Ive been getting into the gps sailing a bit lately and really enjoying the competition and recently lernt how to do a good 2 second .The question i want to put to every one though is should we even have a 2 second class it seems to be more of a trick rather than an actual indication of speed id much rather see a 500m speed class instead so we can compare ours to the world record guys . What does every one else think ?

This question is asked very frequently, so there are people around who share your views.
The main reason it wasn't adopted when the GTC was set up, is the inaccuracy of 1hz devices being able to measure 500m at speed. That's why we have the 5x10 instead, that gives a similar measure of ability as the 500m but can be accurately measured by our devices.
When every body has 10hz devices the reason for not having 500m will be much less, but that will be a long time coming.
And there will still be several of us that prefer not to fix something that isn't broken.
If you really feel the need for a 500m competition that can be done on the GP3s site.

Just putting the idea out there couldnt the average for the 500 m easily be taken on a 1 hz device from the adding the speed readings /3600 untill its over 500m and then average the last 2 readings and asume linear aceleration between them then work out how much of that you need to make the 500 I think the acuracey would actually exceded that of the gps .
My point isnt so much for 500m though as for what does 2 sec actually count for its 2 seconds it seems mostly to give people a big number to boast about in light wind im running down swell and pumping the sail just to get a good 2 second .

There are certainly problems with the 2s, it's the least accurate of the divisions, but most of us have grown up with it and somehow, it's been exalted above it's worth.
I have a feeling we're stuck with it though, it's always nice to know just how fast you've gone, even though for a very short time.

Select to expand quote

decrepit said..
This question is asked very frequently, so there are people around who share your views.
The main reason it wasn't adopted when the GTC was set up, is the inaccuracy of 1hz devices being able to measure 500m at speed.

I don't follow that. If you're averaging 40 knots the first distance reading, on the second, greater than 500 metres is 514.445 metres at the 25 second mark. The computer will calculate 40 knots for that distance as accurately as any other calculation. If a rival does 515 metres in 25 seconds that's 40.04 knots, there's no other way to get a result of 40.04 knots other than to do 515 metres in 25 seconds. GPS download programs can ensure the calculations are on 1 second intervals.

I also think 500 metres is a useful distance, or even 1 km. You can end up too far offshore going for nautical miles.

(I've somehow lost my GT31, did I read somewhere that the recommended replacement doesn't download to macs?)

the speed accuracy is fine, it's the distance that's the problem, an almost 500m isn't 500m.
Daffy has the gW52 working with his Mac, I think he's using a windows emulator.

No, you are right. We shouldn't really use it because it is often too inaccurate to be very meaningful.

But it is still very popular and I must admit I like to see my peak 2 seconds speeds as well.

10 seconds was chosen for the reasons already mentioned by Decrepit, but also because it translates to a distance that can much more easily be found in a lot of places (about 200 to 250 meters, depending on how fast you are going), making it much more accessible for a lot more sailors all over the world!

It is also [i]different[/i] from the old 500m WSSRC distance to deliberately separate it from those speeds. They were traditionally measured in a very different way, and still usually are for windsurfing. The 10 second GPS speed can be done anywhere, not though a fixed timing gate, which is a quite different thing so is not usually comparable.

It is quite possible now to get a reasonably accurate 500m average speed with interpolation algorithms in the 3 dedicated stand alone analysis programs GPS-Results, RealSpeed and GPSAR-Pro). Using 1hz data is allows a bit more error to creep in above the inherent GPS error though.

The use of 10hz or faster logging (the latest UBlox chips will do up to 18hz) allows much more accurate interpolation for 500m distance. Manfred Fuchs gathered test data at Luderitz with a twin 10hz GPS prototype in 2013 which I and others wore, and the results were extremely close to the fixed gate official WSSRC speeds (almost within the error margins of the fixed gate timing itself!).

