Print Page | Close Window

New PY Website

Printed From: Yachts and Yachting Online
Category: General
Forum Name: Sailwave Support
Forum Discription: The place to ask questions relating to this popular sailing results program
URL: http://www.yachtsandyachting.com/forum/forum_posts.asp?TID=9151
Printed Date: 15 Dec 17 at 4:35am
Software Version: Web Wiz Forums 9.665y - http://www.webwizforums.com


Topic: New PY Website
Posted By: AlexM
Subject: New PY Website
Date Posted: 27 Mar 12 at 12:18pm
I thought Iíd start a new thread for those of us using the new PY online returns system. How is everyone getting on? Iíve managed to upload all the results from last year and this yearís winter series but I donít think Iíve got enough data yet to propose PY changes (Iíve got about 146 races on with races varying from 10 starters to 50) because the CF is still at 0-0.1.
Main things Iíve noticed
ē     When uploading the results it does take time to appear on the site (old site was near enough instant) be patient :)
ē     Iíve not managed to upload any results from ďselect file sourceĒ only using the plugins on Sailwave works for me
ē     Iíve uploaded some results which I want to deleted. I use the option to delete them (and they do disappear) but they reappear when I come back to the list.
ē     Reports screen Ė Iíve viewed this on different PCís (Iím not sure if this is relevant) but I seem to get different results on each on. On my Home PC Iíll get two lines for each of the classes, on my work pc Iíll get only a few of the classes and on my ipad it looks fineÖ lol
ē     Is there anyway of seeing the results for all the classes (like the old return page)?
Alex



Replies:
Posted By: AlexM
Date Posted: 27 Mar 12 at 12:27pm
This is my current report page:




-------------


Posted By: AlexM
Date Posted: 27 Mar 12 at 4:35pm
Resolved pt 4 above - press the floating refresh button :)

-------------


Posted By: Medway Maniac
Date Posted: 27 Mar 12 at 6:58pm
That report page would certainly make me hesitate to use the suggested PY's, which suggest inter alia that a Laser is about the same speed as a Phantom.  

Effectively, your sailors would be getting 'group' personal handicaps, the (apparently better) Laser sailors being penalised and the Phantom men on a free ride.  A bit demotivating for the Lasers, and unlike a true personal handicap, they can't even boast about how low theirs is!


-------------
http://www.wilsoniansc.org.uk" rel="nofollow - Wilsonian SC
http://www.3000class.org.uk" rel="nofollow - 3000 Class


Posted By: AlexM
Date Posted: 27 Mar 12 at 7:25pm
I think this is where the confidence factor comes in (0 Low 1 High). As you can see none of our results are ready to used yet (wasn't suggesting this, just showing how it work's)
The results so far are pretty spot on for the Phantom /Laser class as our Phantom fleet is all very new (dare i say it, beginners) and the laser fleet are very good. However I think this is where the club judgement also comes in for the ability of the sailors and wouldn't change the Phantom H'cap
Alex

-------------


Posted By: Granite
Date Posted: 27 Mar 12 at 9:46pm
I am sure that the new site is great, but I don't seem to be able to login anymore. Have they changed the login details?
I would try asking them directly but the contact us link still does not work.


-------------
If it doesn't break it's too heavy; if it does it wasn't built right


Posted By: AlexM
Date Posted: 27 Mar 12 at 9:54pm
Originally posted by Granite

I am sure that the new site is great, but I don't seem to be able to login†anymore. Have they changed the login details?
I would try asking them directly but the contact us link still does not work.


Ive login using my old details
Alex

-------------


Posted By: craiggo
Date Posted: 27 Mar 12 at 10:34pm
Granite, I have the same problem.

I have contacted Andy Wibroe and am waiting for a reply. His email address is Andy.Wibroe@rya.org.uk

Hopefully access will be forthcoming. I assume that all the data from previous years inputs will still be there?


Posted By: ellistine
Date Posted: 27 Mar 12 at 10:47pm
Originally posted by Granite

I am sure that the new site is great, but I don't seem to be able to login anymore. Have they changed the login details?
I would try asking them directly but the contact us link still does not work.
Same problem too.


-------------


Posted By: AlexM
Date Posted: 28 Mar 12 at 9:07am
PM Chrisg he's very good at sorting it out :)

Over night and not uploading anymore results my reports page has changed again.... not liking my 100 h'cap! lol





-------------


Posted By: r2d2
Date Posted: 28 Mar 12 at 9:45am
the same would go for your 300 number:-)



Posted By: Oli
Date Posted: 29 Mar 12 at 5:00pm
this is what we have gotten back so far from 467 races. we are looking at making changes when handicaps reach 0.3cf and above.
 

-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: AlexM
Date Posted: 29 Mar 12 at 5:17pm
Good job! I only started doing this for the club last year so only had last years races (150) so i think i need more races'.
I was going to wait till I got a cf of higher than 0.7 or whenever the comment changed to high. I think if we started changing them at 0.3 (Low) it would be hard to convince some! Is 0.5 medium??

Alex

-------------


Posted By: Hector
Date Posted: 29 Mar 12 at 5:57pm
Originally posted by AlexM

PM Chrisg he's very good at sorting it out :)

Over night and not uploading anymore results my reports page has changed again....



I'm confused,  - how/why has it changed if you haven't uploaded any more results?




I also can't login using the old details - and haven't had any response from RYA yet.



-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: AlexM
Date Posted: 29 Mar 12 at 6:06pm
Looking back at the reports given at the time it seems the date range has changed therefore including more of my results. I think they must be still sorting a few things out.

I'm using my old login, slop_idol are you using yours?

Alex

-------------


Posted By: sargesail
Date Posted: 29 Mar 12 at 8:53pm
Originally posted by slop_idol

this is what we have gotten back so far from 467 races. we are looking at making changes when handicaps reach 0.3cf and above.
 

Wow - so the Blaze has one of the highest confidence for a new handicap and would drop 50 points!  And the 400 comes up to 997 admittedly with low confidence (Which makes the speed order Blaze, 300, 400.  Healthy dose of salt required in all this work I feel.


Posted By: AlexM
Date Posted: 29 Mar 12 at 9:02pm
No you are reading that wrong.
Blaze has a 0.2 (Low) cofidence which you wouldn't use
The 400 has 0 (even lower) confidence!!

The only new h'cap you'd come close to using is the FB at 0.5 but still isnt there yet.

I think a healthy dose of reading what the results mean before dismissing!

-------------


Posted By: sargesail
Date Posted: 29 Mar 12 at 10:25pm
Sorry I was misread by slop idol posting about using them from 0.3 and upwards....I realise its low, but it's still one of the highest in that table.


Posted By: Oli
Date Posted: 29 Mar 12 at 11:56pm
We have discussed changing at 0.3cf due to it being a good amount of data for the sample size. When we can pool like venues in to the system, and the cf goes up we will use a higher cf before we change, but as it is the boats are currently performing to these numbers and as such should be changed. Now if it turns out to be too harsh the data will go the other way and we will change it again. Its great that the system now reflects this, but I think that having a pre defined low medium and high value against the cf number to be like the old style of defining the yardsticks with primary secondary numbers etc which will scare people off changing numbers.

As for the blaze moving as much as it does, I have to say that it is probably along with the fireball the most suited boat to our stretch of water, therefore it will perform above its percieved performance range for most clubs.  There hasn't been a lot of all in races at the club where either a blaze or fireball hasn't won or dominated as fleet the results in the last few years, the Musto is another one that when conditions suit can't be touched on handicap, so every dog does have its day I guess.

As it stands I think this new system is much better and with perhaps a few minor tweaks and the proposed planned additions it maybe as perfect as handicapping can possibly get.  The trick is not to be scared to change numbers if it looks like they need changing, just make sure everyone knows how you plan to implement changes and that the rules used to apply them are the same for all classes.


-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: chrisg
Date Posted: 30 Mar 12 at 8:23am
Alex,

Sometimes I get the problem with not all of the classes appearing in the reports page too. I'll see if Simon has an answer for that, but usually if I press f5 (refresh) not the button on the webpage my classes come back. Dont know if its clicking on reports too soon after logging in or something simple like that.

