New 2006 Texel numbers just out, the A Class now rates the same as the F18's.

What changed in the formula to make this happen?

It's not a yardstick system!

Link to the numbers ?

http://www.watersportverbond.nl/data/DC9_numdet1_26-2-2006.pdf

Last year Texel had two handicap numbers per boat (and an addition 2 for the optional spinnaker). Turns out Race Committee were to lazy to use then and so just decided to use the light weather rating for all conditions. So the double rating numbers was created to be more fair to singlehanders and boats optimized for either light winds or big winds and then next thing the RC's do is simply ignore it all together.

That didn't work, so Texel committee decide to just calculate the average of these two numbers and have that as a single rating for each boat. So now the A-cats have a peachy handicap in light winds and an equally disadvantagious handicap in the rough stuff (for example) and the RC's can't ignore it anymore.

In addition I think Texel 2006 saw a change in the way spinnaker benefits are calculated. These are now considered to make a boat more fast then before.

It is not perfect; but it is a good compromise.




That is about right, in my opinion.

I'm sure some wll bark at this, but then again alot of A-cat sailors seem to steer clear of open class regatta's for some reason, so how do they know ? In the results I've seen over the years in open class racing the F18's are in general about as fast as A-cats. They (the A's) have an advantage in the really light stuff but are slower in the heavy stuff as well. So on average they are about the same to F18's.

Wouter

How do you read Texel vs portsmouth?
Ex. hobie 21spi 95 texel --65 portsmouth
Ex. tornado spi 94 texel --59 portsmouth

These # don't match up ?

matt

texel isn't based off of collected regatta results.
that tornado rating is courtesy some of the hottest olympic sailors racing them all over the country.

... And some goof balls who still sail the Hobie 21s.

In my opinion Texel is right in this respect and US portsmouth is wrong.

Another example is the Supercat rating versus the I-20 rating and we all know how the serious US racers feel about the Portmouth numbers in this case.

Wouter

Actually, any system based on data is superior to one derived by formula. With data based system, the system is continually getting more accurate, as opposed to one which is based on a formula, and represents the potential of a boat of a set length, weight, sail area.

The Portsmouth system is based on bouy racing. If someone applies it to races that are something other than bouy racing (e.g., Steeplechase, where the Supercat competes), then it is a mis-application of the system.

Wouter, some of your posts on the F16 forum stated that a single-handed F16 could not be expected to compete with a 2-up F16, however, the Texel rates the F16 1-up w/ spin faster than 2-up w/ spin. Is Texel right?
Have things changed?
Is it a fair contest for F16 1-up and 2-up to race head to head, first in wins?

I beleive this is where the US F16 class wants to head. In tradewinds we did race first in wins and I (2up) got beat by a 1up. So it is proven that a 1up can definitley hang with 2ups.

So far that I know of
Tradewinds 06 we raced first in wins. No DPN corrections.
GYC April 1 and 2, they will race first in wins, again NO DPN corrections.

I hope that this is what we will continue to do and sooner or later get the DPN numbers to match, so there is just ONE F16 rating.

Again this is just a trend, and nothing is official.

Les,

I'm sorry to inform you that I couldn't disagree with you more. That and the fact that I'm bewildered by you attributing statements to me that I never wrote down or would write down. You whole post is simply incorrect. They contain great one-liners, but incorrect ones just the same. Sorry.




You are wrong here. Both systems are based on data, so they can't be discriminated on this property alone. Texel however gethered data of all boats and sailors to establish a basic describing formula which is then used to rate all boats while Yardsticks use the individual datasets for each boat/crew and hopes that these datasets are unbiased and unbiased in comparison to the other individual sets. It is a very well documented fact that these hopes (assumptions) are NOT reflected in real life. Therefor the Yardstick has a huge fundamental flaw in it that ruins everything. Any person schooled in probability and estimator math will be able to explain this to you. Yardstick systems are very sensitive to what we call outlyers. Especially when there is only a small data set for a given boat (+new classes, dead classes, or classes with a small active race fleet) In real terms the identifier outlier refers to guys like Randy Smyth, Robbie Daniels and Charly Ogletree. The advantage, adn yes superiority, of Texel is that ALL data is gethered in one dataset and used to make make only 1 describing function, the texel formula, which is then used top predict all ratings. Outlying values have far smaller effect on the overall ratings this way and often offsets that find there way into the formula are propogated to ALL ratings which are then cancelled out to one another because all ratign usage is is relative. This all makes Texel alot less sensitive to biased data and makes it also alot more stable. Example : Supercat 20 has been stable under Texel fro 20 years; Under yardstick it has gone from faster than (standard) tornado to ALOT slower then (standard) Tornado simply because the guys sailing the Supercat saw the good skippers change classes.




