Portsmouth DP-N

Jake, Why would that matter?

the old boats that have ratings with tons of data behind them will not change. You only have to run the program for new boats... and you have no data..

So... until you have a data set of 25 races with a fleet of un rated boats racing a fleet of yardsticks...... not a problem.

(I think they actually can run the program... they just can't modify the code or want to rewrite the thing for the 21st century.)

which boats are considered

yardsticks

? And what if that boat's design changes (as part of an OD rule, etc.)?

Say the F18 is the

yardstick

and they allow a modified sailplan as part of their OD rule... or a carbon stick. If either of these change the sailing characteristics (speed, angles, etc) wouldn't it move the entire PHRF table?

I think there was a blog about this issue discussing the Farr 40 along with Time on Distance vs. Time on Time systems.

Being relegated to PHRF myself (not a fan, but not sure I can convince anyone of the merits of another system) and rarely racing OD events, which system best combines both monohull and multihull fleets? I'm not even going to touch foilers at this point because I doubt I'd see a Moth, G4 or N17 anywhere I sail...

I think in my PHRF area I rate about the same as a Melges 24 and a Hobie 16 non-spin. Which seems a little odd?

It matters because if anyone puts in the effort to collect data that can't be processed they would be wasting their time. In the bigger sense, if the numbers are able to still be compiled then it would be nice to know what the committee feels is needed in order to update numbers or assign new numbers so the system can remain relevant to our sport. If they are fixed and unchangeable then it is more motivation to start looking at something like SCHRS that can adapt as our sport changes.

There IS data available and the statistics will work on a relatively small sample size because it's weighted. I could probably dig back into three or four years of EMSA F16/F18/Acat single fleet handicap racing and provide a dozen regattas of data but I seriously don't think anyone since Jamie/Darlene has had a chance of processing it. I have a Fortran background and took a stab at understand it a long time ago but the statistics are really advanced and I concluded that I needed more of a statistics background (or the time to gain the knowledge) to be able to modify it in any meaningful/reliable way.

Jake, if I understand what you and others are saying is that the DP-N system is not being actively managed since the passing of Darlene's era?

If this is indeed the case, is there a more actively managed handicap system? I would suspect if there isn't much activity on the DP-N system (or no one knows how to update it) it will eventually prove to be of little use as boats/designs evolve. Even strict OD boats often allow some changes (maybe not in Star or Etchells..)?

I don't have any factual information about whether or not that is a true statement. I do know that the system is very specialized and when I was involved, just before Darlene's passing, there was not anybody else able to run it although others had tried to pick it up and take that off Darlene. Jamie Diamond got involved at some point but I'm not sure he got it into a workable fashion either - I think he too had intended to make some modifications to it. However, I'm not clear whether Jamie was trying to just run it or modify it. I looked at it for a brief period and quickly learned that it was over my head in trying to modify it but I never tried to compile it as it was. The Fortran compiler was expensive at the time but it looks like you can get free compilers now. Darlene eventually took it back up in a limited fashion. I do not know of anyone that has actually compiled numbers using the actual system since Darlene but I have not been involved in several years.

The software (Fortran) is tricky to run and requires a good deal of specialization with old programming / compiling software. My experience with Fortran originated in the time when it was running on IBM mainframes and you would interface with a command line green monochrome terminal station and had to schedule compiling time with the university. Someone on the committee today may well be able to compile the software and statistics as it stands but it's not something that just anyone could pick up and run with. It's definitely not an excel spreadsheet. 😉

I’d stop harping on USSailing if they made some positive changes that actually supported the sailing community. The fact that you defend DPN committee’s lack of action is indicative of what USSailing is all about! Then to defend the DPN’s committee lack of action as maintaining the integrity of the system when they clearly and by your own admission willfully violated that integrity leaves me scratching my head.

Please don’t put words in my mouth or tell me what I’m thinking! I never put any value in DPN results I clearly understand what the results represent and know they aren’t worth the paper they are printed on. So DPN, SCHRS, Texel it matters not! I just draw the line at completely fabricating numbers, geesh! Please try to pay closer attention it’s not the first time I’ve said this.