Glad you have found the site useful. I'd say numbers with a confidence factor of 0.3 which may still be called low on the site are considerably better than finger in the air stuff so saying you'll adjust when you reach that level is fair I'd say. I'll be suggesting something similar to my club. I've got over 530 races in for my club and only get 4 classes with a confidence factor of 0.7. Havent yet managed to achieve anything greater than that.

I think Andys release email, or it may be in the help files, stated that the confidence factor algorithm will be being looked at by the RYA once there is more data in the system but they are pretty confident it is a good starting point. Mark E helped produce it so may have more details. Mark - any thoughts?

For those of you guys who havent got logins, I have been pestering Andy on a daily basis so hopefully you will get them soon.

Any probs please ask.

Chris


-------------


Posted By: AlexM
Date Posted: 30 Mar 12 at 9:02am
Hi Chris,
I think itís an IE issue because it doesnít seem to happen on Firefox or Safari.
When does it change to medium confidence? I think itís pretty hard to convince people that we should be changing the number if it says low against it.

Alex


-------------


Posted By: chrisg
Date Posted: 30 Mar 12 at 10:02am
Not sure Alex, I'll ask the question. I only have some classes on 0.4(low), then others jump to 0.7 (high). I thought I had some in between but haven't. It may be that we need to rethink the wording of low, medium and high. Any thoughts on what would be better? Perhaps 0.1-0.2 - low, 0.3-0.4 - reasonably, 0.5-0.6 - fairly, 0.7 and above - highly.

I use firefox and i definitely get the problem of not all classes showing up sometimes. I just logged in to check my confidence factors and had to hit refresh about 4 times before they all appeared. Definitely something that needs investigating.
Something I forgot to mention earlier is that when you upload, I think the PYS site does the race claculations immediately but only updates the report for the classes at regular intervals. So if you upload and want to check immediately it may not have updated the report page, thats what the floating refresh button is for on the report page.

Chris


-------------


Posted By: marke
Date Posted: 30 Mar 12 at 10:58am
< ="text/">p, li { white-space: pre-wrap; }

Chris/Alex

The confidence factor is a composite factor looking at sample size, the distribution shape, and the historical trend. Each of these is a useful measure of the validity of the data. The tricky bit is joining the factors together to give a simple to understand single factor. As Andy said in his email, this aspect of the CF will need to be tuned when the PYS has suficient data to be able to get some meaningful data out of it.


The issues that you and AlexM are seeing look to me like an issue with the individual factor scaling. When the pressure is off Simon to get the site released I'll take a look at some specific datasets and see if I can spot any problems with the algorithm or the implementation.


Absolutely endorse Chris' comments about using it as a relative factor until the RYA has had chance to tune it a little - looking at the the data shown, even a low factor is better than no data at all, or murmurings in the bar Wink


Mark



Posted By: Simon Lovesey
Date Posted: 30 Mar 12 at 11:17am
Pleased people are starting to find the new PYS site useful,  as aready stated we are very much in early stages of release and refining the system as we get a better understanding of the vast number of variables at play here.

With regard to the reports,  these are built in advance  (there are many more automatic calculations than V1) -  this means if you are jumping around you are likely to see a lag in the data -  again this is all part of the refinement process and we are working on improvements in this area.

New logins are being released on a batch basis -  this allows us to manage and take on board comments received in a practical manner,  rather than be flooded with the same comment.  The staggered release also allows us to manage server loading.  The good news is once you have received your login it takes no time at all to upload and process several years of data -  so thank you for your patience.


-------------
http://www.sailracer.org" rel="nofollow - www.sailracer.org
Online Sailing Results, GPS Tracking & Event Management


Posted By: AlexM
Date Posted: 30 Mar 12 at 1:02pm
Thanks for all the replies and the hard work!

I think i'm going to go with everyone else and anything over the 0.3 we'll change and assess at the end of each series.

So it's

RS400 from 949 -> 940
Solo from 1153 -> 1147

Seem reasonable

Alex

-------------


Posted By: craiggo
Date Posted: 30 Mar 12 at 7:52pm
I'm a bit gutted that none of the historic data has been ported across. Are we honestly expected to go back and reload all historic data?


Posted By: chrisg
Date Posted: 30 Mar 12 at 10:12pm
Craiggo,

I thought originally that the site was going to cut old numbers off and add new in to it, but as it works the numbers out using a different method I think its best just to start again. However, I can see this causing problems to any clubs that had already adjusted numbers based on the old website. Although not sure there were many to be honest.

I reuploaded every single race I had in the old system (500+) to the new site in about an hour and a half. With the new upload from sailwave file (not uploading from inside sailwave) you can do a whole series at a time and it runs the calcs on every race immediately. The whole site is much less labour intensive. Let us know how you get on.

Chris


-------------


Posted By: craiggo
Date Posted: 30 Mar 12 at 10:33pm
Thanks Chris,

The issue is trying to find all the old .blw files and then making sure that the boats spelling are consistent in all series.

I'll give it a go and let you know if I come across any issues.

Paul


Posted By: chrisg
Date Posted: 30 Mar 12 at 11:23pm
Ahh yes, hope you can find them. It is worth opening the file and just ahving a quick glance down the list of classes to check youve called them all the same thing as its easy to lose some. You need 9 data points I think it is before boats appear in the report page so if you accidentally misname one, it may never appear...

Chris


-------------


Posted By: sargesail
Date Posted: 02 Apr 12 at 8:28pm
Originally posted by slop_idol

We have discussed changing at 0.3cf due to it being a good amount of data for the sample size. When we can pool like venues in to the system, and the cf goes up we will use a higher cf before we change, but as it is the boats are currently performing to these numbers and as such should be changed. Now if it turns out to be too harsh the data will go the other way and we will change it again. Its great that the system now reflects this, but I think that having a pre defined low medium and high value against the cf number to be like the old style of defining the yardsticks with primary secondary numbers etc which will scare people off changing numbers.

Yes - but on the other hand it would be useful to ask for those options, to see how they varied at different cfs.

As for the blaze moving as much as it does, I have to say that it is probably along with the fireball the most suited boat to our stretch of water, therefore it will perform above its percieved performance range for most clubs.  There hasn't been a lot of all in races at the club where either a blaze or fireball hasn't won or dominated as fleet the results in the last few years, the Musto is another one that when conditions suit can't be touched on handicap, so every dog does have its day I guess.

And this is where I struggle with using cfs as low as 0.3.  If the Blaze is 50 odd PY points better then you ought to see a similar change in the 300.  Now I'll accept that there aren't enough races yet for the 300, but you are, of course, changing the Blaze relative to everything else - if that's actually because you're water suits the Blaze then by not changing similar boats you are creating an anomalous situation.  How much is Crew Skill Factor too. 

As it stands I think this new system is much better and with perhaps a few minor tweaks and the proposed planned additions it maybe as perfect as handicapping can possibly get.  The trick is not to be scared to change numbers if it looks like they need changing, just make sure everyone knows how you plan to implement changes and that the rules used to apply them are the same for all classes.

Yes I'm all for experiment, as long as there's a control, and the experimenters understand what they're doing.  The irony in all of the PY work is that you could make statistically perfect fair racing and still annoy just about everyone!


Posted By: Oli
Date Posted: 03 Apr 12 at 8:56am
Originally posted by sargesail

Originally posted by slop_idol

We have discussed changing at 0.3cf due to it being a good amount of data for the sample size. When we can pool like venues in to the system, and the cf goes up we will use a higher cf before we change, but as it is the boats are currently performing to these numbers and as such should be changed. Now if it turns out to be too harsh the data will go the other way and we will change it again. Its great that the system now reflects this, but I think that having a pre defined low medium and high value against the cf number to be like the old style of defining the yardsticks with primary secondary numbers etc which will scare people off changing numbers.

Yes - but on the other hand it would be useful to ask for those options, to see how they varied at different cfs.
 
we will only move boats above 0.3cf at the begininng of a new series, that when you look at the data is only two boats currently as the others dont have eneough data yet, and these will move around a fair bit over the season looking at how they have moved in the last few races, the point is we are trying to get to a balance where we may not have to move a boat as it may only move 2py points either side of its base py.

