| Author |
Message |
    
Mike Watts
Username: Mikew
Registered: 10-2006
| | Posted on Monday, 12 October, 2015 - 12:59 pm: | |
Sorry Alan, tandems get a coefficient of 0.7 - slipped off the list below. |
Alan Hodges
Username: Alan
Registered: 10-2006
| | Posted on Sunday, 11 October, 2015 - 02:11 pm: | |
Mike, where do Tandems fit in because they are not specifically mentioned? |
Mike Watts
Username: Mikew
Registered: 10-2006
| | Posted on Friday, 09 October, 2015 - 06:48 am: | |
Horse pairs scores are multiplied by 0.73 before the percent is calculated, and that does make a horse pair score the best normalised score of the day at Shuttleworth, so it gets 100%, well done them, go horse pairs!
We'll see how these weightings (coefficients) work over the season, but even with the first 4 event results in you can see it evens up the classes somewhat, with an open pony 100% at one event, a novice and a pony pair at others as well as your Bedfordshire horse pair winning. This is probably good, a more level measure....
Also you can see in the points leagues that the top few in each league in the four regions with events so far have easily already got over 75%, the new minimum best score proposed as an additional check for championships qualification.
The exact numbers used are in the thread below about the rule 10.1.
Quite probably the computer will have the odd score glitch, though I hope not with this simple formula...if it does please email me to fix. |
Rosemary Neale
Username: Rosemaryn
Registered: 03-2007
| | Posted on Thursday, 08 October, 2015 - 05:45 pm: | |
Sorry being a bit thick here perhaps but looking at the latest scores from our shuttleworth event i still don't understand how the horse pairs competitor gets 100 percent even though the score was not the best of the day?? |
Mike Watts
Username: Mikew
Registered: 10-2006
| | Posted on Thursday, 08 October, 2015 - 04:43 pm: | |
"Percentage" - nearly simple:
--------------------------------
We take the best score at an event and call that 100%, scale the others off that.
So for any other score, it's basically:
(best at event)/ (score at event) x 100
A bit more....
--------------
Since that wouldn't be quite fair as some classes are faster than others etc, we first multiply every score by a fixed number for its class to try to put the classes on an even basis. Based on analysing a few year's worth of scores we came up with numbers to try to level out the classes for the percentage. (See thread below "10.1" for more detail, including the 'class coefficients')
- it's done event by event (it's not very good comparing raw points across events as obstacle length, going etc varies - so this "Percentage" helps even that out as best we can - there's no perfect way of doing this, but this is pretty straightforward whereas more complex methods probably confuse more and can give weirder effects)
- in the points league for each region, we record for each competitor the best percent they have scored in that class and region in a season.
- the plan is to see how it looks as scores build up - review in the New Year. |
Rosemary Neale
Username: Rosemaryn
Registered: 03-2007
| | Posted on Thursday, 08 October, 2015 - 12:41 pm: | |
Could someone (hopefully Mike) please explain the significance of the percentage score and how this works? Is it a percentage from across all competitors in a class across all regions? |
|