That said, 10 seconds has actually become the new standard. If it was not for the canal in Luderitz, we would not have many 500m speeds being done anymore. The vast majority of speed sailors, including the best in the world, post 10 second and 5 x 10 second speeds. Just check out the GPS-SS rankings if you want to compare your own speeds with those of AA and PD etc. There is also 10 second World Record rankings and and a large number of 10 second National records published by the WGPSSRC which are held to a significantly higher standard than the fun rankings posted on GPS-TC and GPS-SS.

So forget the old way and get with the new age! Go out and see of you can knock off Antoine's 53.86 Kts 5 x 10, or his 54.03 best 10 second run!

www.gps-speedsurfing.com/?mnu=user&val=204543&uid=412

That should give you something to aim for!

I'll be the first to admit the 2sec is a BS division, but at end of day it's just bragging rights in a no sheep station comp. I always check the 2sec first but know the 5x10 is the real one and for me the Nm is a great leveler also. But if I'm faster than my mate in the 2sec I'll take it every time

Just saying, FB

Hey Daff, I believe you're on your way to Exmouth, you'll love it, no world records but very nice me thinks. I managed a 11k run along the coast in tropical waters from Sandy Bay till no water, actually I was lost but that's another story I was camped up at Kurrajong, sail from there or Sandy Bay just down the road. Also check out Shell Beach Sharks Bay on the way up, no wind for me but lots of fun potential, so nice

just over 500m is over 500 , the last second can then be asumed to be linear in acceleration and the correct portion that second added on to make the 500m .

Select to expand quote

firiebob said..
Hey Daff, I believe you're on your way to Exmouth, you'll love it, no world records but very nice me thinks. I managed a 11k run along the coast in tropical waters from Sandy Bay till no water, actually I was lost but that's another story I was camped up at Kurrawong, sail from there or Sandy Bay just down the road. Also check out Shell Beach Sharks Bay on the way up, no wind for me but lots of fun potential, so nice

Actually, I am a bit behind schedule, being still down in Albany! Had a very nice sail today there in 14-19 knots amongst the weed. Hoping for a bit more wind tomorrow and then I will head North, checking it all out on the way, and eventually to the Sharky Bay!

Select to expand quote

firiebob said..
I'll be the first to admit the 2sec is a BS division, but at end of day it's just bragging rights in a no sheep station comp. I always check the 2sec first but know the 5x10 is the real one and for me the Nm is a great leveler also. But if I'm faster than my mate in the 2sec I'll take it every time

Just saying, FB

I agree , on the claiming it if i win it bit too

Select to expand quote

sailquik said..
No, you are right. We shouldn't really use it because it is often too inaccurate to be very meaningful.

Well where does that leave the NM? It must be calculated to 1870m, 1860m or whatever, you'll rarely land bang on the 1852m on the even second. If there's a problem with 500 metres the NM also got it.

Select to expand quote

Ian K said..

sailquik said..
No, you are right. We shouldn't really use it because it is often too inaccurate to be very meaningful.

Well where does that leave the NM? It must be calculated to 1870m, 1860m or whatever, you'll rarely land bang on the 1852m on the even second. If there's a problem with 500 metres the NM also got it.

Yep, but the percentage error is much lower. and as Daffy says, it won't get mixed up with the timed gate 500m

Funny seeing average speed on Strava when I upload my gps tracks it gives you your real average speed, quite slow.

Select to expand quote

choco said..
Funny seeing average speed on Strava when I upload my gps tracks it gives you your real average speed, quite slow.

You mean the average speed for the whole session I assume?

Select to expand quote

Ian K said..

Well where does that leave the NM? It must be calculated to 1870m, 1860m or whatever, you'll rarely land bang on the 1852m on the even second. If there's a problem with 500 metres the NM also got it.

That's correct. There will be some error in the NM but GPS-Results uses interpolation to get the exact distance and the error over a NM is very small as Decrepit said. I think I was incorrect in the earlier post. It may be only GPS-Results that has the facility to do interpolation. Mike right be able to confirm if GPSAR-Pro does as well.

I am not sure how KA-72 calculates it, but even if it just goes to the next full second, the difference would be very small at this large NM distance.