As for the blaze moving as much as it does, I have to say that it is probably along with the fireball the most suited boat to our stretch of water, therefore it will perform above its percieved performance range for most clubs.  There hasn't been a lot of all in races at the club where either a blaze or fireball hasn't won or dominated as fleet the results in the last few years, the Musto is another one that when conditions suit can't be touched on handicap, so every dog does have its day I guess.

And this is where I struggle with using cfs as low as 0.3.  If the Blaze is 50 odd PY points better then you ought to see a similar change in the 300.  Now I'll accept that there aren't enough races yet for the 300, but you are, of course, changing the Blaze relative to everything else - if that's actually because you're water suits the Blaze then by not changing similar boats you are creating an anomalous situation.  How much is Crew Skill Factor too. 

 

the blazes more often than not beat the 300 around the course so the 300 may not change relative to the blazes performance.  my understanding of the system is that we no longer use benchmark boats to calculate other boats py and as such the handicap given to your class is the one that you as a class have sailed to, if you want to achieve a more favourable handicap, slow down and fudge the results, but you are only cheating yourselves if you do this.

We are also trying to move away from people having a best guess at a particular number as this will and does cause more arguments.  we have been given a great oportunity to you a mathematical system that uses our performances to attain our py, so why whouldnt we?As it stands I think this new system is much better and with perhaps a few minor tweaks and the proposed planned additions it maybe as perfect as handicapping can possibly get.  The trick is not to be scared to change numbers if it looks like they need changing, just make sure everyone knows how you plan to implement changes and that the rules used to apply them are the same for all classes.


Yes I'm all for experiment, as long as there's a control, and the experimenters understand what they're doing.  The irony in all of the PY work is that you could make statistically perfect fair racing and still annoy just about everyone!
 
absolutely you can annoy everyone, and you could with the old system by not changing so you'll never win over 100% of the people, but not to change because the minority may not agree with you at first is just stupid and we are trying to encourage people to attend. The majority of people i have spoken to understand what we are trying to achieve and agree to it, yes some move more than others but again it is based on their perfomances so should see it as an incentive to sail improve their sailing (this is true if your handicap moves up or down).


-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: Medway Maniac
Date Posted: 03 Apr 12 at 10:49am
You really do risk opening a can of worms if you apply the suggested PY's.  When the 'sanctity' (so far as the average club member is concerned, I'm afraid) of the official figures is broken, it becomes open house for moaning about PY's, people get demotivated and may turn out less often.

Unless you have a lot (dozen or so) of boats from a given class there's no question that a personal handicap element remains, which is demotivating for the sailors in the class being changed.  With just a few boats involved, if they sail a blinding series of races, they know their reward will be a hit to their handicap - great!  And it's not as if it's like a golf handicap where you can wear it as a badge of your ability.  The feeling is especially bad when clubmates tell them that it wasn't their blinding tactics but the current PY that gave them the victory.  Not a nice atmosphere for a club to have.

We've discussed applying the suggested PY's to all boats at WSC a few times now, but each time have concluded that we'd do more harm to the plausibility of the results for the average member than good.  Each week, only one person would be happy - the winner - while the rest moaned about PY's.  

The average member prefers to use the 'sanctified' (even if they aren't) PY's on the official list and just take the rough with the smooth.  The value of the website so far as we are concerned - and it's a great value - is the ability to feed back our results regularly so that the 'sanctified' figures are brought quickly up to date.  We've already seen the benefit with the Phantom, the major and only real source of complaints before all this started.

We also use the suggested PY's when the class has no official number, such as the V3000, 59er with trapeze, and Alto.  As one of the four V3k owners, I see the necessity (as asst sailing sec I voted in favour), but still find the personal handicap aspect demotivating.  For other reasons, I hopped back in an L3k for the past two seasons, and have to say it was nice to be out of the PY crossfire; back in the V now though - more fun till we get ashore...


-------------
http://www.wilsoniansc.org.uk" rel="nofollow - Wilsonian SC
http://www.3000class.org.uk" rel="nofollow - 3000 Class


Posted By: Guests
Date Posted: 03 Apr 12 at 11:11am
Originally posted by Medway Maniac



The average member prefers to use the 'sanctified' (even if they aren't) PY's on the official list and just take the rough with the smooth.  The value of the website so far as we are concerned - and it's a great value - is the ability to feed back our results regularly so that the 'sanctified' figures are brought quickly up to date.  We've already seen the benefit with the Phantom, the major and only real source of complaints before all this started.

 
 
I think that sums it up for me........If the RYA could carve the PY list in stone and set light to someones bush most sailors would be in heavenTongue 
 
But there is a large minority who don't take things as gospel, who want things to be demonstrably fairer for their water. Those who don't tug their forelock or cross themselves at the mention of the RYA.
 
The debate between the two camps will be endless. A bit like left and right wing politics I'm afraid, reason has nothing to do with it.
 


Posted By: Oli
Date Posted: 03 Apr 12 at 12:46pm
Originally posted by Medway Maniac

You really do risk opening a can of worms if you apply the suggested PY's.  When the 'sanctity' (so far as the average club member is concerned, I'm afraid) of the official figures is broken, it becomes open house for moaning about PY's, people get demotivated and may turn out less often.
But at least we will be using numbers derived from the guys that run the system and not someones idea of a good number.  As for demotivating i see it quite differently (maybe its my glass half full attitude to sailing), but if your number drops then you have done well therefore becoming more of a challenge for you and again if you win then youve done even better (winning isnt everything it is handicap racing after all) and if your number drops thats a huge incentive to up your game and the when you do you will win therefore re-inforcing a positive incentive for you to race.
Unless you have a lot (dozen or so) of boats from a given class there's no question that a personal handicap element remains, which is demotivating for the sailors in the class being changed.  With just a few boats involved, if they sail a blinding series of races, they know their reward will be a hit to their handicap - great!  And it's not as if it's like a golf handicap where you can wear it as a badge of your ability.  The feeling is especially bad when clubmates tell them that it wasn't their blinding tactics but the current PY that gave them the victory.  Not a nice atmosphere for a club to have.
 
We currently have that very situation because of the static pys, so after all these years of them being static surely its time for a change, even if its proven to be the wrong thing only time will tell and if you have the minerals to get of your ass and do something about it.  Ive grown tired of people complaining about their current py (both too low and too high, amazing i know but they do!), we have a system that shows promise to even the playing field when it comes to bandits so lets try it.

We've discussed applying the suggested PY's to all boats at WSC a few times now, but each time have concluded that we'd do more harm to the plausibility of the results for the average member than good.  Each week, only one person would be happy - the winner - while the rest moaned about PY's.  
 
Again most of our members are happy to beat their fellow class member and anything above that is a bonus, now some people are in a class of one and yes it would become a personal handicap, but even at 0.3cf you have to do about 300 races before it qualifies for change and so they will be for the foreseable future still race off the natioanl number (wheres the issue?)

The average member prefers to use the 'sanctified' (even if they aren't) PY's on the official list and just take the rough with the smooth.  The value of the website so far as we are concerned - and it's a great value - is the ability to feed back our results regularly so that the 'sanctified' figures are brought quickly up to date.  We've already seen the benefit with the Phantom, the major and only real source of complaints before all this started.
 
they may prefer the 'sanctified' number but thats becasue they havent had a choice before, at least here are two parties in the race now.

We also use the suggested PY's when the class has no official number, such as the V3000, 59er with trapeze, and Alto.  As one of the four V3k owners, I see the necessity (as asst sailing sec I voted in favour), but still find the personal handicap aspect demotivating.  For other reasons, I hopped back in an L3k for the past two seasons, and have to say it was nice to be out of the PY crossfire; back in the V now though - more fun till we get ashore...
 
weve talked about doing some races (mostly for fun) off a personal handicap, to test the water as it were, it like any py system will have pitfalls so again its something that can be looked at for the future.


-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: AlexM
Date Posted: 27 Apr 12 at 5:39pm
Is it possible to view a full summary table of the results? I would like to check that all the classes have the same name and are pulling though correctly i.e. NATIONAL 12* and 12*

Alex

-------------


Posted By: Oli
Date Posted: 14 Jul 12 at 3:35pm
update of bsc numbers
 
 


-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: AlexM
Date Posted: 30 Aug 12 at 3:53pm
L&LSC numbers
About to amend the numbers for the next series



-------------


Posted By: Oli
Date Posted: 15 Sep 12 at 1:27pm
update of bsc py list.  mutos are now down to 838.
 