Because there is NO integrity with the DPN system the fact it is USSailing sanctioned means nothing, you only have to read your own words to see that statement is true.

Of course you support USSailing telling me or anyone else to go pound sand, that’s kinda your thing.

Only you would tell someone in the sailing community who you pretend to represent to piss off. This position invites the community to move on without USSailing.

I’ll stop whining when you stop being a sock puppet for USSailing. No that's not true, I'll stop whining when I stop caring.

Dave.... you are thinking about a rum....and kitten videos...

The US Sailing DP-N system is dead to me.

It's dead because there is no data.

The fact that most of you refuse to join, then publicly rant about no support doesn't help your case.

No one, and I mean absolutely no one, at US Sailing cares if you use a different system.

We've had these debates for eons here. A standard system is necessary for credibility. No one has crunched numbers to prove that there is a better system.

If one of the other systems is better to use, and has the support of the fleet, I will fully support it as the MHC chair.

Mike

We have (I have personally), on multiple occasions, compared the systems with number crunching. Regardless, it's all handicap racing - it's never going to be perfect. It's a comparison of

worse

or

less worse

.

Darlene bless her heart was easily manipulated by a few grudging sailors back in 2002 to aid their scores in the numbers by having her implement false to no-logic theories to be integrated in to the US Sailing DP-N System. That was my problem then but now it looks like the new numbers have been corrected.
As I recall the Portsmouth System went downhill back in the ‘90s when it was taken over by US Sailing from the guy out in the west when he just got tired of running the system. I remember speaking with him back in 1989 when he was beyond ready to quit.
Then after the heyday of big numbers of Cat sailing races and participants in the 80’s and 90’s there was never enough good data to support the fundamental theory of computation used in the DP-N system.
Since the DP-N method using limited race result data and results from sailors with non-uniform abilities really never has been a reliable method (Garbage in Garbage out) I think a more Boat Measurement based configuration calculation should be used.
The new analytic calculations being used by the Americas Cup guys could probably tell to a whisker how a change of one half millimeter to any parameter or angle could predict a change in performance of the boat not the sailor. Make the boat numbers to be equated not the sailor’s abilities. The abilities should be the score result of a race.
That is why I like One Design Racing but since we have to use Handicap Racing sometimes because this is a dying sport and there are many configurations of boats that want to race at least US Sailing or someone should look at incorporating new technologies to equate dissimilar boats performance. Let only the sailor performance dictate the race results.

Lol. If there is one thing that Darlene wasn't, she WASN'T easily manipulated. That woman was sharp as a tack and wise to any alternate motivations.

Terry

Herb Malm was the west coast fellow and his system was called NAMSA.... which was decidedly NOT portsmouth. NAMSA used his proprietary algorithm that he would fine tune with race data. Every time he got some race data.... he would recalculate his PN Table.... and presto... your ratings would change. You might remember the large NAMSA handicap table posted at Wroten Point Yacht Club, that was a copy of one of his versions.

In the late 90s... the problem boats were Dart 18s. Darts Like the F16s had two configurations one up main only and two up sloop. Darts were struggling to meet the fleet metrics for US Sailing Portsmouth inclusion since the fleet was more recreational in orientation.... while Herb would happily work with Brian to generate a fair rating since he had his prediction equation. Eventually catamaran clubs decided to switch from NAMSA to US Portmsouth. As you say... cat racing declined rapidly in the 90s and to to move forward and attempt to serve the public, Darline sought other ways to include boats in the tables without all of the criteria required by the PN Standard. So..Darline and the Committee (did my best as did bob) and made the best of the situation to serve the cat racing community.

In 2015, the europeans have the lead in using measurement based system that fine tune the math with race data. SCHRS and Texel. I doubt Oracle is going to solve this problem.
Unlike Herb's NAMSA system... SCHRS and Texel are completely transparent (formulaes are public) and they are managed by a public committee). SCHRS uses English as the base language.

US Portsmouth has decided to return to the letter of the standard. In their view its better to have the tables serve the OD classes that meet the criterion with out compromise. The committee could choose to PHRF estimate the ratings for new designs and include them in the table but is choosing not to. So... if you are racing Hobie 16s against Hobie 18s.... no worries... If you are racing N17s or Carbon20s.... it will be impossible under US Portsmouth.

