US Portsmouth v2

[The Q]

I think one of the problems with mapping D-PN to RYA is in America they generally sail on much larger waters. Bigger waters mean fast boats perform better, due to straight line speed.

Even in the UK handicaps don't correspond nationally. If you've gone and bought the ultimate planing machine or foiler, don't expect to beat your handicap on a narrow river club.
For our tiny waters a boat, that tacks well, but doesn't plane, will often perform better.

 

[Curious2]

When I did some playing I got a better mapping to the faster and slower boats with this extremely empirical formula:
RYA= (6 + (DPN /16)) * DPN,
but it will equal 11.58 with a DPN a bit over 89.
There were a fair number of classes that didn't seem to map at all well even with the more complex sum though.

Isn't any conversion method going to have difficulty dealing with the fact that the US PYs are often so old and obviously wrong?

While rating the Moth at 15% slower than a Europe and the Int 14 as slower than a 505 could perhaps be put down to the use of ancient data, any system that (for example) rates the Laser 4.7 as being faster than a Laser Radial and the little Laser 13 cruiser faster than the longer, lighter Laser II with its trap and kite seems to suffer from a lack of a common-sense test.

 

[JimC]

The trouble with doing evaluations of handicap systems is figuring out what "correct" is. I've noticed in the past that sailors will abandon an old rating system because they don't think it works adequately, and then complain its replacement must be flawed because it doesn't give the same results as the old one! Pragmatically I think its better to try to avoid making public judgements, and just do the best you can.

I produced my formula by lining up as many UKPY and DPN numbers as possible and calculating the ratio between them. That gave me a general trend, albeit with some major inconsistencies, so it was just a question of constructing a formula that matched the general trend. By following the trend its not really necessary to make a judgement on whether DPN or UKPY is "wrong" for the major outliers, if indeed its not a case of both being wrong.

I'd be the last person to claim that UK PY numbers are definitively correct, and there are definitely aspects I have my doubts about, but I will say that I think its a pretty good job in the circumstances and for now I can't think of improvements that could be practically implemented. Its worth noting that a lot of work has been put into the UK PY in the last ten years, the data handling has changed greatly, and I'm convinced its much better now than it was then.

 

[Rambler]

The Australian Yardsticks are marked by Australian Sailing as 'reliable', 'probable' and 'tentative' with most being in the latter category but are still just used with no (or very little) complaint. They are still the basis of 'most' (or at least a lot of) club racing in Australia.

Here's a link to them least they be of any use (including the introductory notes); https://cdn.revolutionise.com.au/cups/hyc/files/xetsiuifgr6eyls4.pdf

We've had to deal with three unrated classes in the last year or two; the 49erFX, 13ft skiff and our 59erXS.

The FX one became fairly important when we had two of them as regular participants in our twin wire skiff fleet and at least one was fairly well sailed (National Champion in a year when the Olympian hopefuls were away). Portsmouth didn't rate them either. We managed to be directed to another UK rating ('Inland waterways', or something like that) and did a conversion from there. Mathematically, the conversion came out at about 80 to 81 on the Australian system. The committee was generous and gave them 83. We might be starting to regret that

Neither the 13ft skiff nor the 59erXS has any hope of being rated elsewhere. The 13's just don't mix race (normally), and the XS (really think 29erXS) hasn't had time to develop one. So we figured the 13 is faster than a 29er, so we'll give it 95, and the XS slower than a MG14, so we'll give it 108 (probably rather mean). They are presently club 'training' boats. If someone starts winning in them, it's time for them to buy themselves a boat of their own; although the 13 is intended as the next class we build as we have run out of Formula 15's available for purchase, so we might eventually have to take it more seriously.

But a number of points are worth making.

The first is that there is very little bitching about the ratings (at any club I've been at). It is what it is. On the very rare occasion one has proved to be defective, a fairly civil meeting has sorted it out.

The second is that, sometimes the geography or weather of a location will not give a class the best chance of success, even if the rating is generally appropriate. We have just such an instance where the RS100 races in our gennaker fleet, but on a river with a fast flowing tide will more often than not be disadvantaged against the twin wire boats (but not always - they still get the occasional win). That has not held back the adoption of the class at our club. It is second only the the twin wire Formula Fifteens in one design class numbers.