-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: AlexM
Date Posted: 15 Sep 12 at 4:57pm
Have you started implementing them yet? We just started for autumn. Used the Yorkshire Dales SC format of dampening the changes. Our numbers came out different than I thought they would.... Thought Solo would go down and laser up
in a few weeks they are updating the website again so you can see all data ( ie not just the classes with enough data) and should be able to download and compare with other clubs of the same size/ location to have data for classes not at the club

Class     Last used PN     †'Damped' difference     New L&LSC PN
Enterprise     1117                    -4     1113
Laser             1085                    -9     1076
Miracle             1195                    -3     1192
Optimist             1646                      8     1654
Phantom             1025                    -2     1023
RS 200            1057                         1     1058
RS 300.           1000                      -18     982
RS 400               948                    -9     939
Solo                    1150                      3     1153
Topper             1301                    35     1336

Alex

-------------


Posted By: charlie1019
Date Posted: 16 Sep 12 at 12:26am
Originally posted by AlexM

Have you started implementing them yet? We just started for autumn. Used the Yorkshire Dales SC format of dampening the changes. Our numbers came out different than I thought they would.... Thought Solo would go down and laser up
in a few weeks they are updating the website again so you can see all data ( ie not just the classes with enough data) and should be able to download and compare with other clubs of the same size/ location to have data for classes not at the club

Class     Last used PN     †'Damped' difference     New L&LSC PN
Enterprise     1117                     -4     1113
Laser             1085                     -9     1076
Miracle             1195                     -3     1192
Optimist             1646                     †8     1654
Phantom             1025                     -2     1023
RS 200            1057                         1     1058
RS 300.           1000                      -18     982
RS 400              †948                     -9     939
Solo                     1150                     †3     1153
Topper              1301                     35     1336

Alex


Interesting to see that your results effectively returns the laser to its old handicap of 1077 despite L&L being quite large. I assume this is down to the fact that they have been (when handicap racing) racing generally similar speed boats rather than fast boats that can really stretch their legs and therefore make the most of the larger water?

Good to see both clubs showing a reduced Phantom handicap, proves what we have always believed... ;)


Posted By: AlexM
Date Posted: 16 Sep 12 at 5:21pm
Maybe.... but the 400 does well here too. But after trying to think of reasons I think we've just got a very good laser fleet (some clubs must have to keep the ave where it is)
We did a test on last series results with new and old handicaps and guess what it made little difference !
I've had a look around some of the clubs who are doing the new handicaps and what is clear is to expect the topper to move a lot next time!

Alex

-------------


Posted By: Oli
Date Posted: 17 Sep 12 at 12:19pm
hi alex,
 
yes we made changes earlier in the year, as soon as a boat goes over 0.3cf we make sure that for the next series they use the new number. we now have 3 boats changed, fireball, musto and laser.  we still have a few members who dont like the changes, mostly those that it has affected yet mind, but we are getting there. 
 
ive only had a brief look at comparision results using new numbers and old for the same races and although not making huge changes (predicted it wouldnt make huge changes) it does make the racing closer and can mix it up a bit, which is good as it make syou try harder.


-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: Hector
Date Posted: 17 Sep 12 at 2:01pm
Originally posted by AlexM

Have you started implementing them yet? We just started for autumn. Used the Yorkshire Dales SC format of dampening the changes.
Alex


For anyone wondering what the system Alex refers to is here at Yorkshire Dales SC,  See below from our website.

 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

The RYA System also produces what they have called a confidence level.  'Confidence' refers to the stability of the number - not its accuracy. Confidence in the stability of PNs obviously rises as more race data is input and as we are in the first year of this process we decided to recognise the likelihood of increasing confidence, and so the Sailing Committee has agreed that the changes recommended by the RYA for YDSC will be 'damped' by applying  a % adjustment. Depending on the confidence level the % of any change applied will be as follows   

0.0 - 0.1 = 33.33%       0.2 - 0.3 = 66.67%       0.4 ++++ = 100%


RYA club specific PN list for Club Racing at Yorkshire Dales SC

                 
Data from 03/11 - 03/12  with resulting  RYA recommended PNs
                 
Class Appears Confidence Last used PN RYA Suggested PN Difference % of difference applied  'Damped' difference New YDSC PN
29ER
19
0 924 954 30 33%  10     934
CONTENDER 25 0 993 953 -40 33%  -13     980
FINN 43 0 1060 987 -73 33%  -24    1036
FIREBALL 11 0 980 967 -13 33%  -4     976
LASER RADIAL
56
0.1 1106 1162 56 33%  19    1125
LASER 217 0.4 1082 1087 5 100%  5    1087
MIRROR 25 0 1385 1389 4 33%  1    1386
MUSTO SKIFF 16 0 860 867 7 33%  2     862
PHANTOM
25
0 1030 1014 -16 33%  -5    1025
RS200 58 0.1 1057 1075 18 33%  6    1063
RS600 21 0 920 917 -3 33%  -1     919
RS800 23 0 822 801 -21 33%  -7     815
TOPPER 62 0.1 1297 1343 46 33%  15    1312
VORTEX
145
0.2 937 909 -28 67%  -19     918
                 

Classes not shown did not return enough data at YDSC to provide an accurate 'club' number. They will largely remain on the RYA 'standard'  number. The Sailing Committee reserve the right to adjust PNs in the interest of fair racing - for instance in the event of a boat new to the club with known performance characteristics that affect likely PN at Grimwith. Where possible, this will be done using reliable sources of data-led information.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

The stand out change on our list is the Finn. It was one boat, very well sailed, and also a bit of a series picker. That is. did very well in Wed eve light winds, not so good in a breeze, but still beat 95% of Lasers, RS200 etc in wind.  A little worried that this was in effect a Personal Handicap, but thats a very good reason to 'damp' changes as we did.
The Topper had a big change, but that wasnt a suprise as they dont go well at our large windy venue. (And we have some very good Topper sailors).
Other big change was Vortex which went to 918 (damped from 908) whereas the Natonal PN for the Vortex is 945. They have still won probably 65% of races, and 75% of series, so we feel vindicated that our PN number is more accurate than the RYA number and is likely to drop further.
In general, the changes have made little difference to results, except the margins of victory / defeat have reduced, and the Lasers / Toppers get an occasional look in now.

Shortly going to update our numbers and will see if it goes where we expect  then, but also hoping for increased confidence with another seasons data.


-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: andy101
Date Posted: 17 Sep 12 at 2:12pm
Slightly suprised if it is a generally windy venue the radial moves up compared to the standard laser, I would have thought it would be the other way round as in my experience is radials struggle in the light stuff but go very well when the breeze picks up?


Posted By: Hector
Date Posted: 17 Sep 12 at 2:17pm
All to do with who sails them.
In the Lasers we have a National Youth Squad member who was just top of his age group at the Europeans, 3 ex national Champions (not in a Laser, but top 5 in Laser Nats and race winner at Laser worlds etc), and many other very good sailors.
Radials tend to be beginners, kids coming up from Topper, Ladies who don't sail often and struggle when windy, and Old blokes trying to get more comfy!




-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: marke
Date Posted: 17 Sep 12 at 2:48pm
Hi

Keith's approach is a reasonable one - when applied to your own club data entered into the PYS.  It will take a little while for the confidence factor to be tuned on the website + since most clubs will not have a long period of data you can't take into periodic variations based on exactly who is sailing.

However I would not recommend applying it  to the annual national PNs published by the RYA, as you would be applying a damping factor on top of some inherent damping factors in the 'returns' based system.  As the use of the PYS increases and (hopefully soon) replaces the returns system then this caveat can be ignored.

Mark


Posted By: AlexM
Date Posted: 17 Sep 12 at 3:05pm
Hi Keith
The only thing I didn't take from your dampening was those classes with a cf of zero, I thought it was too much of a personal h'cap. I.e. we'd have the 600 closer to a 1000
Alex

-------------


Posted By: Hector
Date Posted: 17 Sep 12 at 3:58pm
Originally posted by AlexM

Hi Keith
The only thing I didn't take from your dampening was those classes with a cf of zero, I thought it was too much of a personal h'cap. I.e. we'd have the 600 closer to a 1000
Alex


Catch 22 isn't it?
We have a very good Contender (another who's won a race at his worlds), the good Finn, the RS500 National Champion and a couple of other v good individual boats. To not change them at all on the basis of low confidence factor would penalise ( or advantage)  the boats with (slightly) higher factors that we have changed. 
The point as I understand it is that the computed figures are correct. The confidence level is low due to lack of data = far more likely to change compared to classes with more data, BUT the computed PNs are correct nonetheless.

