New Posts New Posts RSS Feed: Dynamic ratings from Sailracer.
  FAQ FAQ  Forum Search   Register Register  Login Login

Dynamic ratings from Sailracer.
 Post Reply Post Reply Page  <1234 6>
Author
  Share Topic Share Topic  Topic Search Topic Search  Topic Options Topic Options
423zero View Drop Down
Really should get out more
Really should get out more


Joined: 08 Jan 15
Location: United Kingdom
Online Status: Offline
Posts: 3420 		Post Options Post Options   Quote 423zero Quote  Post ReplyReply Direct Link To This Post Topic: Dynamic ratings from Sailracer.
    Posted: 29 Dec 17 at 9:49pm
yes, weighed and then the matching PY for your particular chosen class, Toppers minimum and maximum weight range etc.
Class and handicap racing, then it will be your results (personal handicap), all the way through race everyone constantly monitored.
423zero, boat 'Sprint', weight 11 stone, last six positions, Personal PY for this race, class/handicap.
Back to Top
sargesail View Drop Down
Really should get out more
Really should get out more
Avatar

Joined: 14 Jan 06
Location: United Kingdom
Online Status: Offline
Posts: 1459 		Post Options Post Options   Quote sargesail Quote  Post ReplyReply Direct Link To This Post Posted: 29 Dec 17 at 10:06pm
Firstly I don’t accept the opening premise that handicap racing is the bedrock of dinghy racing.

It’s something that Sailracer have been pushing for a good while....but then it’s the basis of their business.

If there was a little more about how the slow boat had its day in strong winds too then that would be reassuring.

And I am absolutely sceptical about the 6% for an old solo, if it doesn’t include Crew Skill Factor.

And yes the logical destination would be crew weight based personal handicsaps. Let’s not go there.

Of course such an initiative might drive us all back go fleet racing!
Back to Top
ian.r.mcdonald View Drop Down
Far too distracted from work
Far too distracted from work


Joined: 24 Feb 06
Location: United Kingdom
Online Status: Offline
Posts: 440 		Post Options Post Options   Quote ian.r.mcdonald Quote  Post ReplyReply Direct Link To This Post Posted: 29 Dec 17 at 10:25pm
so you weigh in and get your py. what happens after the bacon sarnie before launching?

has the world really gone mad?
Back to Top
423zero View Drop Down
Really should get out more
Really should get out more


Joined: 08 Jan 15
Location: United Kingdom
Online Status: Offline
Posts: 3420 		Post Options Post Options   Quote 423zero Quote  Post ReplyReply Direct Link To This Post Posted: 29 Dec 17 at 10:50pm
The technology already exists for all types of factors to be programmed into a personal GPS device on whatever boat you happen to be sailing on any given day, (can't see my club taking it up), but for big events class or handicap it's just a decision away.
Back to Top
Simon Lovesey View Drop Down
Far too distracted from work
Far too distracted from work
Avatar

Joined: 30 Nov 04
Location: United Kingdom
Online Status: Offline
Posts: 349 		Post Options Post Options   Quote Simon Lovesey Quote  Post ReplyReply Direct Link To This Post Posted: 30 Dec 17 at 7:55am
Originally posted by mozzy

The irony is the author only knows Dobson should be doing better because of results from class racing. 

We also have results from large handicap events to cross reference,  Ian Dobson has won the Grafham Grand Prix and Stevie Nichs, and been on the podium for the Tiger,  Datchet Flyer and Brass Monkey in the past, in addition to National and World championship wins in major classes.  
http://sailracer.org/ResultsEntry/Result/Sailor?sortOrder=Rank_asc&SailorName=Ian%20DOBSON 

Originally posted by mozzy

Making the handicaps 'dynamic' I think is missing two points. 
1) The numbers are still generated from performances which are the product of boat and crew. No amount of statistics up-speak will gloss over the fact that they are inseparable and without a clear methodology the talk of 'vast amounts of data' , 'standard deviation and confidence interval' is just massaging numbers until the names the author thinks should be winning, are winning. 
I agree that CSF can have an influence,  but that is the same with PY.  We are only focussing on very large handicap events (circa 100) boats,  with a high incidence of multiple boats from each class,  which reduces some of the CSF influence.  