That is just nonsense. Again a great one-liner but without a scientific basis. I can give plenty of examples were such systems do NOT become more accurate with more data. In soem cases you can even see them become less accurate. One example is of the Prindle 19 ; Raced hard at first, then class dies, then got briefly revived because of Randy and died again. We now have a data set that contains both very fast data and very slow data, yardstick averages this into a rating number that doesn't reflect both situation with high accuracy.




To this part I agree. It is the only part.






I never said that or wrote that. I don't understand why you think I did. From the very beginning the F16's were designed to be comparable in both setups; if I thought that they weren't then I would have redesigned the setup.





It is alot more right then the US Yardsticks, I can tell you that. US PN have us 4% to 5% slower then the F18's. Texel shows us to be in the 1% band around the F18 rating (is often just the rounding off offset). It is my firm believe and personal experience that F16's are very comparable around the course to F18's. Even the Australian yardsticks put them there. It is only the US portsmouth that has been "unwilling" to converge to the right number for years now. And I talk about "right number" because we have such much data now that underlines the correctness of the F18 equality. The US PN numbers are just way of base.




Texel was always more accurate then US PN, I'm sorry, despite the fact that for a while Texel wasn't too accurate itself. Now I feel Texel 2006 (and Texel 2005) are accurate in the absolutely sense as well. With this I mean the true performance is within a 1 % band of the calculated perfromance (rating). That is pretty big achievement as there are so many variables.




Yes; in general I beleive so. With this I mean that 2-up vs 2-up and 1-up vs 1-up is more accurate then 1-up vs 2-up but event the last contest will be accurate with a margin of 2 % at the maximum. Which is accurate enough for 90 % of the racing. Meaning that 90 % of the racing won't see a different final listing even when taking such a offset into account.

Wouter

YES! What bothers me about the Portsmouth tables is than when certain new designs (Taipan F16, for example) arrive on the market they “define” several configurations and get several assigned numbers. My issue is that this flies in the face of the main doctrine of a main table with modification factors.

The modification table says, “Sail lighter and you go faster, take off a jib and sail slower.” Starting with the Taipan 2-up sloop with jib and spin let’s calculate the uni/spin rating. Taipan class rules have no class weight so when sailing uni you MUST use L4 (see new Portsmouth L4 rule) . With this in mind, take the Taipan F16 sloop with spin at 65.3 and apply the L4 light-weight modification of 0.97. Then use the US modification of 1.026 for taking off the jib. This gives you 65.0 for the uni, slightly FASTER than the 2-up sloop (25 seconds /hour).

However, the “special” rating for the uni has been assigned 66.4 which is 2 minutes pre hour SLOWER than the sloop! Why should it get a “special” slower rating when Portsmouth provides for a modification calculation that say it's FASTER?

65.3 x 0.97 x 1.026 = 65.0
Table value = 66.4 ???

The Mystere 4.3 came out with 4 ratings when it was new. It’s down to two. It should be ONE and then use the modifications as EVERYONE else must do.

If you sail one design, do whatever you want. If you sail Open Portsmouth everyone must use ALL the rules ALL the time. Otherwise, it’s just typical Florida BS politics.

Perhaps Portsmouth politics are like the quote from Animal Farm, "All animals are equal, but some animals are more equal than others."

This is exactly what the other systems are showing and what appears to be true ont he water. Mind you those 25 second/hour is a pretty small difference (0.7 %) and less then 1 % = 36 second/hour. I call that pretty comparable around the course.

For the record : The F16 class officially strives for a reduction in rating numbers and even for ratings close to the F18 class. It is the official standpoint of the F16 class that US PN numbers for F16's are too slow ! We believe this to be unfair to the other sailors on different designs.

Wouter

Sorry sparky, for one of the few times I have to agree with Wouter.

A measurement system is far better than one based on returns.

The critical point is that the returns system (protsmouth) also measures the skill of the sailor - consider

Class X is sailed by many very good sailors in active fleets around the world and so is sailed very well; however they do very little handicap racing, those that do handicap race are those at the tail of the fleet - The portsmouth rating reflects this because it scores where the boats finish, and this MUST be a function of the skills of the crew(s).

Consider Class Y that also has a very good fleet of sailors, but they also do many many handicap events and so the Portsmouth for class Y is a "hard" handicap as the boat and sailors have performed well at events.

