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  <1 89101112 14>
Author
Andymac View Drop Down
Really should get out more
Really should get out more


Joined: 04 Apr 07
Location: Derbyshire
Online Status: Offline
Posts: 837
Post Options Post Options   Quote Andymac Quote  Post ReplyReply Direct Link To This Post 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...
Back to Top
pondmonkey View Drop Down
Really should get out more
Really should get out more
Avatar

Joined: 12 Aug 11
Online Status: Offline
Posts: 2202
Post Options Post Options   Quote pondmonkey Quote  Post ReplyReply Direct Link To This Post 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.

  


Edited by pondmonkey - 06 Nov 12 at 3:03pm
Back to Top
MattHarris View Drop Down
Groupie
Groupie
Avatar

Joined: 03 Oct 06
Location: United Kingdom
Online Status: Offline
Posts: 94
Post Options Post Options   Quote MattHarris Quote  Post ReplyReply Direct Link To This Post 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
Back to Top
pondmonkey View Drop Down
Really should get out more
Really should get out more
Avatar

Joined: 12 Aug 11
Online Status: Offline
Posts: 2202
Post Options Post Options   Quote pondmonkey Quote  Post ReplyReply Direct Link To This Post 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.  
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: 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


Edited by marke - 06 Nov 12 at 5:36pm
Back to Top
pondmonkey View Drop Down
Really should get out more
Really should get out more
Avatar

Joined: 12 Aug 11
Online Status: Offline
Posts: 2202
Post Options Post Options   Quote pondmonkey Quote  Post ReplyReply Direct Link To This Post 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.
Back to Top
Andymac View Drop Down
Really should get out more
Really should get out more


Joined: 04 Apr 07
Location: Derbyshire
Online Status: Offline
Posts: 837
Post Options Post Options   Quote Andymac Quote  Post ReplyReply Direct Link To This Post 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.
Back to Top
pondmonkey View Drop Down
Really should get out more
Really should get out more
Avatar

Joined: 12 Aug 11
Online Status: Offline
Posts: 2202
Post Options Post Options   Quote pondmonkey Quote  Post ReplyReply Direct Link To This Post 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???

Edited by pondmonkey - 06 Nov 12 at 10:54pm
Back to Top
Andymac View Drop Down
Really should get out more
Really should get out more


Joined: 04 Apr 07
Location: Derbyshire
Online Status: Offline
Posts: 837
Post Options Post Options   Quote Andymac Quote  Post ReplyReply Direct Link To This Post 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.
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: 1334
Post Options Post Options   Quote sargesail Quote  Post ReplyReply Direct Link To This Post 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...

Back to Top
 Post Reply Post Reply Page  <1 89101112 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