Yes the error in the result would be greater than for the NM. But thinking that any errors would be minuscule, I revisited the kinematics of year 11 physics to work it out. ( GPS turns out to be exactly analogous to the old ticker tape experiments. remember them?
www.physicsclassroom.com/class/1DKin/Lesson-2/Ticker-Tape-Diagrams )

Anyway in my hypothetical calculation I have the rider doing exactly 40 knots for exactly 500 metres and then, hit by a lucky gust maybe, accelerating at 1/2 g until he/she clocks in at the 25th second. ( I think the gust to give a 1/2 g acceleration might trip a catapult, but maybe not.)

At 40 knots the rider passes 500 metres in 24.29803 seconds. (the GPS is unaware of this of course).

The rider now accelerates at 5m/sec^2 for the 0.701970 seconds it takes until the next GPS reading at 25 seconds.

40 knots = 20.5778 m/sec ( All this would be a lot easier if we didn't talk in knots)

So in this 0.701970 of a second the rider accelerates from 20.5778 to 24.08765 m/sec.

the extra distance travelled in this 0.7 of a second of constant accelerate is easy to work out .

d = (20.5778 + 24.0876504) * 0.701970 / 2 = 15.676903 metres

So when the 25 second time interval comes about our rider has done 515.676903 metres.
Average speed = 515.67903 / 25 = 20.62707 m/sec

= 40.09 knots.

So the error in this pretty much worst case hypothetical scenario is less than 1/10th of a knot. I would be happy with that.

(Of course you can crash right on the 500 metre mark and decelerate a lot more quickly, so the scope for underestimating a valid 500 metres with GPS is greater )

Select to expand quote

Ian K said..
Yes the error in the result would be greater than for the NM. But thinking that any errors would be minuscule, I revisited the kinematics of year 11 physics to work it out. ( GPS turns out to be exactly analogous to the old ticker tape experiments. remember them?
www.physicsclassroom.com/class/1DKin/Lesson-2/Ticker-Tape-Diagrams )

Anyway in my hypothetical calculation I have the rider doing exactly 40 knots for exactly 500 metres and then, hit by a lucky gust maybe, accelerating at 1/2 g until he/she clocks in at the 25th second. ( I think the gust to give a 1/2 g acceleration might trip a catapult, but maybe not.)

At 40 knots the rider passes 500 metres in 24.29803 seconds. (the GPS is unaware of this of course).

The rider now accelerates at 5m/sec^2 for the 0.701970 seconds it takes until the next GPS reading at 25 seconds.

40 knots = 20.5778 m/sec ( All this would be a lot easier if we didn't talk in knots)

So in this 0.701970 of a second the rider accelerates from 20.5778 to 24.08765 m/sec.

the extra distance travelled in this 0.7 of a second of constant accelerate is easy to work out .

d = (20.5778 + 24.0876504) * 0.701970 / 2 = 15.676903 metres

So when the 25 second time interval comes about our rider has done 515.676903 metres.
Average speed = 515.67903 / 25 = 20.62707 m/sec

= 40.09 knots.

So the error in this pretty much worst case hypothetical scenario is less than 1/10th of a knot. I would be happy with that.

(Of course you can crash right on the 500 metre mark and decelerate a lot more quickly, so the scope for underestimating a valid 500 metres with GPS is greater )

Thanks for the Arithmetic Ian.

Most of the time the difference is in accelerating at the start and slowing down at the end, but, as you say, it tends to be a small difference most of the time.

We have had requests from quite a few sailors for WGPSSRC World, National and Spot records for 1NM. If we do this (and I think we will) it will be a change to our policy of just ratifying records based on time interval and we will use the interpolation in GPS-Results to be as accurate as possible. It will also require some additions to the Record Rules. It's a bit late for Tony Wynhoven who did the then best NM at LG a few years ago though which would have been a GPS World Record. Sorry Tony.

As the new 10z and 18hz (and higher) GPS's come through (they may be pretty close!), we will also look at ratifying World, National and Spot records for Alphas.