I did an exercise using a few 'typical races' to illustrate to our members what actually happens in a typical race.
With say 20 boats out, the middle 2 or 3 boats sail roughly to their PN. The winner is roughly -80 below PN and the last is roughly +80 above PN. The spread reduces if the conditions don't favour one particular boat or type of boat too much, but increase if it's perfect for one type.
So with a spread of around 160 PN pts in most races, adjustments of +/- 20 pts don't necessarily move results very far.

What was illuminating was what actually happened with a boat that allegedly has extreme of performance such as the Vortex. When sailed by someone good in perfect F3 Vortex Conditions they can easily perform to a big reduced PN (say -100). The suprise was that they didn't lose by much in 'poor' Vortex conditions of F1 (roughly -20). IE they were rarely below half way and so their PN was almost always reduced. They would suffer more in near calm - but we don't get many of those, and if we do they rarely go out!
Taking an average sailor in an average performing boat (say an RS200). In lighter breeze, they would sometimes just get into the top half, sometimes not (largely dependent on beating most of the Lasers) and so would have a neutral PN. But in a stronger breeze our average RS sailor would struggle and be + 40 or 50 pts on their PN. Luckily we also have a top RS200 team who tend to balance that out so the RS200 number hardly changes at all.
So it's hugely to do with who sails what, and as such, we hope that the National database gets more accurate with more returns based on actual performances rather than the guesswork and fear of change / upsetting members thats previously prevailed.



-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: Hector
Date Posted: 17 Sep 12 at 4:03pm
Should also point out that our results database was compromised last year due to issues with corrupted downloads and the fact that we used to allow multiple boats in a series. So someone who sailed the RS500 for 10 races then a Laser for one would get a result (for the series) with a boat labelled RS500/Laser. The PYS system didn't deal with that and so the entire number was compromised - hence only those boats with meaningful results in our table.
We've sorted that for this year and will have far more data as a result.


-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: AlexM
Date Posted: 17 Sep 12 at 4:22pm
Fair point, will have to consider for the winter series.
I've gone back to 2007 to get as many results on the web as possible ( adds a bit of more weight to the change )
Hopefully with the update we'll be able to share results with similar clubs. (how that is decided I'm not sure?)

Alex

-------------


Posted By: Hector
Date Posted: 17 Sep 12 at 5:25pm
Originally posted by slop_idol


 ive only had a brief look at comparision results using new numbers and old for the same races and although not making huge changes (predicted it wouldnt make huge changes) it does make the racing closer and can mix it up a bit, which is good as it make syou try harder.


Almost exactly our experience.Clap


-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: AlexM
Date Posted: 21 Sep 12 at 6:58pm
Major changes on the reporting page of the web page!
Looks like i've got some tidying up to do!

Alex

-------------


Posted By: Oli
Date Posted: 06 Oct 12 at 4:45pm
october update of blackwater sc py list.  a few minor changes.  link to the most recent all in handicap race from last weekend using the updated number. http://blackwatersailingclub.org.uk/images/raceresults/Blackwater%20Cup.htm - http://blackwatersailingclub.org.uk/images/raceresults/Blackwater%20Cup.htm


-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: MattHarris
Date Posted: 22 Oct 12 at 12:09pm
Has anyone noticed a speed improvement on the new website?  Previously mine took a few days to appear, i've uploaded a about 100 races and after 2 hours only two have appeared.  Is it worth me looking back this afternoon or is it still taking a few days to appear?

-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: marke
Date Posted: 22 Oct 12 at 12:26pm
Matt

That issue has been fixed.  It wasn't entirely slowness, it was just getting stuck.  It is processing much quicker now.  We (and several other clubs) now have 500 races loaded.  If it is still going slowly for you I'd report the problem, may be something in the data that is causing the problem.

Mark


Posted By: AlexM
Date Posted: 22 Oct 12 at 12:27pm
It's loads better. You'll see them this afternoon.
Even better now that it's got all the classes.

-------------


Posted By: AlexM
Date Posted: 22 Oct 12 at 12:34pm
Ours to-date


-------------


Posted By: MattHarris
Date Posted: 23 Oct 12 at 9:53am
They did appear later on in the afternoon, although from home some of them appeared virtually straight away.  The new layout for the reports is much better although the confidence factor doesn't seem to climb ver fast.  I've got around 4 years of data uploaded and the highest we have is a 0.5 confidence, has anyone got much higher than this and how much data did it take?
 
Also is anyone struggling to get on this morning, i'm getting an internal server error :(


-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: AlexM
Date Posted: 23 Oct 12 at 10:11am
c.f Iíve seen 0.7 (see above, fireballs) Iím sure the c.f. raises quicker when youíve got more of the same class in a race rather than a few doing many races.

Im getting error as-well. must be working on it
Alex

-------------


Posted By: marke
Date Posted: 23 Oct 12 at 10:56am
Alex

The confidence factor for the calculated PY for a class is a composite of three factors
- sample size (bigger sample = higher confidence)
- variation in calculated PY (lower variation = higher confidence)
- change over time (lower change = higher confidence).

This was a first cut at establishing high confidence.  As the database grows it will be possible to look at other measures.  But in general the more data you have for a class, over a longer period of time and with the resulting PYs in a narrow range then the confidence factor should rise.  0.7 is pretty good.

Mark



Posted By: RS400atC
Date Posted: 23 Oct 12 at 12:26pm
Originally posted by AlexM

Ours to-date


That looks like a set of data that ought to be rebased around the RS400.
What's the second RS400 entry?


Posted By: AlexM
Date Posted: 23 Oct 12 at 12:31pm
Sorry the screen shot was taken after I sorted all the names out. The space was missing out of the name and should be with the other results.

Alex

-------------


Posted By: RS400atC
Date Posted: 23 Oct 12 at 12:53pm
Originally posted by AlexM

Sorry the screen shot was taken after I sorted all the names out. The space was missing out of the name and should be with the other results.

Alex


That space really slows the boat down....


Posted By: AlexM
Date Posted: 23 Oct 12 at 12:58pm
lol, yes didn't notice that!
I see why you asked now and I suppose it could of been a 400 without a kite.

-------------


Posted By: MattHarris
Date Posted: 24 Oct 12 at 11:18am
I was going to post our results to date as they seem quite different to everyone elses, however the websites still not letting me in :(

-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: AlexM
Date Posted: 24 Oct 12 at 12:10pm
Working now

-------------


Posted By: MattHarris
Date Posted: 24 Oct 12 at 1:33pm
I'm having another problem:
for some races when we have multiple races on a single day they are being flagged as duplicates.  there is no ignore button to stop this however if i delete the multiple values the remaining race is accepted.  it only seems to be doing it for certain races where i have entered a series name.  If i remove the series name then it will accept the races....  i'm guessing this may be an issue with how sailwave packages the data before sending or maybe how the RYA site defines duplicates.  Either way is anyone else having this issue?
 
Think this may be why some races weren't showing up the other day as it picked all of the results up as duplicates :(
 
Matt


-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: AlexM
Date Posted: 24 Oct 12 at 1:44pm
I think it's a naming issue. We have 3 races a day but they are all named different R1,R2,R3. could this be the issue?
Here's the output from SW for us -
http://www.llsc.org.uk/LLSC/Results/2012AutumnSaturday.htm

Hope this helps
Alex

-------------


Posted By: MattHarris
Date Posted: 24 Oct 12 at 1:47pm
I tried renaming the races and giving them different start times an hour apart but it didn't change anything.  I've just put them in without series names for the time being so at least the data is on there even if you have to look up the date to find out what race it was for.  Having said that its purely data for number crunching so i can't see this being an issue.  Will put up the results once i've uploaded this seasons information

-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: MattHarris
Date Posted: 24 Oct 12 at 1:54pm
 
So thats the output for the last 3-4 years, although the .7 confidence factor on the GP14 still doesn't give me a printable stamp of approval...