Originally posted by mozzy

2) I don't like the result not being obvious on the water. Corrected time handicapping is bad enough in this regard, but at least the maths are simple enough that you can estimate in your head and understand the calculation to check it ashore. No point being more accurate if people can't understand and believe it.  

I agree that with a small fixed length club handicap race it is possible for sailors to estimate their corrected finish position,  and with similar boats racing each week build up a ready reckoner of minutes you need to beat a slower boat by or minutes behind a faster boat.  BUT for a large average lap race with 100 boats and 40 different classes over three starts,  where Toppers may only manage 1 lap and Moths 5 laps, I think many people would find it a challenge to estimate their corrected finish position.  


Edited by Simon Lovesey - 30 Dec 17 at 7:56am
www.sailracer.org
Online Sailing Results, GPS Tracking & Event Management
Back to Top
Guests View Drop Down
Guest Group
Guest Group
Post Options Post Options   Quote Guests Quote  Post ReplyReply Direct Link To This Post Posted: 30 Dec 17 at 9:38am
I'm not convinced that a variable rating system such as this is desirable - not knowing your PY at the start seems strange and off putting, but maybe this is just my prejudice and resistance to change. Having said that, I do wonder if probabilistic PYs could help. In that method a boat does not have a single PY value but a range associated with some form of distribution. The outcome is that when calculating the results, this distribution is used and a boat may have say a 10% probability of first, 25% 2nd, 50% 3rd and 15% 4th. Then sum the weighted scores.
Back to Top
PeterG View Drop Down
Really should get out more
Really should get out more


Joined: 12 Jan 08
Location: United Kingdom
Online Status: Offline
Posts: 822 		Post Options Post Options   Quote PeterG Quote  Post ReplyReply Direct Link To This Post Posted: 30 Dec 17 at 10:12am
Firstly I don’t accept the opening premise that handicap racing is the bedrock of dinghy racing. 

Indeed, and these attempts to endlessly further tweak handicaps just underline that. If it's really that important to race on a reasonably clear and level playing field then class racingis there, available and works..

Of course there are plenty of people who would choose to sail something that is not locally available in fleet numbers (and others who clearly want to choose a boat they think will give them a handicap advantage!), and that's why we have the PY system. It's a compromise, it's never pretended to be anything else, but it allows people to make a compromise choice between class racing and handicap racing in the boat of your choice.

No handicap system is perfect, and not will it ever be. Of course, handicap systems could and should be reviewed to make them as close a compromise as possible to a level playing field, as they are, and the endless discussions here prove. But if you start using GPS and trying to adjust handicaps for local conditions and even for local conditions of different legs of a race you have to start asking yourself WTF is the point!


Edited by PeterG - 30 Dec 17 at 10:12am
Peter
Ex Cont 707
Ex Laser 189635
DY 59
Back to Top
Guests View Drop Down
Guest Group
Guest Group
Post Options Post Options   Quote Guests Quote  Post ReplyReply Direct Link To This Post Posted: 30 Dec 17 at 10:39am
Originally posted by Simon Lovesey

Originally posted by mozzy

The irony is the author only knows Dobson should be doing better because of results from class racing. 

We also have results from large handicap events to cross reference,  Ian Dobson has won the Grafham Grand Prix and Stevie Nichs, and been on the podium for the Tiger,  Datchet Flyer and Brass Monkey in the past, in addition to National and World championship wins in major classes.  
http://sailracer.org/ResultsEntry/Result/Sailor?sortOrder=Rank_asc&SailorName=Ian%20DOBSON 
The problem is, with the way handicaps are calculated, boats which struggle in certain conditions must be offset by having lenient handicaps in other conditions. So if you point to this result and say he's too low down the fleet, you are at the same time pointing to his previous wins and saying he shouldn't have done so well. 