I can't see us doing 500m records though. That would be too confusing and conflict with WSSRC too much I think.

Meanwhile, addressing the OP. The GPS-TC ain't broken. Most people say they like it the way it is. No need for fiddling.

Select to expand quote

Ian K said..
Yes the error in the result would be greater than for the NM. But thinking that any errors would be minuscule, I revisited the kinematics of year 11 physics to work it out. ( GPS turns out to be exactly analogous to the old ticker tape experiments. remember them?
www.physicsclassroom.com/class/1DKin/Lesson-2/Ticker-Tape-Diagrams )

Anyway in my hypothetical calculation I have the rider doing exactly 40 knots for exactly 500 metres and then, hit by a lucky gust maybe, accelerating at 1/2 g until he/she clocks in at the 25th second. ( I think the gust to give a 1/2 g acceleration might trip a catapult, but maybe not.)

At 40 knots the rider passes 500 metres in 24.29803 seconds. (the GPS is unaware of this of course).

The rider now accelerates at 5m/sec^2 for the 0.701970 seconds it takes until the next GPS reading at 25 seconds.

40 knots = 20.5778 m/sec ( All this would be a lot easier if we didn't talk in knots)

So in this 0.701970 of a second the rider accelerates from 20.5778 to 24.08765 m/sec.

the extra distance travelled in this 0.7 of a second of constant accelerate is easy to work out .

d = (20.5778 + 24.0876504) * 0.701970 / 2 = 15.676903 metres

So when the 25 second time interval comes about our rider has done 515.676903 metres.
Average speed = 515.67903 / 25 = 20.62707 m/sec

= 40.09 knots.

So the error in this pretty much worst case hypothetical scenario is less than 1/10th of a knot. I would be happy with that.

(Of course you can crash right on the 500 metre mark and decelerate a lot more quickly, so the scope for underestimating a valid 500 metres with GPS is greater )

I'll way in if only for a couple of items:

- the GPS's we use today, aren't time-accurate... but they are far more accurate than positional-accurate, by several orders of magnitude. However, they are still not time-accurate, so this "over-distance problem" also exists in the time-space too. eg: the "25 second" is probably more like 24.794 (or whatever your local-timestamps are)

- we use doppler speed (so a time-based measurement) as it is about 2-orders of magnitude more accurate than the tech for position based accuracy. For example, in Ian's case here, that error is bigger than the worst-case Doppler error with the GT31.


With any of these GPS's questions, it mostly *isn't* about "how accurate we think a given GPS is"... it is almost completely about "how verifiable is a given measurement". This is the same requirement for 100m track-running and for anti-doping Tour-de-France. It is all about what we can verify as accurate. If we do go down the path of distance-measurement, error-accumulation needs to be taken into account (of which the above calculation is only part). **


** For reference, any position-based measurement *always* over-calculates your distance... which means our 516m would need to be quite a bit more.
www.i-programmer.info/news/145-mapping-a-gis/9164-gps-always-over-estimates-distances.html

Just my 2 cents but I recon it would make a awesome additional category. The category might also improve NM results (average speeds) as competitors seek faster speeds over longer durations. Its got to be a good incentive and certainly addresses peeks/squirts. The 2s category does seem to get peoples blood to boil sometimes so perhaps this could be a leveller/produce a better context

We already include the NM in the GPS-TC. One of the issues with it is that there are not so many places you can do a really fast NM. But it is a great leveller too, as some of the latest 10 sec spots can't produce really good NM's and vice vera.

I may be wrong Mat, But I think GPS-Results uses the Doppler measurements to calculate the distance.

Select to expand quote

sailquik said..

choco said..
Funny seeing average speed on Strava when I upload my gps tracks it gives you your real average speed, quite slow.

You mean the average speed for the whole session I assume?

Yes that's right for whole session I think my average was around 28-29knts (53kmh) but in reality not that good an average

Select to expand quote

mathew said..

