Inaccurate RYA Teaching : CTS - Quest For a New Method

[LJH]

Originally Posted by Dockhead

One should always be most grateful to those who disagree. If they turn out be right, then they have helped you learn something or overcome some error, and thus given you a big present. If they turn out to be wrong, then they've given you a chance to exercise your brain and have helped you perfect your knowledge. People who agree with you actually give you nothing. And sometimes have caused you harm by seducing you into confirming an erroneous view. I always thought it strange, that people so often thank others for agreeing with them. This wrong-headed, and probably results from insecurity - emotional need to be supported. Actually, all the thanks are owed to those who disagree.

Well put.

It has been interesting, and insightful. I have learnt more in the last week about tidal set than I would have otherwise. It puts me a step or two ahead in my planning for this summer's travels where it will come into play.

[goboatingnow]

One should always be most grateful to those who disagree. If they turn out be right, then they have helped you learn something or overcome some error, and thus given you a big present. If they turn out to be wrong, then they've given you a chance to exercise your brain and have helped you perfect your knowledge. People who agree with you actually give you nothing. And sometimes have caused you harm by seducing you into confirming an erroneous view. I always thought it strange, that people so often thank others for agreeing with them. This wrong-headed, and probably results from insecurity - emotional need to be supported. Actually, all the thanks are owed to those who disagree.

Please do not include me in your deliberations I am not wrong nor is Seworthy wrong. I have never claimed her maths were wrong , in the same way the RYA vector maths arnt wrong either.

It's all about your interpretation of tidal data.

All the rest of the argument is bum fluff. ( as is the construction of examples and the incorrect applications of the RYA method to then say its wrong )

Please

[Seaworthy Lass]

Originally Posted by Dockhead

This is becoming pornographic, and as a moderator, I must warn you to stop posting this smut. . . . before I start looking up plane tickets to Greece in Travelocity. . . .

ROFL
Maybe you should issue a strike?
It may encourage me to behave a little better.
Then again it may not .

[Dockhead]

Originally Posted by goboatingnow

The essence of SWLs method is she applies the proportion to the last tide ( irrespective of her arguments ), the RYA applies the average of they set of closest tides that apply ( whether D is on the rhumb is irrelevant. )

Therefore it boils back to a discussion about the accuracy of tidal time boundaries. I would argue that both SWL and RYAs error circles lie WITHIN the overall passage error circle ( ie an EP work up ). Hence in that regard either method is suitable. aND neither gives more precision .

precision a means relating the model to real life , not comparing each model to each other. In physics you compare theory to practice not theory to theory.

Say I have a theory on Lead acid modelling. , you have one , do we compare models no we not not. We compare to the actual situation, then we can deduce which is the more accurate or precise

sWLs whole claim to fame is using the concept of precise time boundaries to arrive at a more precise result. That only works if you beleive the assumptions is valid.
The arguments about the rhumb line are a red herring.

Dave

Entirely agree about the rhumb line. That's just chatter from those who don't understand vector triangles.

The rest is a question with an empirical answer. An interesting question. A CFer has kindly donated the very expensive Admiralty Alltides high resolution (5 minute) tide program. I'll be back in Blighty in a few days and will run some scenarios to see whether we are "splitting hairs" or not. I will put a bottle of 25 year old single malt on the proposition that we are not.

[goboatingnow]

I think its very unfortunate that these examples get brought up. It competent obscures the fact that a model is merely a guesstimate at real life. Inaccurate or misplaced precision then gives the unreliable indication that the answer is representative of real life. Nothing could be further from the truth and one of the things that must be clearly pointed to neophytes, is just because a model says something is possible doesn't ever mean it should be attempted.

The limitations o f applying tidal data to CTS is the most important thing taught not the theoretical CTS.

Dave

[goboatingnow]

Originally Posted by Dockhead

Entirely agree about the rhumb line. That's just chatter from those who don't understand vector triangles.

The rest is a question with an empirical answer. An interesting question. A CFer has kindly donated the very expensive Admiralty Alltides high resolution (5 minute) tide program. I'll be back in Blighty in a few days and will run some scenarios to see whether we are "splitting hairs" or not. I will put a bottle of 25 year old single malt on the proposition that we are not.

Correct and of it can be determined using real data and real perturbations , that SWLs method is better in more cases then the RyA then you shall have a witty concession speech. I don't see any reason to suggest that her method significantly reduces the overall passage error circle to the point where the RYA error circle lies wholly or partially outside it.

That's the test.

Dave

[goboatingnow]

So applying a real multi hour CTS with leeway , wave induced offsets, helmmsns errors , variations in boat speed , tidal inaccuracies , tide data extraction issues. , does SWLs method reduce the resulting error circle enough ( or at all ) to expose the RYA error circle.

We shall see , though I suspect the circles will remain well inside the overall error circle.

Remember in practice , for those of us that plot one to two hour CTS , or recompute CTS as we go, then all the methods converge on the one hour plot , which in both cases provide the same result.