So... what goes around comes around... Our choice now is a measurement system using a formula tuned with race data!

OR US Portsmouth Somebody will have to champion a fleet of N17s racing other boats including lots of yardsticks to have enough data to get a statistical rating. It may not be an accurate rating however since there are lots of other assumptions made by the PN system that may not be currently valid. (to reprise... garbage in... garbage out)

Bottom line... its time for the US cat fleets to move on.... fleets with the new boats coming to play should plan accordingly.
Part 2,
Sailors who want to improve the OD configuration should just get their stuff measured and use the calculators to generate your specific SCHRS rating.

it's all handicap racing - it's never going to be perfect. It's a comparison of

worse

or

less worse

[.quote]

MOST of the big boat racing in the country happens under PHRF Handicap racing.... NOT one design. Catamarans need a viable handicap system just to have a race these days.

What is the purpose of characterizing the racing as

worse

or

less worse

Handicap and One design involve tradeoffs... if an organizer leaves a rump group of disparate boats in handicap fleet ... chances are they go home... don't get a one design boat and just day sail. So... choices to be made.

Now picking handicap systems. also are not a matter of worse and less worse..

You could compare them on accuracy
You could compare them on the wind ranges where they are accurate.
You could compare them on the transparency under the hood of the tables.
You could compare them on how quickly they can generate a fair rating for a new boat.
You can compare them on the amount of effort needed to maintain them and their governance.

ah... choices!

 

I'm fine with the flaws of handicap racing (i'm just there for the bar, anyway... you've seen my sailing). What I posit is using a handicap system that is actively being managed/updated.

I get the feeling that DP-N may not be as actively managed as other systems for a number of reasons (regattas not sending info, boat/classes too small, no one knows the system, etc).

Based on the actively managed criteria alone, which handicap system is most used?

P.S. I am a current USS member.

I'd also venture a guess that designing or updating a handicap system (for any sport) would make an awesome statistics Ph.D. dissertation.

Surely there are some statistics genius types in the global sailing community?

And with the ability to generate VPPs on just about any boat (pretty soon I'd figure someone will make a smartphone app and crowd-source that), it might just help generate the needed data with statistical significance?

Since they are using various platforms for the Hobie Alter Cup, do they digest any of that information for use in the DP-N calculations?

There is a fundamental disagreement in handicap world between stats guys and measurement guys. The measurement guys will argue... If you had perfect data betweeen two fleets... One of them the N17 where the sailors are practicing 200 days a year and the F18 fleet (Yardstick) where the sailors are practicing a whole lot less... this basic asymmetry will skew your stats so that the N17 rating will be too fast. In the real world....for other classes, a rock star sailor competing in a carboard box will beat the top rec sailor on his 5 year old boat. The stats are screwed! The stats guys will say... Hey... you are trying to model a nonlinear performance curve of classes that span 4 generations... No chance in hell!
YMMV.

A VPP application does not give you a ratings table.... it gets you polars to sail to for your design. VPP approaches inform handicapers who create rating sytems like the latest failure (HPR) which lasted for a few years before the owners decided it was not for them. So... no pot of gold at the end of the VPP rainbow.

The Alter cup... is just a one design regatta. No other boats are on the course to compare with. So... it is of no use to the stats based guys or the measurement guys.

Sure, but would it not give information on the actual performance of the design platform? Since you're rotating crews and it is a race under somewhat controlled conditions (how the course is set, etc). I would figure no one is really sand-bagging at Alter Cup.

But I suspect using this information would require the RC to take times, rather than just score finish positions?

I am happy to set my goal as sailing to my actual PHRF number or VPP speed. This would tell me I'm pretty flawless at turns and straightline speed.

http://www.schrs.com/

Here yee, here yee, here yee. Look who go it right!

A Class Classic and A Class Flying. BRAVO!!

MarkS I finally agree with you, SCHRS is the way to go!

