Inaccurate RYA Teaching : CTS - Quest For a New Method

[HappySeagull]

...not sure if relevant but length of cts=(2 hours boat-speed+"ratio found" times boat-speed...)

[goboatingnow]

Originally Posted by HappySeagull

Yes
All these methods,RYA,Bowditch, whatever) are to instruct and avoid the scenario of NOT taking drift into account....

...where a clueless, hapless boat heading just "rhumb", finds herself miles downstream and lost.

Yes,goboatingnow -incremental "improvements" is what it comes down to comparing the methods.

SWL's method works. I just think it need be simpler explained. A simpler refinement of averages for instance won't lose any notable accuracy.

Determine this proportion accurately by measuring the distance from S to B (=SB) and also from B to L (=BL) and calculating:
SB / SB+BL
= 1.6 / 1.6+2.1 = 0.43

Multiply this by the strength of the current for this last hour and you now know the boat displacement due to current for this period.
= 0.8 x 0.43 = 0.3 nm

Mark 0.3 nm on the final current vector and label it K

I'd like the SWL method evolve a simpler method of "averaging "just using a ruler and /or a compass. I haven't considered what it would be, but I have an idea percolating that there is one.

Of course it works but unless you run real life comparisons you can't tell its any better. What you are doing is just assuming the tide data is as per the tide data given. In real life the numbers are gradually and constantly changing. Therefore the only method that demonstrates if its any better is to use much more accurate tide data over far shorter times and hence more numerous data points. Then you can do a real comparison.

Dave

[goboatingnow]

Originally Posted by HappySeagull

oh, and Unfortunately, this is where the RYA method can fail badly if D does not end up very close to B.
I'm repeating myself and maybe it was mentioned before that, but what if I just "improved" the RYA method by using a CTS drawn as a 1/2hour increment, "where D does not end up close to B".

ie: instead of "estimate boat goes" xmiles in yhours, draw arc" maybe "xmiles in y hours+1/2hr, draw arc in certain situations" will "cure" the problem. This steals from your method but ...

The RYA regularly teaches 30 minutes tide vectors ,especially where such tide data exits. , ie tidal stream atlas etc

[goboatingnow]

As I've said before I can't speak for the RYA, but it looks to me that they assume the more likely situation where tidal set goes from (for example)1.3 kts in the final whole calculated hour, to 1.1 kts in the final partial hour, so averaging the last 0-30 minutes won't be grossly inaccurate - certainly given all the other errors that "real life" will inevitably throw into the mix.

At long last someone begins to understand the effect of the assumptions of tide data are all important in deciding how to apply the theory. And then sees that the pursuit of further precision in the face of real life errors is just garbage in garbage out issues.

Dave

[Lodesman]

Originally Posted by goboatingnow

At long last someone begins to understand the effect of the assumptions of tide data are all important in deciding how to apply the theory. And then sees that the pursuit of further precision in the face of real life errors is just garbage in garbage out issues.

Dave

Been with you on this point all along Dave.

[Lodesman]

Originally Posted by HappySeagull

...not sure if relevant but length of cts=(2 hours boat-speed+"ratio found" times boat-speed...)

Nicely demonstrated Seagull. Of course it's relevant - it proves your triangle is correct.

[goboatingnow]

Originally Posted by Lodesman

Been with you on this point all along Dave.

Yes indeed you have. Thanks

Dave

[goboatingnow]

Originally Posted by HappySeagull

Start->

trying to finish!->

aiming at Destination start to finish. Drawing's not perfect but close enough. This is what I "see" as a rhumb line argument versus a single CTS.
Faster corrections will keep it closer to the rhumb,but every increment of "pause,think, correct,alter" will take it further from it. Oversteering(damping) is going to be yet another issue that has to apply to any steering whatever, so it might be cancelled out for any method.

I don't understand what you are trying to show in this example. Only some one very Inexperienced would adopt a singe CTS over several hours , relying on a final tide many hours in the future to actually allow you to reach your destination. , its the height of folly. Of in real life anything went wrong you could be swept into danger or find impossible to reach the destination . If the tide is faster then your boat, then you look at solutions that involve ( a) getting out of it using eddies etc or ( b) you cross in slacker periods.

All these examples point to the dangers of maths over experience.

Dave

[Seaworthy Lass]

Originally Posted by Lodesman

At the risk of sounding like a stuck record, this is what I did with my method. I've been calling it "my method" just to distinguish the way I approached the problem, but I consider it an extension of the basic tidal solution.