Consider class Z. Not very active, and those that do sail sail Portsmouth events and finish at the back of the fleet; over time the handicap of class Z gets more favourable; then Hotshot A comes along, sails Class Z well and cleans up.



This is why Portsmouth does not work.

Wouter,

“I always valued the versatility of the F16 in the way of 1-up and 2-up sailing as the most important. I'm actually quite convinced now that we've done the right thing and made it work out on the water as well. Within limits of course as some difference will always remain.” Wouter

“But it is proven the sloop is a faster F16.” Robi

“I'm definately faster sloop, and I believe sailing uni is more challenging to do well.” Eric

“Of course a singlehander is always disadvantaged in the really strong winds” Wouter


These are a few quotes I found with a brief search of the F16 forum, however, you have posted so many times of such length that I must admit, it is by no means an exhaustive search. I just don’t have the time to read all that you write. I included some things that were in the forum by Eric and Robi (your lack of a post to disagree made me think you agreed, based on your past practices) so when I stopped reading the F16 forum, I left with the distinct impression that F16 1-up and F16 2-up were hoped to be equal, but that it is an impossible task and there will always be conditions where one is favored over the other. You have not convinced me otherwise, but go on and try. I expect this and am sure there will be great entertainment value for the forum.


Your problem with the yardstick and the probability problems you state are true with small sets of data, however, Portsmouth collect a year’s worth of data and only change the numbers when the data is compelling. The issues you have are minimized by this process. Any person schooled in probability and estimator math will be able to explain this to you.

Wouter, you posted “Example : Supercat 20 has been stable under Texel fro 20 years; Under yardstick it has gone from faster than (standard) tornado to ALOT slower then (standard) Tornado simply because the guys sailing the Supercat saw the good skippers change classes.”

You neglected to include that during this timeframe there were other factors that also affected the relative numbers, like the use of a stiffer Tornado platform from Marstom, the rig development and sail development that has been seen in the Tornado Class, even before the addition of spinnaker and 2nd trapeze. The competitive fleets and best sailors in pursuit of Olympic glory have all made the Tornado faster over the years. Since Texel uses ALL the data, it must also reflect this change in the speed of the boat. I contend that the data submitted to USSA and the Portsmouth number correlate. This is what is seen in the fleets using Portsmouth and serves those fleets adequately.

And you said “US PN have us 4% to 5% slower then the F18's.” Portsmouth is reflection of the data. Data is not wrong, it is data. As more sailors race F16 in the US (admittedly, a VERY small number so far) true capability of the F16 and F18 will be reflected in the data. For the time being, the data says that the F16 is slower than the F18, and the F16 1-up and F16 2-up are not equal. This is only fair to those who sail those boats.

Scooby, you proposed:

“Class X is sailed by many very good sailors in active fleets around the world and so is sailed very well; however they do very little handicap racing, those that do handicap race are those at the tail of the fleet - The portsmouth rating reflects this because it scores where the boats finish, and this MUST be a function of the skills of the crew(s).

Consider Class Y that also has a very good fleet of sailors, but they also do many many handicap events and so the Portsmouth for class Y is a "hard" handicap as the boat and sailors have performed well at events.

Consider class Z. Not very active, and those that do sail sail Portsmouth events and finish at the back of the fleet; over time the handicap of class Z gets more favourable; then Hotshot A comes along, sails Class Z well and cleans up.“

First, Portsmouth compares the speed of all types of boats, not just those that race Portsmouth. Your description of Class X and Class Y still result in the same Portsmouth number because all data from all boats is submitted. The only exception is that the data comes from USSA sailing clubs and regattas, not from worldwide. Portsmouth is a reflection of what is happening in the fleets using Portsmouth, i.e., it services its users.
Class Z is truly being reflected by Portsmouth numbers, and I believe that is one of the things that make Portsmouth better. If a “dead boat society” ends up with a more attractive number over time as a result of the data, that makes it an attractive boat for those new to racing or cat sailing. If they like winning, they will recognize the best way to get better is to race one-design in large fleets, which many consider the best way to “improve the nut on the end of the tiller”. If a great sailor decides to win the Portsmouth game by using a boat with a great number, then they only get away with that for a limited time, as the number will continue to reflect the data. To win by getting the best Portsmouth number boat is a hollow victory and temporary. Anyway, the great sailors seem to gravitate to the best competition, which is what seems to be happening in the US F18 fleet.

Wouter and Scooby, et al: I do not think any handicap system is flawless, however, I believe that the US catamaran racing fleets are best served by using Portsmouth in its entirety, including wind adjustments and the other HC adjustments for changes to the one-design boat. Possibly, Texel, ISAF, SCHRS, et al, do best at serving their catamaran fleets. I am not the right person to judge what is best for you, and visa versa.

