racing wisdom - the beat and mark rounding

[flaming]

For that reason I dont understand Flamings comments about the GPS - surely the VMG that matters isnt the VMG to windward but the VMG to the mark. Its getting to the mark fast that is important.

If your windward mark is in your cone - i.e you will have to tack at least once to get there, then you should be aiming to sail your optimum VMG to wind.

My comment around GPS to find this is twofold.

1. Even the new ones don't update quickly enough. So it looks bloody brilliant when you pinch a bit at first, but it isn't really... You need water speed to see when your speed starts to die, by the time your GPS catches up you've lost boatlengths, and then you shove the bow down to accelerate again, and your GPS doesn't keep up, so you're going low for too long. No good.
2. Using VMG to wp means that windshifts and where you are on the course form part of your VMG.

To expand on the second point. Imagine you are sailing along at you best upwind angle and speed. But you're measuring your progress by VMG to wp. As you sail across the course your VMG will change, decreasing the closer you get to the layline, until it =0. Then you should tack, when you should be pointing straight at the mark and your VMG should equal your speed So how do you know if you're still sailing the boat to its optimum? You don't! Ok, so it might be a gradual decrease on a long leg, and it IS useful information to the nav but it certainly doesn't work as a trimming aid.

But imagine your VMG to waypoint decreases suddenly. What just happened? Are you slow? Off the wind? Or was it a windshift? Using GPS only, how the hell are you supposed to know that? Using a readout of VMG wind, and TWD, you can see straightaway what just happened. Doesn't really matter if your calibration is off, and the numbers are different on each tack, it's the comparison from minute to minute, second to second that matters.

To decide if your boat is sailing at the right speed, speed through the water is the only thing you should use. To decide if you're sailing the right angle to the wind, your wind instrument is the only thing you should use. GPS DOES NOT WORK as a trimming aid, it's a NAVIGATION aid. Those things are different.

 

[lpdsn]

... GPS DOES NOT WORK as a trimming aid, it's a NAVIGATION aid. Those things are different.

You make some very good points, and you are right when talking about commercially available displays, but these days it is getting easier and cheaper to get at the raw instrument data. There are several products out there that will feed NMEA data out over WiFi, where it can be picked up by a tablet or laptop and recorded for later analysis.

It shouldn't be too difficult to write some software to analyse the raw GPS/GNSS position data to get around SOG averaging algorithms. Similarly, you could calculate the VMG in several alternative ways rather than as just the VMG to the windward mark.

 

[lw395]

You make some very good points, and you are right when talking about commercially available displays, but these days it is getting easier and cheaper to get at the raw instrument data. There are several products out there that will feed NMEA data out over WiFi, where it can be picked up by a tablet or laptop and recorded for later analysis.

It shouldn't be too difficult to write some software to analyse the raw GPS/GNSS position data to get around SOG averaging algorithms. Similarly, you could calculate the VMG in several alternative ways rather than as just the VMG to the windward mark.

The raw data is generally equally unsuitable as a trimming aid.
There are GPS receivers that will measure speed from the Doppler, but even that is too noisy to see short term changes in a yacht's speed and heading.
Taking the raw position data is simply too noisy and full of random error.
It typically shows my client's test antenna to be doing half a knot or so, and that is a high quality antenna bolted to a mast with a very good view of the sky.
When you start thinking about GPS signals bending around your rig, other boats etc, the errors will be even worse.
And position derived speed data is always out of date, since by definition, it relies on your previous position.
Yachts are just too slow for it, it works really well on cruise missiles etc.

I use GPS speed a little on my dinghy for tuning purposes, it is useful for getting your head around the general gain to be had from sailing high with an asy kite up, but no use on a minute by minute basis.
Also fun for bragging rights and having once or twice got 20 knots on the screen, it makes you think what pushing 30 knots must be like in a dinghy.

 

[flaming]

You make some very good points, and you are right when talking about commercially available displays, but these days it is getting easier and cheaper to get at the raw instrument data. There are several products out there that will feed NMEA data out over WiFi, where it can be picked up by a tablet or laptop and recorded for later analysis.

It shouldn't be too difficult to write some software to analyse the raw GPS/GNSS position data to get around SOG averaging algorithms. Similarly, you could calculate the VMG in several alternative ways rather than as just the VMG to the windward mark.

Even if that was possible, why bother when a bog standard Log and Wind instument (which any serious boat will of course already have) will do it all without any fancy calculations?

 

[lpdsn]

Even if that was possible, why bother when a bog standard Log and Wind instument (which any serious boat will of course already have) will do it all without any fancy calculations?

Because you could go on to do a lot more analysis. Even LW395's point about noisy position data could be handled with proper scientific techniques.

Log and wind instruments won't, for example, tell you how much you lost on each tack.

