Inaccurate RYA Teaching : CTS - Quest For a New Method

[Seaworthy Lass]

Originally Posted by jackdale

The RYA method is a good approximation. Even over an hour current can change substantially. If you want a more accurate CTS then you need to account for leeway which can be done in advance only if you know the type of vessel you are sailing, the wind speed and the point of sail.

The problem with current velocity "predictions" is that they are subject to all sorts of meteorological conditions. Barometric pressure that is not normal (1013 mb) will lower or raise tides which will influence currents. Wind has a huge impact on current. Some of the early current prediction software had wind as a factor that could be input.

To be honest when I cross Georgia Strait which has changing currents especially when you get near the Fraser River, which can have strong outflow and requires a separate calculation, I use my tide tables to estimate what the current will be about half way across. Crossing the Strait to get into the Gulf Islands is problematic because one of the passes has not anchorages in which you can wait.

I do not think that any method can make an accurate prediction.

Like I tell my students about plotting. A three bearing fix (or better a couple or more transits) is the best position, an estimated position is good so is a running fix, a DR is the worst - but it beats doing nothing.

Thanks for posting Jackdale. Wish you were around the last few days when I was having issues with an RYA instructor who was insisting the method was mathematically precise and that the CTS determined by the RYA method would get you accurately to B (not D at any stage).

I agree tidal data is not accurate and that boat speed and leeway are hard to predict, but I think it is important that any method used to compute the CTS should not compound the errors so much (the RYA figure was 14% out in the last example I posted).

Could you please look at my RYA examples and confirm I am applying the method correctly?

[mcarling]

Originally Posted by Seaworthy Lass

I fully acknowledge the imprecision - tide data, estimated boat speed etc.
But it does not help if you start with a faulty method. Errors are cumulative usually!

Yes, that's right and I already acknowledged that your method is more precise than the RYA method. Unfortunately, it's also a bit more complex and therefore more prone to human error. I'm not sure that's a good trade in an introductory course. I do believe that instructors should know the limitations, faults, and weaknesses of the methods they teach, so I'm disappointed with RYA in that regard.

[mcarling]

Originally Posted by jackdale

The RYA method is a good approximation. Even over an hour current can change substantially. If you want a more accurate CTS then you need to account for leeway which can be done in advance only if you know the type of vessel you are sailing, the wind speed and the point of sail.

The problem with current velocity "predictions" is that they are subject to all sorts of meteorological conditions. Barometric pressure that is not normal (1013 mb) will lower or raise tides which will influence currents. Wind has a huge impact on current. Some of the early current prediction software had wind as a factor that could be input.

To be honest when I cross Georgia Strait which has changing currents especially when you get near the Fraser River, which can have strong outflow and requires a separate calculation, I use my tide tables to estimate what the current will be about half way across. Crossing the Strait to get into the Gulf Islands is problematic because one of the passes has not anchorages in which you can wait.

I do not think that any method can make an accurate prediction.

Like I tell my students about plotting. A three bearing fix (or better a couple or more transits) is the best position, an estimated position is good so is a running fix, a DR is the worst - but it beats doing nothing.

I agree with all of the above. The best dead reckoning method is better than throwing a dart at a chart pinned to the bulkhead while blindfolded.

[mcarling]

Originally Posted by Dockhead

This is very true, and that is why, instead of making bombastic pronuncimentioes based on some WAG, I am going to test my hunch with an experiment. My hunch may not be right, but it at least derives from multiple data points.

I just re-read my posts in this thread and I found nothing bombastic.

Your proposed experiment, as described, depends on precise input data where none are available and therefore cannot produce any meaningful results. If you have fun doing it, then it may be worthwhile for you. If you want to design an experiment that can produce meaningful results, then I suggest reading Karl Popper (heavy reading) or Nassim Nicolas Taleb (light reading).

[Andrew Troup]

re "I do not think that any method can make an accurate prediction..."

I don't think the method makes the prediction in this case; it processes the predictions.

Prediction is what you have to do, in order to come up with the input data to the computation;
such a prediction may be inaccurate or accurate.

Regardless of that, it seems to me that a computation method should not introduce capricious and possibly significant error* when a comparably practicable method does not.

*ie, larger than any error in the prediction-based data inputs

(The RYA method has another problem apart from the capricious error; it fails to solve one class of scenario altogether.

But most egregious: it instructs students NOT to plot their CTS to the destination, but to a rather misleading and arbitrary hypothetical destination. So misleading that an instructor was grievously misled.

It's not good, I reckon, when a method gets in the way of understanding the underlying first principles.)


Point 2: Please, everyone, try to live up to the forum posting standards re civil discourse.

The precursor to this thread was continually skirting disaster in that respect, and would have been considerably more valuable if it had been more mutually respectful, it seems to me.

[mcarling]

Originally Posted by Andrew Troup

It's not good, I reckon, when a method gets in the way of understanding the underlying first principles.

In my opinion, this is the strongest argument yet against the RYA method.

[Seaworthy Lass]

Originally Posted by mcarling

Yes, that's right and I already acknowledged that your method is more precise than the RYA method. Unfortunately, it's also a bit more complex and therefore more prone to human error. I'm not sure that's a good trade in an introductory course. I do believe that instructors should know the limitations, faults, and weaknesses of the methods they teach, so I'm disappointed with RYA in that regard.

Easy approximation methods are OK just as long as their limitations are understood and stressed when taught.

One of my concerns is that many (all?) of the RYA graduates believe the method is mathematically precise. I spoke to another cruising sailor who completed her yacht master certificate recently and she was warned to use a sharp pencil for the vectors for precision, otherwise she would be failed!