Finally, where a mixed fleet consists of a reasonable number of each of an individual class of boat, people much more focus on how they are doing against their class than the overall 'fleet championship'. And it only takes, say three or four boats of a class for that effect to kick in. Which is probably what helps moderate any bitching about the ratings.

 

[Foredeck Shuffle]

@Rambler or anyone else in the know, were there ever any attempts to map the ASY PN rated boats to PY boats or even DPN boats?

We have not had problems with people worried about ratings. The only concern has been the issue of a new boat with no prior rating comes into the yard and we have to start figuring that out. This is the real issue we want to solve, ratings that are as fair as are reasonably possible so we do not disenfranchise someone with a bad boat rating.

Funny, a well intentioned person asked me last night if I had considered ORC ratings. I would love to try ORC, except the costs are too high. ORC would require hull measurements and that process would be several hundred dollars the first year. Certificates cost $9 USD per foot, per year. An ILCA would cost $126/per year. That does not have legs.

 

[The Q]

The RYA has a limited Data list as well as the full PY list.

portsmouth-yardstick

www.rya.org.uk

 

[Foredeck Shuffle]

The RYA has a limited Data list as well as the full PY list.

portsmouth-yardstick

www.rya.org.uk

I merged the RYA PY Current List, Old List, and Limited Data lists, keeping the Current as the preference.

But I have not yet seen anyone attempting conversions from ASY PN to PY or DPN, or the reverse. I haven't done the math for the ASY PN so I have no idea if it is a reasonable idea.

 

[Rambler]

Can't help you on the ASY to PY issue.
On the new class of boat issue, I've mentioned above how we dealt with it. I suppose at club level our own approach to the owner would be..

"Welcome aboard. Nice to have you join us with your lovely new boat. As you know, your class doesn't yet have a rating, so we're going to have to make one up; judging, or rather, guessing, how you boat preforms against existing rated classes. The initial rating will be provisional until we can get more data and will probably be consertative. But hopefully after a season or two we can work out what the proper rating should be."

Normally, at club level, I wouldn't expect a problem with that approach. At a National level, it is harder to get empathy for dealing with the problem; although a statement that the rating is tentative in the absence of data, intended to be consertative, can be revised at club level and will be revised as soon as data is available, might get you somewhere [although US sailors seem to take this far more seriously than I'm used to dealing with in Australia].

But an interesting side issue is that sometimes manufacturers want their boat rated as 'faster', not 'slower' (and therefore more likely to win), for marketing issues.

It is sad to see how much the PY list has shrunk and to see so many well known classes are on the 'Limited Data list'. I suppose that's the equivalent of the Australian "tententative' designation, although for us it remains on the 'front page' so to speak.

I've also long been aware that PY has three or four digit ratings whereas we have two/three digit ones. Can't say I've ever seen anyone argue about the extra digit.

When I was trying to rate the FX and having to look at UK numbers compared to our,s I was surprised by the number of classes rated faster in one country than the other. So I couldn't just say...
"Ok, so the FX is half way between X class and Y class in the UK, so we'll do the same here" because I'd notice the UK rated Z class as being faster than Y, whereas we rated it slower (or something like that).

In short, this is very rough science being applied on very little data. In Australia we just say so and deal with it. And as far as I could see, with very little agrivation about it.

PS: Perish the thought you go to some measurement rating. What an aweful idea when you can fairly easily get a system that works well enough without that hassle. Indeed, given the range of dinghies; from ancient heavy wooden classes to modern lightweight, fast planning skiffs with wings, I can't see how any measurment rule could do better than the current system.

 

[The Q]

Years ago I had a one off dinghy and got stuffed by the " guessed at " handicap.

So when my home design boat, a mini keel boat, was coming to completion , I decided to give the handicapper some thing to start with.

So therefore there are several formula for calculating a keelboats handicap from its data, PHRF is an example, I used that and work by Professor Wolstenholme https://blogs.city.ac.uk/linstat/
To generate a handicap figure, which the handicapper accepted after reviewing the Excel spreadsheet.

Sadly those formula don't work for a non displacement dinghy, ( I tried) but one day someone will crack that.

I believe some day all boats will have a computer generated handicap, ( modified by local permanent conditions) that everyone will accept, because everything is based on various laws of science. It's just no one has cracked it properly yet.