-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: AlexM
Date Posted: 24 Oct 12 at 2:12pm
0.7 Is very good c.f.
Take a look at Yorkshire Dales implementation of the figures thatís what weíve followed:
c.f. 0.1 used 33% of the change
c.f. 0.2 used 66% of the change
c.f 0.3+ used 100% of the change
Interesting to note youíve got similar results to L&LÖ.
Laser down weíre 1074 (something I wasnít expecting at all!! I would have bet anything on this going up)
RS 400 down weíre also 942

Are you over in Southport? You could class our lakes similar and we could use some of your high c.f. classes e.g GP14. I think this option was originally mentioned when the site got setup?


-------------


Posted By: MattHarris
Date Posted: 24 Oct 12 at 4:25pm
Yeah I'm from Southport, not sure on the lakes similarities.  I agree they're both small but with the islands we find that we can be quite restricted and we sadly never get the same blasting in the 400 without running out of space. 
If you want to use any of the figures please feel free, although I think we'd get some grief trying to use the L&L ones, especially the enterprise one.  Having said that the figures are quite similar although.
 
With regards to the yorkshire dales idea i think we'd have to change them based on the confidence wording to get more traction with members (0.2-0.3, 0.4-0.6 and 0.7+) but thats for another day entirely.  For the moment I'm just making sure its done ready for the return and for bar discussion over winter so the idea has been thoroughly chewed over before maybe starting a discussion next year.


-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: marke
Date Posted: 24 Oct 12 at 4:40pm
Matt

Its far too early in the development of the PYS to be worried about 'absolute' values of c.f.  With a few years experience the c.f will be modified and tuned and then we can start talking about absolute thresholds.  For now the message for clubs should be that a cf of >0.5 is bloody good and should give a lot more confidence than a class with a cf of 0.1.  Keith's approach at Yorkshire Dales seems like a good way to approach it at this stage.

We started adjusting our PYs locally a couple of years ago, and on all sorts of measures we can demonstrate that the racing has got closer.  There was a little bit of grumbling when it was brought in - but much less than I expected and once you have made the move its a lot easier to keep it going. 

mark


Posted By: Hector
Date Posted: 24 Oct 12 at 10:25pm
I think the RYA have named the 'confidence factor' badly. It smacks of error / inaccuracy when it's not meant to. I think it should have been named Stability Factor as that is a more accurate description of what it is. 

Mark and I are clearly on the same page RE application of the PYS. 
The outcome of our changes has been much closer results, and a little more variety in boats in the top 5. Same ones seem to win - dependent on wind, but the top boats from other fleets are more likely to beat the not so good sailors from the 'favoured boat' - which I'd say is a good thing.

We didn't just decide to 'do our own thing'. I attended two talks / meetings on the subject by Andy Wilbroe / Bas (at The Dinghy Show).
The logic regarding confidence factors as I understand it following those talks is described on our website and pasted in part below

NOTE: They (RYA) have also give a confidence level.  'Confidence' refers to the stability of the number - not its accuracy. Confidence in the stability of PNs obviously rises as more race data is input and as we are in the first year of this process we decided to recognise the likelihood of increasing confidence, and so the Sailing Committee has agreed that the changes recommended by the RYA for YDSC will be 'damped' by applying  a % adjustment. Depending on the confidence level the % of any change applied will be as follows   

0.0 - 0.1 = 33.33%       0.2 - 0.3 = 66.67%       0.4 ++++ = 100%

The percentages we chose were partly as we had no high confidence figure after one years input, and secondly that to do less would have let some key classes escape any worthwhile alteration, and third, the figures are correct - they just may not be stable. The Highest c.f. was the Laser at 0.4, and that only moved a paltry 5 points from the standard number so we were Very Confident we could implement that 100%.  The rest followed logically.

I think the RYA have named the c.f. badly. It smacks of error / inaccuracy when it's not meant to. I think it should have been named Stability Factor as that is a more accurate description of what it is. 

I'll post our latest PNs here soon - once agreed at club. Our Laser figure drops slightly - Like Mark, I'm  suprised at that - but others have dropped more so overall they gain on most other classes. With very much increased / high confidence figures for more classes, we'll be applying more of the change from National PN to Yorkshire Dales PN for our next series, and no-one has complained about it at all, on the contrary, most comments are very positive.



-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: JimC
Date Posted: 25 Oct 12 at 6:19am
I dunno, confidence seems like a reasonable term to me. The wilder the data in all respects the less confidence you have it it. I don't see a connection with stability at all.


Posted By: marke
Date Posted: 25 Oct 12 at 9:46am
From a scientific perspective I think it is a confidence factor (in this application I think consistency is part of the accuracy assessment - unless of course you are making all RS400 sailors at your club tow a bucket!).  In the interests of avoiding argument it could perhaps be called the "anti-grumbling factor" or "getting the nay-sayers to shut up factor" - or perhaps following the Duckworth/Lewis approach from cricket and call it the Wibroe/Edmonds factor. Smile

We are seeing exactly the same impact from local PY adjustment as Keith - "Same ones seem to win - dependent on wind, but the top boats from other fleets are more likely to beat the not so good sailors from the 'favoured boat' - which I'd say is a good thing".  In several of our trophy races this year we have 5 or more different classes in the top 10, and in our overall club championship (series of ~12 trophy races held in conditions throughout the year) we had 8 different classes in the top 10 (including for the first time in a decade a well-sailed laser).  


Posted By: AlexM
Date Posted: 25 Oct 12 at 9:55am
I like that - "anti-grumbling factor" or "getting the nay-sayers to shut up factor"   

I'm still amazed about the PY for the Laser

We've changed ours for this Autumn/before Christmas series, will have a review at 1st Jan for 2013 and then probably see where we are when the new RYA PY's come out.

-------------


Posted By: Hector
Date Posted: 25 Oct 12 at 1:09pm
From the RYA / PYS site:
The reports page will currently show a list of Portsmouth Numbers based the results a club has uploaded. Each is given a confidence factor based on the volume and quality of the data behind each PN. As the data increases and proposed number becomes more stable it will be given the RYA seal of approval and a high confidence factor.

So to get a high confidence requires a stable number - hence the c.f figure is a measure of stability.
Nuances aside, I stick by my assertion that stating you have low confidence in a figure smacks of possible error, and if 'some people' want to resist the change, they can more easily hang their argument on that point.


The RYA site goes on:

ALL CLUBS are reminded that this number may still not be perfect and the club will still need to take the ultimate decision as to if they use that number and or an adjustment of that number.
Hence the YDSC decision to recognise the low stability of most of last years Club numbers and hence damp the adjustment from National numbers. This year there is far higher confidence from the PYS site, and after a year of racing and results we are also more confident ourselves.

To simplify my point  - 
Are the RYA confident that the process accurately and correctly calculates current Club PNs?   YES
(I understand the process is based on YR2 but adjusted with algorithms that take into account data spread, volume etc).

Are the RYA confident that the figure produced are likely to remain stable? DEPENDS
(On quality and volume of data)





-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: AlexM
Date Posted: 30 Oct 12 at 12:31pm
Is the data graph for the class working for anyone else? I get all the info apart from the actual line.

Alex

-------------


Posted By: MattHarris
Date Posted: 02 Nov 12 at 1:10pm
Just checked and it looks ok on mine.  Whilst checking I also noticed that I've got an "RYA Technical Approved" stamp on the results for the GP (0.7 cf) and Mirror Single Handed (0.6 cf).  Not sure whether this is a new feature or just I've not noticed it before.
On an aside we had a sailing committee meeting last week at which I brought up the PYS website.  Was interesting to see the mix of responses across the sailors.  Ultimately they decided not to adopt the numbers and stick with the RYA list and have another discussion in 12 months with another years data on hand, we should hopefully have a few more stamps of approval by then too.