This sort of advice perpetuates the myth that the RYA method is totally accurate if the data is accurate. Some instructors are obviously really stressing that point.

I think many graduates complete the course believing this to be the case. The limitations of the data are drummed into them, but not that of the method of computation. My biggest concern is that students need to be taught the limitations of the method as well.

If we can find a better way that is still reasonably simple, that is even better. I think my method is not complicated once you understand it (you need to work through a few examples).

Even "eyeballing" the relationship between SB and LB in my method and applying that proportion to the last current vector would give a much more accurate result than the RYA method without much more work.

[mcarling]

I continue to hesitate on whether the Seaworthy Lass method should be taught in the RYA theory classes. I do believe it should be taught in the RYA Instructors classes, to ensure that the instructors understand the limitations of the RYA method.

Anyway, this all shows that Seaworthy Lass understands dead reckoning better than the vast majority of CF members.

[conachair]

Originally Posted by Seaworthy Lass

Easy approximation methods are OK just as long as their limitations are understood and stressed when taught.

One of my concerns is that many (all?) of the RYA graduates believe the method is mathematically precise. I spoke to another cruising sailor who completed her yacht master certificate recently and she was warned to use a sharp pencil for the vectors for precision, otherwise she would be failed!

This sort of advice perpetuates the myth that the RYA method is totally accurate if the data is accurate. Some instructors are obviously really stressing that point.

I'm not quite sure about that, I think most people who have any kind of instinct for what's going on knows that it's a CTS to D , which may or may not be at the destination but will be along the rhumb line. And in fairness as a CTS & time to D it is mathematically correct., but as it's just to the nearest hour the CTS & time to destination might be out.

Anyway, I much prefer a CTS to a waypoint a few miles short ( and maybe a bit upstream), then do the rest as pilotage.

[Seaworthy Lass]

Originally Posted by Andrew Troup

......
But most egregious: it instructs students NOT to plot their CTS to the destination, but to a rather misleading and arbitrary hypothetical destination. So misleading that an instructor was grievously misled.

According to the RYA, plotting the CTS to the destination is a cardinal sin (or more precisely the 8th deadly sin LOL).

Even where the current in the final portion of the journey is zero , students are still taught to ignore the destination point and aim for some other point.
That goes against the grain of common sense, yet still this principle is adhered to .

[Seaworthy Lass]

Originally Posted by conachair

I'm not quite sure about that, I think most people who have any kind of instinct for what's going on knows that it's a CTS to D , which may or may not be at the destination but will be along the rhumb line. And in fairness as a CTS & time to D it is mathematically correct., but as it's just to the nearest hour the CTS & time to destination might be out.

Anyway, I much prefer a CTS to a waypoint a few miles short ( and maybe a bit upstream), then do the rest as pilotage.

The RYA instructor giving me grief in the other navigation thread certainly didn't know that!

And if it is so instinctive, why didn't one single person (and there were over 10,000 views of that thread) speak up to support me when for four days I insisted the method was an approximation and that the rhumb line would be hit at this point if the data was good? .

[Weyalan]

One thing I was taught, way back in the dim and distant past (when Jesus was still playing quarterback for the Galilee Saints) was that although these new-fangled things called computers were, potentially, phenominally accurate, that accuracy was predicated on accurate / correct input and as such, it was important to be able to make a quick and simple "back of an envelope" calculation that could give a ballpark solution that one could confidently say was within an an order of magnitude of the correct answer... this could be compared with any computer-generated solution to give confidence that the computer solution had not been corrupted by incorrect data input.

To an extent, I feel the same way about (simple) DR calculations. In this day and age, electronic solutions (via GPS and chartplotters) are so readily available so as to make DR (almost) redundant. Neverthelessthe ability to make a relatively simple back of the envelope calculation / vector diagrom is a worthwhile skill.

[jackdale]

The SWL method is faulty.

The rhumb line is crucial.

Example

Travelling from Cautionary buoy to Entrance Island
Current 000 T
Current 1st hour 2 knots
Current 2nd hour 1 knot
Current 3rd hour 0 knot



Givens
Distance = 14.69 miles
Boat speed = 5 knots
CMG = 239 T

To calculate
SMG = 4.4 knots
CTS = 229 T

Chart work

[Jim Cate]

G'Day all,

I have followed this thread with interest, and have to admire the energy and attention to mathematical purity that is demonstrated.

But really, in the real world of cruising, there are few areas that have ANY tidal current data available. And as Jackdale points out, our yachts seldom achieve a constant speed for very long, nor a constant or known amount of leeway, nor have helmspeople or autopilots that can maintain a course to an accuracy of a degree or two or three...

So, I'm left with the feeling that these are academic arguments with little practical application for folks who are not traversing the Channel or Straits of Georgia or even SF Bay. And even in these well studied regions, I really doubt that one can consider the tidal currents as a nice uniform laminar flow. I suspect that these currents are full of swirls and eddies that change from hour to hour and day to day, just like the oceanic currents have been found to do. With these uncertainties, what is the point of attempting to achieving one degree accuracy in CTS? A frequent reassessment of the bearing to the destination with a thought to the general tidal flow and one's previous track is a general technique that has brought us to our intended target so far. Not elegant, and perhaps not providing the optimum sailing distance, but workable in the real world where GPS positioning is readily available, and independent of precise tidal flow information.

All in all it seems that this thread is more an attack against the teaching methods of the YRA than a useful addition to the skill set of the average cruising sailor. But, as I said, I admire the tenacity and fervor of the participants!

Cheers,

Jim

[jackdale]

BTW - here is a Power Point Show that I use to teach set and drift

http://www3.telus.net/jackdale/navle...S/detercts.pps

This is why the rhumb line (CMG) is essential.