I picked above was the one closest to 22kts.
You can see a similar pattern, but I spoiled this one by going upwind on the exit. (note the bottom line here, that's the +/- value, and it's very low all the way through)
And here's a GW60 view of the same alpha.
It;s very similar but the +/- numbers go up in the gybe because it's worn on the wrist. As you've done wearing on the head would give +/- numbers similar to my DIY logger worn on my head.
So no you don't need a motion, but it is a better unit.
And yes if you really want to improve stuff, there's nothing like being able to analyse your results. I highly recommend GPSSpeadreader for this.
ecwindfest.org/GPS/GPSSpeedreader.html
Payment is voluntary beer money. It will do a direct upload to the GPSTC, and an overlay on google earth.Select to expand quote
I'm a light weight, haven't been weighed for a long time, but doubt I weigh more than 65kg, have to sail
I know you problem well, I've been as low as 62kg, but I'm about 66kg now, and 76 years old.
I have to point out that these alphas are done in very flat water! I'm hopeless in chop. I come from a wave sailing background and I don't have a gybe that works in short estuary chop
Thanks Decrepit. I realise now that there is the opportunity to review a lot more of my Alpha's in "GPSSpeedreader" than I have been seeing evidence for up to date - I'll try to get to grips with it.
Select to expand quotePhromsky said..
Thanks Decrepit. I realise now that there is the opportunity to review a lot more of my Alpha's in "GPSSpeedreader" than I have been seeing evidence for up to date - I'll try to get to grips with it.
Yep, the other useful function for alphas is the 60m and 70m, if there is a difference to the 50m alphas, you know you're going too wide.
I had a patrik95 / 62 wide and a 6.6, very flat water. Usually I just try and stay upwind, before the jibe try to go a bit faster without bearing off too much.
I hope sailquick will also share his knowledge soon, he is the one of the best with Decrepit in this forum.
Phromsky. Have a read of this thread: 
www.seabreeze.com.au/forums/Windsurfing/Gps/Alpha-Graphs?page=1
Phromsky....... have a look at 'GPS Action Replay' analysis program.
It has a gybe analysis function where it automatically sorts the answers to all of your questions.
It also has a tacks analysis function.
Select to expand quoteTe Hau said..
Phromsky....... have a look at 'GPS Action Replay' analysis program.
It has a gybe analysis function where it automatically sorts the answers to all of your questions.
It also has a tacks analysis function.
If you mean the % of average speed in, to min gybe speed, gpsspeadreader also does that.
But I don't find that very useful for alpha improvement. It doesn't consider turn radius, this is important for alphas. If you go wider than 50m you have to get back up wind, slowing the alpha, but the gybe analysis will say this is a better gybe. A snappy gybe with a low min speed, means you can go a bit downwind and accelerate quicker to a higher speed. This also wont look good on gybe analysis because the run in is faster making the ratio worse.
Here's the GPSSpeadreader gybe analysis for my PB 26kt alpha
Turns 44.339, ----- which is the percentage of min speed to ave going in. This isn't classed as a good gybe, needs to over 50% to be good.
It's probably more appropriate for slalom racing with wider boards.
GPS Speedreader is great! ..I found a 22 knot Alpha for 6 Dec 2020 that I didn't know about, because it's a shade slower than the one identified by KA72 on the same day - As seen below it is second in the list. This one actually looks in better shape to me than the slightly faster one. Although I've blown out to more than 60 metres wide I managed to get back in, maybe on a lucky wind shift
Have to admit though, that all four of my 22 knot Alpha's appear to be affected by significant +/- values, and if said values were to be applied as a subtraction it would more or less take all four a shade under 22 knots - So thats a bit troubling. ..As well, for Decrepit, they appear to be close on double your GW60 +/- values. On all these occasions my GW60 had been on my wrist, so no doubt that's a significant factor, but I'm also wondering if GPS satellites are thinner in the sky over NZ.
Glad you like speadreader, it 's a great helpSelect to expand quotePhromsky said..>>> and if said values were to be applied as a subtraction it would more or less take all four a shade under 22 knots - So thats a bit troubling. ..
Are you sure? 22.863 - 0.126 = 22.737Select to expand quote
As well, for Decrepit, they appear to be close on double your GW60 +/- values. On all these occasions my GW60 had been on my wrist, so no doubt that's a significant factor, but I'm also wondering if GPS satellites are thinner in the sky over NZ.