 

[Curious2]

I merged the RYA PY Current List, Old List, and Limited Data lists, keeping the Current as the preference.

But I have not yet seen anyone attempting conversions from ASY PN to PY or DPN, or the reverse. I haven't done the math for the ASY PN so I have no idea if it is a reasonable idea.

Several ASY numbers are created by conversion from RYA yardsticks so it's clearly possible.

 

[Curious2]

A naval architect and dinghy sailer in the UK came up with the following formula on a Yachts and Yachting thread;

"I initially had a much more complex formula with things like RM to weight ratios and based the natural log of the PY number, but it did not give a much better fit and was harder to get input numbers for (wetted surface area, sail carrying power etc).

So, for boats with no spinnaker, hydrofoil or trapezes:

PY = 1800 - (144 x Length) - (37 x Sail Area) + (1.2 x Weight)

where Weight is total displacement of boat plus crew"

The graph showed a good fit for such a simple regression formula for boats of the right type;

One could then use the modification factors in the Australian yardstick to allow for spinnaker (3% faster) and trapeze (also 3% faster) if used. Obviously neither of them is very accurate but you may be able to get a decent figure and at least you can explain how its derived.

The full thread at Y&Y is here; https://www.marinebusinessworld.com/forum/forum_posts.asp?TID=11438&PN=1&title=the-fuller-number

 

[pqbon]

A naval architect and dinghy sailer in the UK came up with the following formula on a Yachts and Yachting thread;

"I initially had a much more complex formula with things like RM to weight ratios and based the natural log of the PY number, but it did not give a much better fit and was harder to get input numbers for (wetted surface area, sail carrying power etc).

So, for boats with no spinnaker, hydrofoil or trapezes:

PY = 1800 - (144 x Length) - (37 x Sail Area) + (1.2 x Weight)

where Weight is total displacement of boat plus crew"

The graph showed a good fit for such a simple regression formula for boats of the right type;

View attachment 574722

One could then use the modification factors in the Australian yardstick to allow for spinnaker (3% faster) and trapeze (also 3% faster) if used. Obviously neither of them is very accurate but you may be able to get a decent figure and at least you can explain how its derived.

The full thread at Y&Y is here; https://www.marinebusinessworld.com/forum/forum_posts.asp?TID=11438&PN=1&title=the-fuller-number

Is this old? DZero and Aero9 have PY numbers...

 

[Xeon]

Is this old? DZero and Aero9 have PY numbers...

If I remember it was 2014 . One of many at that time ( well 2012 to 2022) that involved iGRF/GRF against the rest . They all ended with him accusing the RYA of corruption and anyone that disagreed with him of being blind or stupid or a sheep. Or all three.
I definitely don’t want to go down that rabbit hole again 😀

The one thing I do know is the reason I don’t win many races in my D-zero is nothing to do with the yardstick and everything to go with my lack of skill. 😂😀

 

[Rambler]

The Bethwaites, in their various writings and books - and especially Julian - have come as close as anyone to describing the science of dinghy boat speed. Combinations of righting moment, sail area, underwater surface area and weight; with different factors predominating in different wind speeds.
But crew weight is a major factor in all this since it is a high percentage of overall weight and of righting moment. It makes it hard to turn it into ratings

 

[Prism]

One of many at that time ( well 2012 to 2022) that involved iGRF/GRF against the rest

Hang on, that’s my formula and I’m not GRF…

The Bethwaites, in their various writings and books - and especially Julian - have come as close as anyone to describing the science of dinghy boat speed. Combinations of righting moment, sail area, underwater surface area and weight; with different factors predominating in different wind speeds.
But crew weight is a major factor in all this since it is a high percentage of overall weight and of righting moment. It makes it hard to turn it into ratings

Yes, but… 95% of performance is baked in with the key dimensions.

I haven’t, and won’t, publish most of my work but I am very familiar with the Bethwaite‘s work. I am confident that I have built on what they have published, as I am sure they have many unpublished jewels. Righting moment is important, but the formula above is intended for hiking singlehanders - they all have similar RM and similar design wind speeds so sail area works as a good proxy. The formula above is a reasonable first stab if you have a design with no published yardstick. Multiply by 0.97 if it has a carbon mast, etc.