-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: Hector
Date Posted: 02 Nov 12 at 2:10pm
Originally posted by MattHarris

Just checked and it looks ok on mine.  Whilst checking I also noticed that I've got an "RYA Technical Approved" stamp on the results for the GP (0.7 cf) and Mirror Single Handed (0.6 cf).  Not sure whether this is a new feature or just I've not noticed it before.
On an aside we had a sailing committee meeting last week at which I brought up the PYS website.  Was interesting to see the mix of responses across the sailors.  Ultimately they decided not to adopt the numbers and stick with the RYA list and have another discussion in 12 months with another years data on hand, we should hopefully have a few more stamps of approval by then too.


At the Dinghy Exhibition presentation last year the RYA said  that the National PY number is simply not the right number for the majority of clubs. It is an average of some very iffy returns from the last decade and beyond, and unless a club is that exact 'average' then the figures are wrong for them.
So was there any sound logic to the decision Matt, or was it a case of  not believing the system works / heads in sand / fear of change or backlash / fleets protecting their percieved advantage on your water? Wink

One thing that helped me to persuade some of the more doubtful both on the committee and otherwise, was to print off a few  'Typical Race' results - showing the individual PNs that would have resulted.

Looking at several races the same thing occurred almost every time:

Wiith a roughly 30 boat fleet, the top 2-3 were often from the same class, and had sailed to somewhere around minus 80-100 points below their Yardstick, while the last few boats were sailing to +100points above their yardstick.  Not suprisingly, the middle few sailed roughly to their yardstick (and were often very sailors). Everyone in- between was approx pro-rata to their position.

In other words, for the current winners to not win would need a 100 point PN swing from the mid fleet guys and 200 points from the backmarkers. None of the Class related swings proposed were anywhere near that - especially for those classes with decent confidence, and so there was / is little prospect of big results swings in most races.
The result of the changes we've made has been closer results, with more chance of a mix of fleets in the top 5 rather than a 'Class A N other' whitewash. The same classes still win when you'd expect them to (when conditions suit them) - they just don't dominate the all top places quite as much.

It's worth noting that when we agreed when implementing the changes that we'd review after  the summer series (on the basis that if they were wrong or the racing was spoilt, that we could simply go back to National PNs). In fact, the changes have been widely accepted - even welcomed and increasingly thought a great thing. So much so that we've been encouraged to implement the latest updates in time for our winter series - rather than waiting until next season - we are that pleased with and confident in them!Clap
These latest changes are already more subtle - indicating that the PNs recommended by the PYS system last year weren't far off the mark.

I'd encourage all clubs to consider implementing the changes - at least in part.
For clubs who want to wait  - why not run a shadow series with the adjusted results - discuss those and aim to implement as soon as possible - mid season shouldn't be a problem.

Cheers
Keith



-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: AlexM
Date Posted: 02 Nov 12 at 2:13pm
RYA Technical Approved" stamp - Yes i've only just seen this too. That's why I was after the graph so I could take a print screen as backing. I think the stamp appears at 0.6 and above.

That's a shame about your Sailing Committee especailly as you've got really good data for the GP and Mirror. And lets be honest the changes to be applied hardly make any significant difference to results.

Edit - better post from Keith while i was typing.

-------------


Posted By: JimC
Date Posted: 02 Nov 12 at 2:44pm
Originally posted by Hector

It is an average of some very iffy returns from the last decade and beyond,

You must have misunderstood. It isn't. They don't use returns more than three years old.


Posted By: MattHarris
Date Posted: 02 Nov 12 at 3:04pm
I think there was definately a degree of a fear of change, the concerns coming from those who thought they had most to loose (a dropping handicap).  There are also quite a few classes only sailed by one or two individuals which brought up a discussion on whether it was a personal handicap, again a valid point.

The other points which didn't help were rescoring the previous series results and whilst everything became much closer (3 points separating the top 4 vs 7pts) there was no overall change.  This lead to a query of why bother if its going to cause an upset? - which I can fully understand.  Equally the other problem was adding 3 pts to a streaker.  Whilst mathematically there is virtually no difference (something like 9 secs an hour?), everyone knows the streaker already has a generous handicap on brand new boats and so it leads to a drop in confidence of the other numbers produced.
 
 
It was very good to be able to say that L&L (who are seen as a larger local club) are using their numbers and with more clubs addopting them it may be easier to implement the numbers next year.  We shall see, either way I look forward to how the numbers change in March with what should be a large amount of real race data available.


-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: AlexM
Date Posted: 02 Nov 12 at 3:59pm
Matt, could you send me the pdf of your classes over 0.6? Thinking of running the idea past the committeeto see if we'd use some of your data (not that we have many GP's)
Alex

-------------


Posted By: Oli
Date Posted: 02 Nov 12 at 4:00pm
we pushed really hard at the beginng of the year to implement our schedule of using any new number that has a cf over 0.3 at the beginning of every series.  i will find the list of changes and post them here later.  it has worked quite effectively, and keeps people on their toes! 
we made a rule that if you win 3 consecutive races and your handicap is below the 0.3cf we would change the handicap to whatever number was given, as yet no one has managed that, please draw your own conclusions... is it stable numbers? unstable sailors? bandits dont exsist?Wink


-------------
https://www.facebook.com/OJSPhoto" rel="nofollow - ojsphotograpghy
https://www.youtube.com/ojsphotography/" rel="nofollow - youtube


Posted By: Hector
Date Posted: 02 Nov 12 at 4:20pm
Originally posted by JimC

Originally posted by Hector

It is an average of some very iffy returns from the last decade and beyond,

You must have misunderstood. It isn't. They don't use returns more than three years old.


Apologies - what I meant to say was that the historic numbers were kept the same or very similar by decades of iffy data. The reason the numbers that we're all familiar with and that hardly ever changed or only did so a little was because there was
  1. A very low rate of returns (I seem to recall about 25% of clubs).
  2. Of the returns, the majority (ISTR approx 70%) returned  'No change'.

On investigation, the No Change returns were usually because the results person hadn't done the labourious task of calculating the actual return. So a majority of returns constantly weighted the average towards No (or little) Change - but with no actual analysis of results.

Of those returning a change, there were few who'd calculated actual PNs - again most made an educated guess.

Or at least thats what I understood from the RYA presentation(s).



-------------
Keith
29er 661 (with my daughters / nephew)
49er 688 (with Phil P)
RS200 968
Vortex (occasionally)
Laser 2049XX


Posted By: sargesail
Date Posted: 02 Nov 12 at 8:27pm
What I have found very interesting looking at the PY recommendations results for my classes which have been posted here for various clubs, (the RS300 and the L2K), is that (unsurprsingly) at the clubs with very good sailors the recommendation is a drop, and for those where they are average a rise.

Especially in the 300 the population of sailors is small enough that I am confident that we are handicapping the individuals not the class.


Posted By: pondmonkey
Date Posted: 06 Nov 12 at 10:02am
Originally posted by sargesail

Especially in the 300 the population of sailors is small enough that I am confident that we are handicapping the individuals not the class.

almost certainly... and isn't that the fundamental paradox of this entire system?  The boats who need this system the most are the ones whose data is so unrepresentative of the 'class average'.  

I know our own club data is fudged by blending in results from the fleet boats who aren't even racing in the handicap fleet.  Somehow this is supposed to give our data some added credibility, but I see at nothing more than a farce and ultimately a disservice to those who are actually relying on credible data to help improve their PY racing.   


-------------


Posted By: Contender443
Date Posted: 06 Nov 12 at 1:29pm
Originally posted by pondmonkey

Originally posted by sargesail

Especially in the 300 the population of sailors is small enough that I am confident that we are handicapping the individuals not the class.

almost certainly... and isn't that the fundamental paradox of this entire system?  The boats who need this system the most are the ones whose data is so unrepresentative of the 'class average'.  

I know our own club data is fudged by blending in results from the fleet boats who aren't even racing in the handicap fleet.  Somehow this is supposed to give our data some added credibility, but I see at nothing more than a farce and ultimately a disservice to those who are actually relying on credible data to help improve their PY racing.   
Local conditions and course sailed also come into play as well. I am sailing against well sailed Laser 2s on triangle sausage courses. When they can trapeze and carry their kites on a reach I don't stand a chance in the 100. It also means they can trapeze the beats and are fully powered up when I am de-powering. Then the runs are invariably with the tide so I spend more time beating than running...
 
Blinking Laser 2s.....(mild rant over).


