New Posts New Posts RSS Feed: New PY Website
  FAQ FAQ  Forum Search   Register Register  Login Login

New PY Website

 Post Reply Post Reply Page  <1234 14>
Author
moomin View Drop Down
Groupie
Groupie


Joined: 19 Jan 05
Online Status: Offline
Posts: 57
Post Options Post Options   Quote moomin Quote  Post ReplyReply Direct Link To This Post Posted: 14 Nov 12 at 11:30am
Thanks Marke and Alex,
Question answered: it looks for the number of different competitors in the sample - perfect.  No complaints about the new system, a massive step forward.  Thanks for the effort to the team that have put it together.
Moomin
Back to Top
AlexM View Drop Down
Really should get out more
Really should get out more


Joined: 10 Jan 06
Online Status: Offline
Posts: 857
Post Options Post Options   Quote AlexM Quote  Post ReplyReply Direct Link To This Post Posted: 14 Nov 12 at 11:20am
My question is what happens if your data continues to follow the same pattern over a range of 80 points but as you sail more races, you will have more data points. So will your CF increase despite the fact it's still only you sailing and it becomes a personal PY?

If that pattern continued i wouldn't expect the c.f. to move from 0 (I haven't seen this happen in any of our single boat classes) but you are correct if it did it would be a personal h'cap. It does take a lot of data for the c.f. to move and then our club only takes a % of the movement when the c.f. is low (see earlier posts)

Simple question: Does the PY calculator look for a difference between the number of races and the total number of starters, that would indicate there are multiple boats from the same class entered in the races when assigning a CF.

It does - it produces the table below at the bottom of each race you enter:





Back to Top
marke View Drop Down
Far too distracted from work
Far too distracted from work


Joined: 16 Jun 08
Location: United Kingdom
Online Status: Offline
Posts: 209
Post Options Post Options   Quote marke Quote  Post ReplyReply Direct Link To This Post Posted: 14 Nov 12 at 11:18am
moomin

The PYS confidence factor is a composite metric based on
  • number of samples
  • number of different competitors in sample
  • variation in sample over time
The PYS cannot take account of CSF for a club fleet - however if you have a fleet with a range of skill levels then the PYS number for your boat should be a pretty good measure of a local PY.  If it is for a single boat then less so - and you need to think about the implications for any local adjustment (perhaps considering results obtained for you class at similar clubs - you'll need to consider what 'similar' means (classes sailed, water conditions etc.).  However for a class with a single competitor the cf should never get particularly big - only one competitor and surprisingly I would guess that like most competitors your actual PN varies in a range of 50-100 points across a season so that variation over time will be high (= lower cf).

At my club we take the PYS output as a starting point for the local adjustments and then do some analysis of individual data points across multiple seasons to analyse relative CSF values between our main fleets (remarkably similar with the exception of one fleet).  As several others have noted the introduction of local adjustments has resulted in more competitive racing and a greater mix in the results at my club - this may be one factor in higher turnouts.

Remember this is the first year that the PYS has tried the cf - it is very likely to be tweaked and improved as more experience is gained.

mark


Edited by marke - 14 Nov 12 at 11:19am
Back to Top
moomin View Drop Down
Groupie
Groupie


Joined: 19 Jan 05
Online Status: Offline
Posts: 57
Post Options Post Options   Quote moomin Quote  Post ReplyReply Direct Link To This Post Posted: 14 Nov 12 at 11:03am
Thanks Alex,
Not sure you've quite got the jist of what I'm asking.
Your PY in the RS100 varies over a range of about 85 (995 to 910 ish).  The 400 results vary over a similar range 80ish (895 to 975ish).  The 400's have a higher CF factor therefore becuase they have more data points - more data points better confidence, and the  RS100 CF is low because there are fewer data points. The only reason the 400s can have a higher CF is due to the number of data points because from a rough eyeball of the graph the distribution actually looks less stable. 
 
My question is what happens if your data continues to follow the same pattern over a range of 80 points but as you sail more races, you will have more data points. So will your CF increase despite the fact it's still only you sailing and it becomes a personal PY?
 
Simple question:  Does the PY calculator look for a difference between the number of races and the  total number of starters, that would indicate there are multiple boats from the same class entered in the races when assigning a CF.