So I had 9 sats, in the sample above, what did you have? Mine was on my wrist as well, so that shouldn't be any different, unless you wear it upside down, or something. The antenna is under the GPS sign on the strap, that should be facing the sky. If you want to wear it the other way up, there's a screen rotate function, so you can keep the antenna up.
They have had problems in the past in Eastern Australia when sat reception was very poor, causing loss of data. So it may be similar for NZ.
Ah well what's going on with the bottom line that's in highlighted text? ..it says:
42.4s 22.863 +/- 0.876kn 498.7 m
...
I always wear the watch in the conventional way, on top of my wrist facing up. In fact I've frequently noticed and often felt somewhat surprised that the antenna almost always seems ideally positioned for a good view of the sky. Having said that I gybe with my front hand underhand.
I don't know how to bring up that column that shows the number of satellites that were in view?
Select to expand quotePhromsky said..
Ah well what's going on with the bottom line that's in highlighted text? ..it says:
42.4s 22.863 +/- 0.876kn 498.7 m
...
I always wear the watch in the conventional way, on top of my wrist facing up. In fact I've frequently noticed and often felt somewhat surprised that the antenna almost always seems ideally positioned for a good view of the sky. Having said that I gybe with my front hand underhand.
I don't know how to bring up that column that shows the number of satellites that were in view?
in GPSSpeedreader to show # sats, Settings -> Columns, then select what you want to see (# of sats is one option)
To be honest I haven't taken a lot of notice of that bottom line, but I have a feeling it could be the greatest error, not the average.
I'll see if I can figure it out and get back
looks like I'm right.
here's the overall alpha view
so the alpha has an average of +/-0.030, and bottom line says +/- 0.248.
And here's a look at mid gybe where accuracy is the lowest
Not quite the same I see the worst accuracy there as 0.274. May be it's gybe mid point? I'll ask Peter just what it is.
Select to expand quotedecrepit said..
looks like I'm right.
here's the overall alpha view
so the alpha has an average of +/-0.030, and bottom line says +/- 0.248.
And here's a look at mid gybe where accuracy is the lowest
Not quite the same I see the worst accuracy there as 0.274. May be it's gybe mid point? I'll ask Peter just what it is.
Is that 19 sats??
Select to expand quotesailquik said..
Is that 19 sats??
Yep, obviously not a gw60, must have grabbed a DIY file by mistake
Select to expand quotedecrepit said..
Glad you like speadreader, it 's a great helpPhromsky said..>>> and if said values were to be applied as a subtraction it would more or less take all four a shade under 22 knots - So thats a bit troubling. ..
Are you sure? 22.863 - 0.126 = 22.737
As well, for Decrepit, they appear to be close on double your GW60 +/- values. On all these occasions my GW60 had been on my wrist, so no doubt that's a significant factor, but I'm also wondering if GPS satellites are thinner in the sky over NZ.
So I had 9 sats, in the sample above, what did you have? Mine was on my wrist as well, so that shouldn't be any different, unless you wear it upside down, or something. The antenna is under the GPS sign on the strap, that should be facing the sky. If you want to wear it the other way up, there's a screen rotate function, so you can keep the antenna up.
They have had problems in the past in Eastern Australia when sat reception was very poor, causing loss of data. So it may be similar for NZ.
"They have had problems in the past in Eastern Australia when sat reception was very poor, causing loss of data. So it may be similar for NZ." , was this visible in any of the recorded data , in general GNSS or just specific GPS ?
Select to expand quotedecrepit said..>>>>>Not quite the same I see the worst accuracy there as 0.274. May be it's gybe mid point? I'll ask Peter just what it is.
I asked, "The bottom line in an alpha analysis shows a much higher ? number than the result line. Is this the average during the gybe? "
So Peter Answers, " you are correct, the bottom line number always shows averages, while the results show what the error estimate is if the errors are indeed independent (which sometimes is too optimistic)."
And there's also a new version here
ecwindfest.org/GPS/GPSSpeedreader.html
solves the problem of a wrong date appearing when saved to SBP
Thanks for that Decrepit. I realised the first part of the answer earlier today, when I decided to try adding up all the SDOP errors for a couple of my 2 second High's (recorded at 5 Hertz). Dividing them by 10 then of course produced exactly the same values as represented in bold on the bottom line. ..was much easier than doing same for an Alpha.
Regarding the second part of the answer, how does that work?
And which of the two possible speeds that can be arrived at are we obliged to embrace in respect to the GPS Team Challenge?