-------------
Bonnie Lass Contender 1764


Posted By: Andymac
Date Posted: 06 Nov 12 at 2:53pm
Originally posted by pondmonkey

I know our own club data is fudged by blending in results from the fleet boats who aren't even racing in the handicap fleet.  Somehow this is supposed to give our data some added credibility, but I see at nothing more than a farce and ultimately a disservice to those who are actually relying on credible data to help improve their PY racing.   
 
I don't know about that.
I would be the first to admit that I'm no expert on data/statistical analysis, but I would have thought that a seperate 'fleet' racing around the same cans at/around the same time as the handicap fleet can only help endorse the other numbers (think of it as a 'control sample').
Perhaps you have stumbled upon an additional data source which clubs do not normally submit...


Posted By: pondmonkey
Date Posted: 06 Nov 12 at 3:01pm
Originally posted by Andymac

Originally posted by pondmonkey

I know our own club data is fudged by blending in results from the fleet boats who aren't even racing in the handicap fleet.  Somehow this is supposed to give our data some added credibility, but I see at nothing more than a farce and ultimately a disservice to those who are actually relying on credible data to help improve their PY racing.   
 
I don't know about that.
I would be the first to admit that I'm no expert on data/statistical analysis, but I would have thought that a seperate 'fleet' racing around the same cans at/around the same time as the handicap fleet can only help endorse the other numbers (think of it as a 'control sample').
Perhaps you have stumbled upon an additional data source which clubs do not normally submit...

Well it's of course just opinion, but I would have thought that generating data by stealth means, when the competitors themselves don't know they are racing each other, can only produce an anomaly and is hardly representative of a boat-on-boat competitive sport.  Although for racing against a spreadsheet, I guess it something more to argue over in the bar, therefore adds 'intrinsic value' to the PY offering as it stands.

Even at my very modest level of competence, the tactics I'd employ in a handicap race and significantly different to that of a fleet race.  Compound this by the fact that the air disturbance, wind shift pattern (and if you were on the sea) tidal flow of all sailing together is grossly different from splitting out over 'categories' separated by whole starting sequences, well I think you've got a minefield of crap data.  

But I'm no stats whizz either... I'd just go with basic premise that if folks don't know they are racing each other, then I'm afraid it's not a race and any data generated is a load of old cobblers.

  


-------------


Posted By: MattHarris
Date Posted: 06 Nov 12 at 3:18pm
I'd agree with you, if boats aren't starting in the same flight together then you shouldn't be putting the results together as if they had sailed the same race.  If they all started at once and then fleets were drawn away from this thats fine. 
However from what you've described the results mean absolutely nothing as if they weren't jockying with each other at the start or obstructing each others wind/tactics - in short they weren't racing each other.  Its like taking the results of the first lap of qualifying on F1 and suggesting theres no difference between this data and the first lap of the actual race.

-------------
Phantom 1175 - Alice

http://www.thelostpenguin.co.uk" rel="nofollow - http://www.thelostpenguin.co.uk


Posted By: pondmonkey
Date Posted: 06 Nov 12 at 3:24pm
Originally posted by MattHarris

I'd agree with you, if boats aren't starting in the same flight together then you shouldn't be putting the results together as if they had sailed the same race.  If they all started at once and then fleets were drawn away from this thats fine. 
However from what you've described the results mean absolutely nothing as if they weren't jockying with each other at the start or obstructing each others wind/tactics - in short they weren't racing each other.  Its like taking the results of the first lap of qualifying on F1 and suggesting theres no difference between this data and the first lap of the actual race.

absolutely +1 - it's not a good data set, thus does a disservice to those relying on quality data to help improve the system.  


-------------


Posted By: marke
Date Posted: 06 Nov 12 at 5:34pm
So you've done the analysis to prove that its a bad dataset then have you?

I looked into this issue about a year ago and compared performance from boats when racing on handicap and in class racing (we do both).  I used almost 8000 results from 4 years worth of races.  I couldn't see any trends that would point to different performances from when a boat is sailing on single start handicap race, single start class race, or multiple start class race.  I was trying to work out what was wrong with the laser handicap at our club - but looked across several classes.

Of course I know that on a forum anecdotal evidence always trumps real data.Wink

I think you might be surprised at how many clubs do what draycote do when submitting results - consolidating fleet and handicap results as long as the boats were sailing the same course at roughly the same time.  For many clubs it is easier to do that - as it happens we don't consolidate, but I don't worry about consolidated results skewing the analysis.  You can't honestly believe that this dataset is worse than a club submitting a "no change" return because they can't (or can't be bothered to) submit real data -  as Keith indicated above this is what really does the damage.

Mark


Posted By: pondmonkey
Date Posted: 06 Nov 12 at 6:26pm
Mark, I don't need to run the stats to know that boats which have not signed on to race together are NOT racing; neither under the RRS or from basic common sense that would suggest that people need to know they are racing to be complicit in the act.

I had not realised that what Draycote did was so commonplace and to be honest, this only goes to further consolidate my opinion that PY is a wholly unsatisfactory method of running proper racing. I am not suggesting PY is a cause, by rather a reaction to widespread consumer choice, and whilst ordinarily I'm all for free market economics, I do wonder if something has been lost at a lot of clubs by the reduction in fleet racing and in turn, this accounts for dropping participation in the sport as a whole.

Anyway, this is well off topic, I'll leave the number crunching to sailing's bean counters... PY serves them very well.

-------------


Posted By: Andymac
Date Posted: 06 Nov 12 at 10:12pm
Thank you Mark for that lucid post.
 
Jimbo, to use your F1 analogy, think of it as a F1 'race' going on around a race track at the same time as a menagerie race that started 30 seconds behind... everyone will be 'racing' for track position at some point whether or not they are in the same 'race'.
The staggered starts or 'flights' is very much a red herring when you also consider that many clubs use average lap times, which mean a faster boat that started in a different flight ahead of a slower boat could actually have a longer race time. It will all average itself out in the end.


Posted By: pondmonkey
Date Posted: 06 Nov 12 at 10:37pm
It's Matt analogy, not mine, but it doesn't matter, I agree with him... Sorry boats that don't even know they're racing are not racing; either under RRS or basic logic. Or has PY racing reached a new all-time low where finally at least no one is pretending there's a race on!!!

It's a joke right???

-------------


Posted By: Andymac
Date Posted: 06 Nov 12 at 11:23pm
Apologies for the misappropriation.
I think you are missing the point that I am trying to make, in that everybody is 'racing' around the same course (if not each other) at the same time. i.e. trying to get around the course as quickly as possible.
I don't understand why you would accept the results between say a 5o5 and a Mirror dinghy which briefly 'started' a race together, whilst dismiss a fleet of twenty SMOD's racing around the same course at the same time.


Posted By: sargesail
Date Posted: 06 Nov 12 at 11:53pm
Ok - let's try this Andy...

The 5o5 and the Mirror start together and they sail the course.  The difference between their elapsed times and exposure to conditions is largely driven by their performance difference.  (And there are those eg HISC which quantify this for tide in their handicap racing by adjusting PYs).

It is possible to argue that the Mirror is overall likely to suffer by sailing in a dying breeze more than the 5o5 across the years (timescale required to build data)....if you accept this you would also have to accept that it is factored into the PY already....but this is a side issue (until we come to separating their start time).

Now I would be the first to say that I have no qualms with the Draycote Dash methos of producing a single set of results for three fleets which start over a 6 minute period (although it can produce some anomalous results as evidenced by the race last year in which the later starting slow fleet caught most of the medium and fast fleets).

So we need a way to remove the anomalies....how do we judge them?

Then how long is acceptable?  Let's say the fleet has two general recalls (it's a fleet so even in Clubracing it happens) on a 5 minute start sequence....so the separation is at least 15 minutes.  Still OK?  Let's say the Mirror takes 20 minutes longer to sail the course - the 50% takes an hour -  that's fully 35 minutes of different conditions or half the 505s race length...

If that's OK what about 505s do the morning and Mirrors the afternoon - still reasonable?

To me if the separation is more than you could reasonably bring together the handicap and fleet times and produce one set of results then the data is flawed...




Print Page | Close Window

Bulletin Board Software by Web Wiz Forums® version 9.665y - http://www.webwizforums.com
Copyright ©2001-2010 Web Wiz - http://www.webwizguide.com