Edited by moomin - 14 Nov 12 at 11:08am
Moomin
Back to Top
AlexM View Drop Down
Really should get out more
Really should get out more


Joined: 10 Jan 06
Online Status: Offline
Posts: 857
Post Options Post Options   Quote AlexM Quote  Post ReplyReply Direct Link To This Post Posted: 14 Nov 12 at 10:57am
Hadn't noticed that but......... there is a reason for that!
Back to Top
Peaky View Drop Down
Really should get out more
Really should get out more


Joined: 15 Nov 08
Location: United Kingdom
Online Status: Offline
Posts: 2771
Post Options Post Options   Quote Peaky Quote  Post ReplyReply Direct Link To This Post Posted: 14 Nov 12 at 10:55am
Does that suggest that 100's sail faster in warm water? OMG, maybe Graeme was on to something...
Back to Top
AlexM View Drop Down
Really should get out more
Really should get out more


Joined: 10 Jan 06
Online Status: Offline
Posts: 857
Post Options Post Options   Quote AlexM Quote  Post ReplyReply Direct Link To This Post Posted: 14 Nov 12 at 10:38am
Hi Moomin,
Iím not sure that I can answer your question but Iíll give myself as an example. All the single boat classes in our club still currently have a 0 c.f. I think your assumption of a single boat sailed in consistent way is wrong as even if you win every race your position from the mean could vary drastically and then your sailed to PY will be different. Have a look at my 100 PY below, itís all over the place (and I would hope I sail consistently bad ):



When you have lots of boats it takes the average of their sailed to PY and as you can see on our 400ís the graph the range is smaller thus giving us a good c.f.

Back to Top
moomin View Drop Down
Groupie
Groupie


Joined: 19 Jan 05
Online Status: Offline
Posts: 57
Post Options Post Options   Quote moomin Quote  Post ReplyReply Direct Link To This Post Posted: 14 Nov 12 at 9:53am
Can anyone answer this question that is confusing me at the moment.  I thought the new PY system was supposed to ensure the PYs were based on a good sample spread across a fleet i.e. the confidence factor, and avoid personal handicapping.  What I'm reading now is that if I am the only person sailing a particular class then to start off with the CF will be low (due to low number of appearances) but as more races are sailed will the CF rise or does the system recongise that it is a single boat and prevent personal handicaps? 
If the CF based on the stability of the PY then surely a single boat will actually create a higher CF faster than a small group, as the boat will be sailed to a more consistent way. Hence the high CF relating to the Scorpion in previous posts which seem to have been generated by a single good sailor.
Before you start I know PY sailing is not perfect and if I don't like it  then buy a different boat and go fleet racing etc etc.  I enjoy sailing my choices of boats and won't be deterred from sailing based on a PY, more of a stats/data geek interest!
Moomin
Back to Top
Peaky View Drop Down
Really should get out more
Really should get out more


Joined: 15 Nov 08
Location: United Kingdom
Online Status: Offline
Posts: 2771
Post Options Post Options   Quote Peaky Quote  Post ReplyReply Direct Link To This Post Posted: 13 Nov 12 at 11:18am
Exactly. The PYS only rates boats relative to each other, not in absolute terms.

There is a risk that some clubs will pick and mix from other clubs suggested PY's - especially those that have been designated as having a high Confidence Index. For example, take Thornbury's Topper PY, Yorkshire Dale's Vortex PY and Hayling's 400 PY, because the clubs have a lot of 'confidence' in those numbers. But that would be mathematical nonsense.
Back to Top
RS400atC View Drop Down
Really should get out more
Really should get out more


Joined: 04 Dec 08
Online Status: Offline
Posts: 2956
Post Options Post Options   Quote RS400atC Quote  Post ReplyReply Direct Link To This Post Posted: 13 Nov 12 at 11:01am
Originally posted by Peaky




Anyone? Is there a standard boat that is benchmarked off?[/QUOTE]

Apparently not.
I don't think the RYA handle the data properly regarding returns from clubs where most of the handicaps are changed either.
For instance HISC change the numbers, every boat in the fast py races are running on lower py's.
The slow fleet may have increased PY's but 99% of the time they are not in the same race.
So if they get some data that says B14 should be 750, they need to rebase that because the B14 is racing against a Flying Dutchman at 810 or something.

So the info they have is B14 should be 750/810 of the FD PY

The data is being mangled by both clubs and RYA.

Since clubs like HISC do the same thing to their numbers every year, the effect is that boats like the RS400 keep dropping by the same amount every year. But, the Merlin is dropping faster.
Back to Top
 Post Reply Post Reply Page  <1234 14>

Forum Jump Forum Permissions View Drop Down

Bulletin Board Software by Web Wiz Forums® version 9.665y
Copyright ©2001-2010 Web Wiz
Change your personal settings, or read our privacy policy