I'm not saying I'm going to do it - I haven't even started anything concrete on the intelligent AIS alarm I've been considering - but it would certainly be practical for somebody for a winter project.

 

[lw395]

Because you could go on to do a lot more analysis. Even LW395's point about noisy position data could be handled with proper scientific techniques.

L.....t.

Such as?

 

[lpdsn]

Such as?

A simple book on the subject.

http://www.amazon.co.uk/Errors-Obse...49081341&sr=8-14&keywords=treatment+of+errors

There are dozens of others. The subject is taught to pretty much every undergraduate in the physical sciences.

 

[lw395]

A simple book on the subject.

http://www.amazon.co.uk/Errors-Obse...49081341&sr=8-14&keywords=treatment+of+errors

There are dozens of others. The subject is taught to pretty much every undergraduate in the physical sciences.

Yes if you read and understood any of those books and bothered to look at the random error in a gps position, you'd know that you can only average out the errors by looking back, which introduces delay, which makes GPS very poor as a real time trimming aid. It's fairly fundamentally related to the sampling interval of GPS position and the uncertainty in each position.

Personally, I find the most useful thing is the little indicator on the B'n'G which flags speed rising or falling.
Assuming hours on the water sailing level in One Designs is a given of course.

 

[Birdseye]

The raw data is generally equally unsuitable as a trimming aid.
There are GPS receivers that will measure speed from the Doppler, but even that is too noisy to see short term changes in a yacht's speed and heading.
Taking the raw position data is simply too noisy and full of random error.
It typically shows my client's test antenna to be doing half a knot or so, and that is a high quality antenna bolted to a mast with a very good view of the sky.
When you start thinking about GPS signals bending around your rig, other boats etc, the errors will be even worse.
And position derived speed data is always out of date, since by definition, it relies on your previous position.
Yachts are just too slow for it, it works really well on cruise missiles etc.

I use GPS speed a little on my dinghy for tuning purposes, it is useful for getting your head around the general gain to be had from sailing high with an asy kite up, but no use on a minute by minute basis.
Also fun for bragging rights and having once or twice got 20 knots on the screen, it makes you think what pushing 30 knots must be like in a dinghy.

I'm not convinced by this. I accept the points about the short term accuracy of the gps but I have just as many doubts about the short term accuracy of wind instruments and logs. In both cases your system smooths the fluctuations of the data, plus in tidal waters you have the issue of tide speed varying second by second as you sail over sea bed contours. Seems to me that to some degree you need to separate the short term trimming from the course setting.

Lets invent a situation. The wind is fluctuating between 3 and 10 degrees to stbd of the upwind mark ( unless its a laid course you would be lucky to get it this close) and your boat sails at 6kn at 34 deg to the wind powered up and without pinching and 6.5kn at 37 deg slightly free. So what do you helm on either tack to get to the mark the fastest? A bit free and faster or closer and slower. Your VMG to the wind wont tell you anything about the tides and there's no reason to think its any more accurate than the gps VMG to your mark. If you maximise VMG to windward then you need to adjust tack length to reflect the fact that the mark is off the wind a bit - in this case you would be doing longer stbd tacks than port ones to hit the mark. How much longer - how do you judge without the gps?

As you have probably gathered I'm not a hot shot racer so I dont know how things are done at the highest level. I'm talking club racing here, when significant wind shifts apart, the sails would be trimmed to the desired course and the boat helmed to the tell tales .

 

[TallBuoy]

Thanks guys but the idea of these threads is to try and get some discussion going on the racing board rather than out out my problems. .

and jolly interesting threads they are too !

 

[lpdsn]

Yes if you read and understood any of those books and bothered to look at the random error in a gps position.

That's a bit rude and opinionated. For a start I studied Physics to degree level. And passed. And for a couple of years I even earned my living developing instrumentation that analysed data from very unreliable Boron Triflouride detectors.

A running average is a pretty simple technique. It's what boat instrumention does as it is satisfactory for the general market. That doesn't mean that you can't go better. There's various techniques, and if I were to actually go ahead and do something on this I'd probably start off by looking at something that identifies and discards or otherwise massages anomalous readings and uses a weighted average to minimise delays. But I'd adapt it based on sample data. Statistical treatment of data from detectors is pretty standard stuff and there are libraries available (no doubt they've moved on since I last did any work on this, but they'll still exist). Processing time is not going to be the problem.

Anyway, I'm not going to write reams on this as I don't have the spare time to get into lengthy debates. There are other things to consider. The main causes you mentioned sound like they will introduce systematic errors - and systematic errors that will vary, albeit on a longer time frame than the sample period. Real-time trimming has come to the fore but the main thing I was advocating was using GPS data to answer the question earlier on in the thread about the time lost during a tack. There theres even more opportunity to process and re-process the data to get meaningful results.