A better way to highlight the problem would be show how in class racing Dobson has never finished outside the top xx% in class events, making this result a significant outlier and that it's unlikely to be a duff day as he still had a large margin over the nearest fireball. 

I'd prefer it if the article didn't focus on individuals however as it comes across as fudging results until the names you know are winning. It would be better to illustrate the point with class averages. 

Don't the GL handicaps deliberately exasperate this problem by choice anyway?
"We are therefore aiming to handicap to the potential of the boat. An extreme example is the foiling moth, we will handicap for foiling conditions only"
Originally posted by Simon Lovesey

 
Originally posted by mozzy

Making the handicaps 'dynamic' I think is missing two points. 
1) The numbers are still generated from performances which are the product of boat and crew... 
I agree that CSF can have an influence,  but that is the same with PY.  We are only focussing on very large handicap events (circa 100) boats,  with a high incidence of multiple boats from each class,  which reduces some of the CSF influence.  

Originally posted by mozzy

2) ....No point being more accurate if people can't understand and believe it. 
I agree that with a small fixed length club handicap race it is possible for sailors to estimate their corrected finish position,  and with similar boats racing each week build up a ready reckoner of minutes you need to beat a slower boat by or minutes behind a faster boat.  BUT for a large average lap race with 100 boats and 40 different classes over three starts,  where Toppers may only manage 1 lap and Moths 5 laps, I think many people would find it a challenge to estimate their corrected finish position.  
[/QUOTE]
Like I said, it's bad enough as it is. But at least you can check and understand how the result is derived.  Whether that's on the water or when the result go up, this new method seems to push the point of understanding further away from the finish line. 

I like pursuit races, although I don't think the result is fairer, the result is instant, and you get that racing feeling of chasing and being chased on the water. 

I don't doubt that the authors are trying to produce fairer handicaps but what i really don't like about this article and the 'method' page on the GL website is neither actually give a method. They describe the aims; but the devil is in the detail. 


Edited by mozzy - 30 Dec 17 at 11:00am
Back to Top
Simon Lovesey View Drop Down
Far too distracted from work
Far too distracted from work
Avatar

Joined: 30 Nov 04
Location: United Kingdom
Online Status: Offline
Posts: 349 		Post Options Post Options   Quote Simon Lovesey Quote  Post ReplyReply Direct Link To This Post Posted: 30 Dec 17 at 11:17am
Originally posted by mozzy

I don't doubt that the authors are trying to produce fairer handicaps but what i really don't like about this article and the 'method' page on the GL website is neither actually give a method. They describe the aims; but the devil is in the detail. 

Dynamic GPS Handicapping Method 
  1. Using the GPS data (length,  angle,  speed,  time, VMG etc) for each leg a Course Speed Profile (CSP) is calculated for each class.
  2. The database of previous races is then referenced to filter down to similar races 
  3. The dynamic handicap is then calculated from the suggested handicaps for those races (in a similar manner to PY)
We are very much at an experimental stage,  but the early outcomes are encouraging,  showing the faster boats getting a boost in the lighter conditions and the slower boats benefitting when the breeze starts to pick up. 
www.sailracer.org
Online Sailing Results, GPS Tracking & Event Management
Back to Top
Guests View Drop Down
Guest Group
Guest Group
Post Options Post Options   Quote Guests Quote  Post ReplyReply Direct Link To This Post Posted: 30 Dec 17 at 11:37am
But how is CPS calculated? Those variables define the type of leg it is, then I guess you use the time it takes a class to complete a leg compared to other boats to produce a PY.  

What is the output of CPS? Is it a PY number specific for a leg angle and wind force?

Then in the article you talk about 'modes'. Its not clear how this relates to CSP. Do you take the characteristics of the leg to determine what mode the boat should be in for that leg, and have a PY for each of that classes modes? Then their PY for the whole leg will be weighted by the length of each leg and which mode the model determines the class should be in for that leg? 

Where did the met data come from? 
Back to Top
 Post Reply Post Reply Page  <1234 6>

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