- the GPS's we use today, aren't time-accurate... but they are far more accurate than positional-accurate, by several orders of magnitude. However, they are still not time-accurate, so this "over-distance problem" also exists in the time-space too. eg: the "25 second" is probably more like 24.794 (or whatever your local-timestamps are)

I can understand why GPS is time based, but let's face it, humans are averse to time. We associate time with ... late for work, over all ready, when I'm 64.... If you go down to the bank at a speed sailing venue the chatter is ( apart from alphas ) all about Peak Speed, Nautical Miles and Distance travelled.



Another distance category, 500 metres, would be very popular, especially if the GPS programmers could arrange an on-the-bank display without need for download.

"When I'm 64!" Hey! I resemble that comment!

Mat Robertson just found and sent me a link that validates all my theories on Positional data v's Dopler data. And it illustrates perfectly why we should not use positional data if we can possibly have an alternative.

Researchers have explained why consumer GPS always overestimates distances. Overestimating distance will always overestimate positional speed by the same amount. In their experiments that found that overestimation was between 12% and 20%!! In practice for out use, the overestimation is less than that as we generally go in relatively straight lines, but it is significant nonetheless.

tech.slashdot.org/story/15/11/15/1536235/gps-always-overestimates-distances

Full article quoted: www.i-programmer.info/news/145-mapping-a-gis/9164-gps-always-over-estimates-distances.html

Select to expand quote

Ian K said..

mathew said..

- the GPS's we use today, aren't time-accurate... but they are far more accurate than positional-accurate, by several orders of magnitude. However, they are still not time-accurate, so this "over-distance problem" also exists in the time-space too. eg: the "25 second" is probably more like 24.794 (or whatever your local-timestamps are)

I can understand why GPS is time based, but let's face it, humans are averse to time. We associate time with ... late for work, over all ready, when I'm 64.... If you go down to the bank at a speed sailing venue the chatter is ( apart from alphas ) all about Peak Speed, Nautical Miles and Distance travelled.



Another distance category, 500 metres, would be very popular, especially if the GPS programmers could arrange an on-the-bank display without need for download.

//

?rel=0

Mat. Probably more like 24.999

Ian. I take your point about time getting a bad rap, but you can find just as many, or more instances in sport where the time is the quoted measure. e.g.. 6.1 seconds for the drag race. 9 seconds for the fastest man in the Olympics. 1min 34.2 secs for the Downhill ski run, etc. Oh, and at my speed spot, the talk on the bank is about speeds for 2 seconds and 10 seconds. We don't get many NM's.

Interestingly, the OP question was about the 2 second category. It's worth pointing out that a speed sailor covers 40-50 meters in that 2 seconds at top speed. If you blink you will miss a lot of the scenery!

i like the 2 sec peak as it shows what was possible when everything lines up for a perfect run which doenst happen very often.

A speed run is typically made of 3 parts - the approach, the sling shot and then the drive off the wind

the 2 sec is all about the sling shot where is the 5x 10 is more about holding the speed on the drive that comes after.

It shows what is possible on the gear your using and is useful as a counter point to the 5x10,

As you improve your techniques you can see the gap between the 2 and the 10 get smaller.

for speed strips its really useful to help fine tune between runs as it gives instant feed back on how good the sling shot was

Select to expand quote

Roar said..
i like the 2 sec peak as it shows what was possible when everything lines up for a perfect run which doenst happen very often.

A speed run is typically made of 3 parts - the approach, the sling shot and then the drive off the wind

the 2 sec is all about the sling shot where is the 5x 10 is more about holding the speed on the drive that comes after.

It shows what is possible on the gear your using and is useful as a counter point to the 5x10,

As you improve your techniques you can see the gap between the 2 and the 10 get smaller.

for speed strips its really useful to help fine tune between runs as it gives instant feed back on how good the sling shot was

Well said.

Oh, where can I get one of your slingshots? Mine seems to be broken!

Select to expand quote

sailquik said..
>>>
Well said.

Oh, where can I get one of your slingshots? Mine seems to be broken!

Not sure I've every had any.