Lodesman, as far as I understand it, your method was to test out numerous proportions for the last hour before arriving at an approximation of the proportion. Yes, it will work, it is just extremely laborious. I agree your technique it is better thsn the RYA method though .

Originally Posted by Lodesman

This 'obsession' as you call it, is necessary only as far as the rhumb line provides a reference for completing the tidal triangle solution; without it, there are too many "unknowns" to complete the equation.

Yes, you need a triangle joining A, B and C (or K in my case) to solve the equation, but if you can't make your finish point B as you are using whole hours of displacement, why select the line AB (the rhumb line) to end up on? It is far more logical and induces less errors selecting the line CB (your boat displacement due to boat speed vector)! In other words arc off your distance directly to B, not the rhumb line.

Originally Posted by Lodesman

With your solution, you're basically comparing the deltas of the bracketed whole-hour boat vectors to estimate the length of the final part hour tidal vector. You can do the very same thing with the RYA method - for instance if the 2-hour triangle intercepts AB at C, short of B, and the 3-hour triangle intercepts the rhumbline past B at point D, then the ratio of CB and BD can be used the same way you use your KB (iirc) ratios to determine the final partial hour of tidal vector. As C B and D are on the same line, you can measure the last partial hour mathematically as CB/CD.

Originally Posted by HappySeagull

Attachment 55159
CTS is 356
and again, an eyeball would work along current vector too.

Lodesman and Happy, you are a month behind me LOL.

In my quest to come up with a better technique, on the 19th Jan I posted a description of this method with diagrams in post#623 of:
http://www.cruisersforum.com/forums/...-94879-16.html
I called the two points D1 and D2.

This method will not work in some circumstances, eg in the ones where the RYA method is is in error by more than 30 degrees, using this method will also produce wild results.

I abandoned this method in favour of my technique given in post#2 of this thread (a very detailed description with diagrams were given in posts# 621-628 plus 666-668).

[Seaworthy Lass]

Originally Posted by HappySeagull

...not sure if relevant but length of cts=(2 hours boat-speed+"ratio found" times boat-speed...)

Originally Posted by Lodesman

Nicely demonstrated Seagull. Of course it's relevant - it proves your triangle is correct.

I love it when the penny starts to drop .

I made this point about determining time taken a while back, and included it in my detailed step by step example in post #628:

Originally Posted by Seaworthy Lass

Time taken = the length of KB divided by the boat speed
= 7.7 / 4 = 1.93 hours = 1 hour 56 minutes

(It can also be determined by adding up the hours of 'boat displacment due to current' = 0.5 + 1 + 0.43 = 1.93 hours)

NOTE: It is not possible to do this using the RYA method, as they do not consider the last bit of current tidal data at all (they just extrapolate what has already occurred).

[goboatingnow]

Originally Posted by Seaworthy Lass

I love it when the penny starts to drop .

I made this point about determining time taken a while back, and included it in my detailed step by step example in post #628:

NOTE: It is not possible to do this using the RYA method, as they do not consider the last bit of current tidal data at all (they just extrapolate what has already occurred).

Actually you can. The RYA method can be used to simply sweep the arc and plot the position of the boat closer to the destination. The rhumb line is only used to facilitate " one" vector construction. One that is intuitive to the student.

I have always explained that the arc method merely represents a portion of a circle upon which the boat can be depending on its initial CTS. While the rhumb usually represents the closest point , its not the only point. In cases of large tides relative to the boat I explain that the arc may not cut the rhumb

Its a useful demo. Because I use it to warn students the foolishness of computing large CTS computations that rely on tides well into the future to actually get the boat home.

Of course nobody sails like this in real life anyway so its all moot.

As I said none of these mathematical constructions show any real improvement in accuracy as we have not compared them to models that provide more tide quantums

Dave

[goboatingnow]

Rather then people doing huge quantities of CTS compution using "stylised" tide data from exercise books, People need to plot sceanarios from actual tide data extracted from say tidal diamonds.

Using this its easy to see that attempts at accuracy, including increasing precise method of evaluating the final tidal hour are quite ridiculous in the vast majority of cases. The whole process is open to significant errors of extraction, including not actually being in the place where the tide is stated, assumptions about average tides over an hour etc .

Unless we can "benchmark" these methods we can never really know which is better or just exactly what we mean by "better" I , way back in several threads and repeated it several times, the RYA method ( if there is such a thing in reality) is based on certain assumptions about the nature of real tides, SWLs is based on a strictly interpretation of time quantum's and applying rigid boundaries to the tide data.

Until we do those examinations, the accuracy of either method simply boils down to hearsay and debates about the "fit of method to the data assumptions".