 

[lpdsn]

... I even earned my living developing instrumentation that analysed data from very unreliable Boron Triflouride detectors.

And I should also say I was not involved in the cold fusion debacle. Our stuff worked.

 

[Kerenza]

On our boat we have a mantra - tactics, helm,trim,helm and so on until tactics shouts again.
Helm uses experience and skill to drive course set by tactics. Trimmers trim to the course from driver, also watching instruments, speed especially. Tactics also watches instrument trends and instructs helm. It's a team game.
With expensive software like adrena you even get a (dynamic) point on the lay line to aim for, and with tide trend taken into account you even know which lay line is best.
My helm after many years of racing is now getting into the swing of instrument assistance and often can't believe the changes of tide stream and direction we see moment to moment, which of course we ignore in the short term.
Replaying a recent race in light winds, the difference on a (symetrical) downwind leg where we sailed by instinct for the first half (around 170 deg) then following the instrument suggested angles (around 145deg) later, proved conclusively the calculated angles were the better option, although we could have trusted the instruments even more.
Speed, vmg, polar speed, targets, optimum speed all have to be taken together to make sense.

As has been said before, gps for navigation, log for performance, so gps tells you where the destination is - how you get there is up to you!

You're right that instrument calibration is essential, although constants will eventually work themselves out and the other thing we learned is all or nothing, if you're going to follow the instruments don't throw in too many instinct based changes.

 

[lw395]

I'm not convinced by this. I accept the points about the short term accuracy of the gps but I have just as many doubts about the short term accuracy of wind instruments and logs. In both cases your system smooths the fluctuations of the data, plus in tidal waters you have the issue of tide speed varying second by second as you sail over sea bed contours. Seems to me that to some degree you need to separate the short term trimming from the course setting.

Lets invent a situation. The wind is fluctuating between 3 and 10 degrees to stbd of the upwind mark ( unless its a laid course you would be lucky to get it this close) and your boat sails at 6kn at 34 deg to the wind powered up and without pinching and 6.5kn at 37 deg slightly free. So what do you helm on either tack to get to the mark the fastest? A bit free and faster or closer and slower. Your VMG to the wind wont tell you anything about the tides and there's no reason to think its any more accurate than the gps VMG to your mark. If you maximise VMG to windward then you need to adjust tack length to reflect the fact that the mark is off the wind a bit - in this case you would be doing longer stbd tacks than port ones to hit the mark. How much longer - how do you judge without the gps?

... .

If you know how speed vs heading compromises, you drive to maximise progress to windward.
On a yacht with polar data to hand you may do that numerically, knowing your ideal wind angle and target speed for any seastate and windspeed.
In a dinghy you will work without instruments (apart from a compass), using the tell tales and jib trim etc, referenced to experience against others of the same class.
GPS hardly comes into it.
Most of the best people will do very well on a dinghy or dayboat with no gps allowed.

 

[flaming]

Lets invent a situation. The wind is fluctuating between 3 and 10 degrees to stbd of the upwind mark ( unless its a laid course you would be lucky to get it this close) and your boat sails at 6kn at 34 deg to the wind powered up and without pinching and 6.5kn at 37 deg slightly free. So what do you helm on either tack to get to the mark the fastest? A bit free and faster or closer and slower. Your VMG to the wind wont tell you anything about the tides and there's no reason to think its any more accurate than the gps VMG to your mark. If you maximise VMG to windward then you need to adjust tack length to reflect the fact that the mark is off the wind a bit - in this case you would be doing longer stbd tacks than port ones to hit the mark. How much longer - how do you judge without the gps?

Hitting your boats optimum performance and navigating are two totally separate things.

The former has a frame of reference that is all relative to the water, since what is powering your boat is the difference in velocity between the water and the air. So to gauge if you are sailing your boat to its optimum you have to use instrumentation that is relative to the water, not the ground.

Navigating is then relative to the land, so things like differences in tidal flow etc will come into play, but, assuming you're not overstood, under no circumstances should you sail your boat at a non optimum VMG to wind because the tide is with or against you, or it's not a true beat.

I suspect that part of your confusion is that on your boat you are both the helm (and thus responsible for speed) and the Tac/nav.

Think of how it works in a more traditional setup. With us, I am the helm. My responsibility once we're off the line is speed, speed and only speed. I rarely know where the first mark is, and it simply doesn't matter to me. I sail the boat to its best VMG, talking to the trimmers, and the navigator decides which tack he wants me to sail on. Then tells me when to tack.

What he is doing is evaluating how the speed I'm giving him is best used on the course. Where is the best tide? How short of the "no tide" layline should we tack if we're being swept up to the mark, or how much should we overstand if we're being swept back? But whilst we're still in the "cone" - needing to sail upwind to the mark - then we only ever want best VMG.