Dave

[Dockhead]

Originally Posted by goboatingnow

I think its very unfortunate that these examples get brought up. It competent obscures the fact that a model is merely a guesstimate at real life. Inaccurate or misplaced precision then gives the unreliable indication that the answer is representative of real life. Nothing could be further from the truth and one of the things that must be clearly pointed to neophytes, is just because a model says something is possible doesn't ever mean it should be attempted.

The limitations o f applying tidal data to CTS is the most important thing taught not the theoretical CTS.

Dave

Dave, if you think logically and systematically about it, the relationship between a certain amount of uncertainty in the inputs to the uncertainty in the output is a mathematically solvable question. All of our methods have a built in error, including SWL's. Whether that error is significant or not in view of the uncertainty of the inputs CANNOT be answered with an indignant sniff, or the wave of a hand. The answer can be won only after the expenditure of a certain amount of work, which I am prepared to do, despite being not retired and not actually having time for all this mental masturbation.

[goboatingnow]

Originally Posted by Dockhead

Dave, if you think logically and systematically about it, the relationship between a certain amount of uncertainty in the inputs to the uncertainty in the output is a mathematically solvable question. All of our methods have a built in error, including SWL's. Whether that error is significant or not in view of the uncertainty of the inputs CANNOT be answered with an indignant sniff, or the wave of a hand. The answer can be won only after the expenditure of a certain amount of work, which I am prepared to do, despite being not retired and not actually having time for all this mental masturbation.

Well yes assuming you get real life data in small enough quantums. Again the first thing you have to have is a comparison with real life. So we need to start with an actual sailed voyage or preferably a series of them to the same place ( at the same time )

Then you build you passage plan model in theory modelling these perturbations mathematically ( which includes a method to compute CTS) then you compare the theory results ( one for each CTS model) to the reality , then you plot the intersecting error circles.

It would be fabulous, because noone really knows how effective accurate CTS is in real life. , the fact is I haven't been on a voyage yet that didnt have to recompute its CTS. ( that of courses ruins the thought experiment )

Good luck with that

Dave

[Dockhead]

Originally Posted by goboatingnow

Correct and of it can be determined using real data and real perturbations , that SWLs method is better in more cases then the RyA then you shall have a witty concession speech. I don't see any reason to suggest that her method significantly reduces the overall passage error circle to the point where the RYA error circle lies wholly or partially outside it.

That's the test.

Dave

OK, then we totally agree!!!

We shall see!

[goboatingnow]

Originally Posted by Dockhead

OK, then we totally agree!!!

We shall see!

Yes , and by taking odd ball numbers. But buy using a real non symmetrical multi hour passage model and comparing it.

[Dockhead]

Originally Posted by goboatingnow

Well yes assuming you get real life data in small enough quantums. Again the first thing you have to have is a comparison with real life. So we need to start with an actual sailed voyage or preferably a series of them to the same place ( at the same time )

Then you build you passage plan model in theory modelling these perturbations mathematically ( which includes a method to compute CTS) then you compare the theory results ( one for each CTS model) to the reality , then you plot the intersecting error circles.

It would be fabulous, because noone really knows how effective accurate CTS is in real life. , the fact is I haven't been on a voyage yet that didnt have to recompute its CTS. ( that of courses ruins the thought experiment )

Good luck with that

Dave

Real life will need to wait until we've entirely settled the math.

I will have high resolution tidal data from Admiralty expressed in 5 minute intervals. With that I can run a high precision model of a few real life type (taken from the tidal atlases and related to real p,aces) passages.

Since Admiralty also publish one hour resolution data in the atlases I use, we will have an excellent comparison with the results given by a few approximating methods like RYA and SWL.

I've said this before, but I think it's worth repeating here that I don't think we need even one hour tide data for excellent results. I think if we merely know the limits and timing of the limits of the tidal streams, we can calculate superbly accurate CTS, if only we can come up with the right math.

The usable level of precision is another question, of course, which we have already discussed.

[Dockhead]

Originally Posted by goboatingnow

Yes , and by taking odd ball numbers. But buy using a real non symmetrical multi hour passage model and comparing it.

Oddball numbers are really useful. There is a powerful logical principle called reductio ad absurdum which uses oddball cases to test principles which should be universal. How really universal they are can be very efficiently tested sometimes using extreme ranges of parameters.

[LJH]

Originally Posted by Dockhead

Entirely agree about the rhumb line. That's just chatter from those who don't understand vector triangles.

The rest is a question with an empirical answer. An interesting question. A CFer has kindly donated the very expensive Admiralty Alltides high resolution (5 minute) tide program. I'll be back in Blighty in a few days and will run some scenarios to see whether we are "splitting hairs" or not. I will put a bottle of 25 year old single malt on the proposition that we are not.

It will be interesting to see the high resolution problem solving compared to the hourly scenarios. Since we have been changing the tide set based on sailing time and not when the tidal hour changes the higher resolution solution may be surprising....

Looking forward to hearing about results.

[Seaworthy Lass]

Example 5

CURRENT
1st hour: 3.6 knots 270 T (pushing you west)
2nd hour: 1.6 knots 270 T
3rd hour: slack

Boat speed 4 knots
Destination is 8.4 nm due south

All perfectly reasonable realistic figures

What is the CTS and time taken?

Try the RYA method and SWL method and work it out mathematically and see how results compare