Les,



The data will not be wrong but the processing method or the intepretation of this data may very well be. My argument focusses on the latter part. Here Yardstick systems have a well documented fundamental flaw. Texel uses what is called a parameter driven/describing fucntion system identification method. Under a certain set of conditions it can be proven that this later methode is more accurate and less sensitive to outliers then a straight forward statistical approach (like yardsticks are).

It will be far to complex to get into detail. I will try to think up a few examples to make thing clearer. If anyone is really interested then I advice then to pick up some text books on system identification.




That is not entirely the case. It is more an intepration of what the data is showing. Apply a different processing method (in developping the ratings) and you'll can get a different intepretation (reflection).





This is a different topic all together. This is a seperate issue from whether the used rating system is more fair (accurate) then another. This one is best in serving the needs of a given community is not hard science while accuracy is.

While I absolutely certain of the superiority of Texel 2005 in the area of accuracy, based on hard science, I can't be that in the way of which system serves which community best. We should take care to keep these two issues apart.





I know this to be nonsense. The F16 rating hasn't been converging under US PN for years now, despite the gulfport guys and gals sending in all their data. Some would say that that is because the Taipans+spi are really not that fast. But that would be really funny as the Australian yardstick system VYC is rating the Australian Taipans WITHOUT a spi faster as the US Taipans WITH a spi (relative to the F18's). They can't both be right at the same time. This is actually a great example of how much offset can be propagated into yardstick ratings.

When sailing under US PN I would have been in direct competition for the title of club champion and I'm really not that good as sailor, we have a few ex European and World champions racing in our club races. The US PN ratings for F16's are seriously off the mark and I know that from personal experience.

I'm also very happy that Chuck is putting up wagers that he and his wife will beat F18's to the beach, no use of peachy PN ratings. In the first such comparison they actually put deed to claim and finished ahead on elapsed time.

Gary Maskiel and others are doing the same. I'm sorry to say but the US PN rating for F16's is just wrong. But then again it is wrong to the advantage to the F16 owners so we will not complain to hard about it.


Regards,

Wouter

I think I have a great example of where the true system is not accurately reflected by an generally understood statistical model.

Say we have a dessease among the populace where 1 in every 100.000 people IS sick. This is the true situation which in fact we do NOT know but try to discove using statistics.

In a lab we "proof" a test that is 95 % accurate in establishing "Sick" over "not sick" when testing a truly sick person, one that was initially diagnosed as possibly sick because of some sympthoms.

Now the government with its pseudo scientists decided to test the whole populace and see what they have on their hands.

A populace of 10 million people is all tested using this 95 % test. Say you are also tested and the test says "Sick" in your case.

Question : Do you now need to worry and take immediate treatment ?

What do you think ? Answer truthfully !


...


Proper developped probability math will show that there are only 100 truly sick people in the population of 10 million while the populace wide testing will earmark no less than 500.000 people as sick !

So 499.900 people who are tested sick are AREN'T truly sick ! They are just diagnosed the wrong way.

In effect you (as a diagnosed sick person) will have only a 100/499.900 = 0.02 % chance that you are TRULY sick and need to worry and get treatment.

In effect the government just wasted a whole lot of money to test 10 million people, creating upset over patently useless test results.

But how can this be. The test data that confirmed the 95% accuracy of the test itself was correctly developped and "data doesn't lie" ?


When wrongly applied the intepretation method of a truthful data set often DOES lie. Meaning that it gives a totally false picture of what is really going on. In this case the dataset used to proof the test isn't to blame but the processing and application method is.


Along the same lines one can compare the Yardstick and Measurement based rating systems. It then turns out that yardstick systems are more sensitive to flaws then the parameter-model driven systems like Texel where the develloped model is checked against the entire dataset for validity rather then created by them through some statistical processing.

So one must be really careful with too simple notions in the field of statistics and the processing of statistical data.

Wouter

Hi all,

...it is 15 mph out of the east...


...seas are a nice 1-2 feet with small whitecaps..



....the air temp is 84 f.......and nice blue sky with a few white puffy clouds...


..the ocean temp is 78 f -ish and one can see down about 30 feet of various shades of blue...


..why don't you all come down and go sailing?


...you can use my boat and we can line up some other ones too


...afterwards, there will be a cooler of cold ones and we will cook on the beach,....using Texel and Portsmouth tables to start the fire!


...just wanted to put things in perspective..






regards,

Bruce
St. Croix
US Virgin Islands
x17