I’ve also proposed a probabilistic yardstick method. I won’t bore you with the details here, suffice to say I’m a believer in it and am actively developing it further.

 

[Prism]

In short, this is very rough science being applied on very little data. In Australia we just say so and deal with it. And as far as I could see, with very little agrivation about it.

This is the truth!

 

[Foredeck Shuffle]

Some of this good but if the boat is what is traditionally defined as a skiff or sportboat the basic numbers here are inadequate in capturing a rating. Boats with large wings are essentially trapeze boats. Some boats are over powered in light air and do not continue speeding up as much as others do until they plane, then the rating again hops rather than gradually increase. This makes creating a mean average more challenging.

It was a lot easier when most boats had similar SA/D and D/L numbers but those averages keep improving.

 

[Rambler]

It was a lot easier when most boats had similar SA/D and D/L numbers but those averages keep improving.

It was an easier world. I'm assuming you're referring to Prism's or Curious's numbers; and Prism clearly has a much better odea of the underlying pronciples than I do..

But as he (I think) acknowledges this is a very rough science. Sometimes you just have to guess and adjust if there is no more scientific method. A 'tentenative' rating.

Maybe recognise that part of why we have so few problems (at least anywhere I've been involved) is we do try and match like with like.

Our club fleet is divided into gennaker boats (RS100's to 49ers), other monos, cats and trailerable yachts. And if you had foilers you'd definitely want them in a seperate division too. If it wasn't for the division breakup, I acknowledge it would be pretty shitty racing if I had to do it every week on an around the bouys course. Apart from anything else, starting a skiff to leeward of a slow yacht with twice your sail area blanketing your wind enough to stop you breaking free is going to set back your race from the outset.

Once a year we do have an all in, up the river marathon race. The weather will very much determine the result; that's no big deal - it is what it is and everyone accepts it.

But it is all driven off a single set of ratings that do the best they can. "The best the can' is sometimes the best you can do

 

[Xeon]

Hang on, that’s my formula and I’m not GRF…


Yes, but… 95% of performance is baked in with the key dimensions.

I haven’t, and won’t, publish most of my work but I am very familiar with the Bethwaite‘s work. I am confident that I have built on what they have published, as I am sure they have many unpublished jewels. Righting moment is important, but the formula above is intended for hiking singlehanders - they all have similar RM and similar design wind speeds so sail area works as a good proxy. The formula above is a reasonable first stab if you have a design with no published yardstick. Multiply by 0.97 if it has a carbon mast, etc.

I’ve also proposed a probabilistic yardstick method. I won’t bore you with the details here, suffice to say I’m a believer in it and am actively developing it further.

Sorry I didn’t make that clear enough.

I know the formula isn’t iGRFs as will anyone that clicks on the link and reads the thread.

Sorry again for any confusion caused but the comment was about the general tone of the thread and to be honest the general tone of all yardstick threads on y&y over the last decade .

 

[Curious2]

Some of this good but if the boat is what is traditionally defined as a skiff or sportboat the basic numbers here are inadequate in capturing a rating. Boats with large wings are essentially trapeze boats. Some boats are over powered in light air and do not continue speeding up as much as others do until they plane, then the rating again hops rather than gradually increase. This makes creating a mean average more challenging.

It was a lot easier when most boats had similar SA/D and D/L numbers but those averages keep improving.

Why not pump the numbers for some US classes in and see? Large wings can just attract the same ratings hit as trapezes - there are three winged boats in the graph above and they fit the model quite well. Plugging the formula + trap and kite corrections + yardsticks into a multiple regression may at least give you a ballpark figure that will allow you to make other corrections.

The Australian and NZ yardsticks were created when there were lots of "skiff types". They aren't perfect, for the reasons you give, but they keep most people fairly happy; everyone knows that some boats cannot be beaten in some winds and others cannot win in other winds, and that's just the way it has to go.

Even with very similar boats, no yardstick system can work perfectly as well all know. Something like a Laser versus OK duel can just get down to how square the square run actually is. At least it's not like my job of trying to work out local modifications for an unusual sailing area with a fleet that includes two multiple world champs on foilers, windsurfers, wings, cats, trailerable yachts and Optimists.