Comments such as "large errors" or " never get you to your destination" are incorrect attempts to move the model into real life without actually providing any backup as to the real life accuracy

Its also worth pointing out that the RYA Method , is merely the one promulgated by many authorities, usually only in one hour calculations

Dave

[Lodesman]

Originally Posted by Seaworthy Lass

Lodesman, as far as I understand it, your method was to test out numerous proportions for the last hour before arriving at an approximation of the proportion. Yes, it will work, it is just extremely laborious. I agree your technique it is better thsn the RYA method though .

I'm sorry, but you clearly don't understand it. It isn't "guess and test" as you seem to suggest. I take the speed made good for the 'whole-hour' portion (AD) to determine time to cover AB - the extra partial hour is added to tidal vectors and the boat speed vector - these are measured and calculated, not estimated nor guessed. This is plotted with the actual tidal set of that partial hour. Nine times out of ten, where there is a natural change in the tidal vectors the resultant triangle will close at or very close to B; in the case of the freak changes in set suggested by some of your scenarios, it would require further refinement, using the new speed made good to determine the final part hour. I don't see that as particularly laborious - I don't even need to draw any lines from C to B, as I just use the compass to tick off distance along the 'rhumbline'.

Originally Posted by Seaworthy Lass

Yes, you need a triangle joining A, B and C (or K in my case) to solve the equation, but if you can't make your finish point B as you are using whole hours of displacement, why select the line AB (the rhumb line) to end up on? It is far more logical and induces less errors selecting the line CB (your boat displacement due to boat speed vector)! In other words arc off your distance directly to B, not the rhumb line.

AB is your desired resultant - it is also the only fixed known. You don't know the length of CB, so you also can't know the length of AC. If you don't know the length of two sides, you can't make a triangle. By picking an arbitrary time component, you are able to fix the lengths of AC and CD - the rhumbline (AB+) gives you a reference to place D. Apparently the RYA picks the nearest whole hour, but there is nothing preventing you from calculating with fractions of hours, too. There are no errors induced in this method - it gives you a precise result for the information you put in.

You have an issue with extrapolation from that result - and you are correct that if the tidal set varies widely in that last partial hour, an extrapolation won't give you a good result.

Originally Posted by Seaworthy Lass

Lodesman and Happy, you are a month behind me LOL.

In my quest to come up with a better technique, on the 19th Jan I posted a description of this method with diagrams in post#623 of:
http://www.cruisersforum.com/forums/...-94879-16.html
I called the two points D1 and D2.

This method will not work in some circumstances, eg in the ones where the RYA method is is in error by more than 30 degrees, using this method will also produce wild results.

I abandoned this method in favour of my technique given in post#2 of this thread (a very detailed description with diagrams were given in posts# 621-628 plus 666-668).

I'm sorry I must have missed this post. I had asked previously what your method was, but at the time you were moving onto the new, improved SWL method - by post #623 most of us were losing the will to live compartmentalized into separate discussions within a monstrous thread, so not paying due attention to all the posts.

When I first considered how to resolve the multi-tide single CTS problem, I had considered constructing AC¹D¹ and AC²D² and comparing D¹B to BD², but rejected it for what I thought was the more intuitive method, as it was a refinement of what was presented as the RYA method. They are really just two ways of skinning the same cat.
I can't visualize where it would go wrong - can you point me to a scenario where this method will not work?

[HappySeagull]

Originally Posted by goboatingnow

I don't understand what you are trying to show in this example. Only some one very Inexperienced would adopt a singe CTS over several hours , relying on a final tide many hours in the future to actually allow you to reach your destination. , its the height of folly. Of in real life anything went wrong you could be swept into danger or find impossible to reach the destination . If the tide is faster then your boat, then you look at solutions that involve ( a) getting out of it using eddies etc or ( b) you cross in slacker periods.

All these examples point to the dangers of maths over experience.

Dave

I am astounded and a little hurt. Nobody recognizes my best ART! and yet it is Very Evocative ...to me at least. (Oh! the loneliness of the artist!)


However, You might like this effort. More accessible.

I call it "Rowboat without Range Marks"

If there's someone with a boat that has never been in this scenario,
I haven't met them.

of course, they are both from the same "period".
I call this period "Aiming for Destination"

This one- perhaps I shall rename it- "The Current is Stronger on the otherside of the Bay" same series of course.


Or, Rather "The DRIFT is stronger...becausea breeze might be additional to the current

[HappySeagull]

Originally Posted by Lodesman

...I can't visualize where it would go wrong - can you point me to a scenario where this method will not work?

me too, please. It seems dandy to me. I've been trying it on various scenarios seems ok.