Bob

I'm not sure what side he is arguing for anymore...but he did make some valid points about Portsmouth being a set in stone kind of rating system. If that's truly the approach of those currently running the Portsmouth tables, the un-flexible nature won't suit any new (or small) classes, so I'm in favor of scooting over to SCHRS.

One thing that the Portsmouth system seems to have going for it is different numbers for various wind conditions. I imagine all boats are non-linear in their wind-speed/boat-speed performance. Some more than others. Seems like DPN captures that - at least empirically.

So my question - are there any predictive systems that accommodate this?

Thanks Jake.

I think the one thing that we can all agree upon is that we need a handicap system that is fair to the sailors, and easy for the organizers.

If we want to propose to move to a new system, it would be good if it had a champion, if not an administrator. Not in the number management sense, but in the sense of helping OAs use it.

Having recently tried to help Craig and some others get numbers for new designs, I agree that DPN is not working in today's environment.

In my experience with US Sailing, nothing happens without a proposal, and even less without an actionable plan, with leadership. Wildly demanding change, then sitting back waiting for someone else to solve the problem, falls upon deaf ears.

Mike

First I want to thank Brucat for helping me try working with Portsmouth committee. I have also been chatting with Mark in the background trying to see if SCHRS is viable for Sail Series. Below is the only issue I see with switching and thought this would add to the current discussions.

The only problem I see with SCHRS is getting individual numbers for modified boats which is fairly common. You would need to submit the changes to get the correct new number. To even make this more difficult, you would have to get sails measured etc... To compound this, a lot of sailors don't register and just show up which would make getting them a number impossible. It would also require a lot of work for those running the ratings system on an individual boat basis.

The only way I can really see this working accross all races would be to have some sort of temporary or even permanent correction factor similar to what Portsmouth uses. Any input would be helpful.

You needed to stop at nothing happens. If you have a plan and a proposal it gets shot down by people like Mark and Mark has many like minded friends at USSailing.

Isn't the DPN Committee chair a leadership position? Why would it take someone from the outside to do all the heavy lifting and solve all the problems? How about getting the folks that agreed to take the position do something other than keep the seat warm!

Wildly demainding change? Really? Asking that a system be maintained and numbers not be pulled out of thin air is wildly demanding change!? Sorry, I didn't realize I was being unreasonable. Please carry on doing whatever you where doing...

Oh and I'd be happy to take point on this but you know there is that whole “We're going a different direction” thing so...

Just because you publish a table with a rating for a N17 at Beaufort 4 does not mean that you have data (100 qualified races of a fleet of N17s racing against a fleet of yardsticks, (Hobie 16s and F18s). However a table with ? Holes is not useful. Behind the scenes, all race data are tabulated by wind speed (each Beufort number) and boat class. The software calculates the DPN as a weighted average of all the windspeed ratings and I think amounts to 70% of the B4 rating. The wind tables were generated with fixed percentages when needed because a table with holes is useless.
The committee split out B4 ratings for multihulls and not dinghies for a reason. Newer multi designs flew hulls much earlier then older designs. Of course dinghies are either on plane or making a big hole in the water from B4 on up. Now, the latest cat designs are even more efficient and some boats are flying hulls in B3. But Once again... you need a TON of data to have a valid ratings table... and that just doesn't exist.

Second point. Handicappers have found that Users, both sailors and organizing authorities hate the complexity of multiple number handicaps. Often, their favorite scoring program won't easily manage a matrix of ratings. Race committees are not good at determining a SINGLE Average windspeed for a 45 minute race... (oh... the wind was 6 to 12 over the race... well... that data can only go into one bin... B2 or B 3) Only the good pro's record this data during a race to have an actual objective basis. .. Texel published two ratings for their table for one or two seasons; Trapezzing and non trapeezing conditions. Most euro clubs did not bother and used the single number rating. So... they went back to managing the wind speed with a single number rating.

Next point, the precision by which two boats are scored on the race course is very different between one design and handicap racing. You can win a OD race by a boat length... In a handicap race... the difference between them is lost in the noise and they would be tied on points. That's why it is silly to use a timing system with 10ths of seconds measured and you truncate the rating.

Moreover, Handicap ratings can't resolve finish positions any finer then the noise added by the racing rules of sailing of the game. So, take a fleet of 10 identical boats 30 feet in length. 10 get perfect starts on a square line. One goes off on port and the rest on starboard. At the weather mark... 9 30 foot boats will have right of way on the single port tacker. that means he will be 270 feet behind... depending on boats speed, x amount of seconds behind... his sailed rating will be slower then the other boats. Bottom line... He sailed perfectly but NOT to his class rating... he really should be tied for first in handicap world and he is actually last on the water. Basically this looks like noise but is really just a bias in favor of boats on starbord. It takes a lot of qualified data to average out all of the sailing noise from the bias in the data and you still can't get around the bias for starbord around the race track. Bottom line... handicap is not as precise as one design and determining a rating table can only be an approximation with limited resolution.

The good news is that most racing does not require that much precision. Even in a one design race of Olympic sailors... Identical boats will finish 10 minutes apart on the race course... translation the winner sailed their boat to their rating... eg 65.0.... the last place boat sailed to a rating of 80 or higher. Rut ro! ... You can't expect more from a handicap system rating different designs.

Measurement systems take wind speed into account with their righting moment factor. (I don't know the specifics off hand).
DPN Portsmouth takes wind speed by biasing the rating to the B4 condition.
Portsmouth Tables of ratings for windspeed attempt to manage windspeed directly .. however, YMMV as to how good the basis of the table actually is for your fleets.

Personally, I used to think that wind rating handicaps help the game of racing because every sailor has a story when at some windspeed a boat class just flies a hull and leaves the fleet behind .... The idea is that the RATINGS should reflect this reality. One more knot and the fleet can fly a hull and ratings work again... The wind ratings were an obvious recognition of this realty. I liked using wind ratings to pretend to solve the issue AND I thought the time was coming to have rating for B2 and B3 split out.

When I relaxed a bit and take handicap for what it is... we are probably better off with a single number system that manages the non linearity internally. This is the approach of SCHRS and Texel.

Craig,

For the FL300 the modified boats won't be a problem. Measuring sails is not difficult and we aren't talking an entire fleet... one boat, correct? Please don't let the unknowns spook you. Psssst.... it's your freaking regatta you can do pretty much anything you want. Isn't there a regatta in Holland that created a handicap system specifically for their regatta?

It's time to move on DPN simply isn't the way to go anymore. You want help cracking this nut let me know. If you want to use Steeplechase as dry run we can do that too, the fleet will probably be identical. Dude we can do this!

It's time to solve this problem or at the very least make it less bad.

Dave

Craig... Every other form of handicap racing puts the responsibility on the owner to get their damn boat measured for the race. Only beach catamran world tolerates a bunch of beach bums who show up on the day of the race with a boat and expect to race without prior paperwork or registration.

Yes, it is more trouble to get your changes measured before you go racing. ... IMO... since you made the changes.... you own the responsibility to the fleet to get a fair rating.

There are people who measure and it costs you a fee to have your boat done. You also have the option of being a corinthian sailor and measure your own gear and calculate your rating.. Corinthian Sailing runs on the honor system.

Still, It is the OWNERS responsibility... not the OA or the person running the US ratings service.. At best... his job is to hunt down the owner with the measured changes and a one of a kind rating and make sure a copy gets sent to ISAF SCHRS or Dutch Texel.

I am not

spooked

, that is why we are discussing now. Not just considering for Florida 300 but also Hiram's Haul and recommending to other distance races in the series.

I need to put together a proposal for the Sail Series board to approve and you know some of the board hates change. Most of my time before and during the event is tied up with website and content. If you are willing to help us implement, it might be the factor to swing votes.

Mark, here is an interesting case:
Hans Geisler has shown strong interest in the full Florida 300 for next year. His G-Cat 5.0m has a square top main and just ordered a reacher type sail that will not use a spin pole but be attached to the front bows with a traveler system to switch sides. What would we need from him to get a rating?

FYI: Sail Series Promotions is a paid US Sailing member in good standing. I was willing to help with providing results from all Endurance Series events but haven't receved a reply from any of my emails this year (except Brucat). I also recall distance races are not applicable for portsmouth.