News   |   Events   |   Forum   |   Photos   |   Links    


Forum Home   Start New Topic   Edit Profile   Register  

1   2   3   4   5  

StatNav - New Zealand's Unofficial Orienteering Rankings

Show Profile  Chris Forne Posted: 14 May 2013, 9:07 AM  
Cool !!!!

Show Profile  Dwayne Posted: 14 May 2013, 9:55 AM  
Of course the current ranking leader would like it

Show Profile  thomasr Posted: 14 May 2013, 3:23 PM  
Well played statnav.

Show Profile  AlisterM Posted: 14 May 2013, 9:03 PM  
Good job StatNav!
There has been an amazing amount of work gone into this with the compiliation of results from many events presumably sourced from WinSplits and Club websites. Would StatNav like their job made easier by clubs submitting OE2010 result files?

Show Profile  Marquita G Posted: 14 May 2013, 9:24 PM  
So who exactly IS StatNav??

Show Profile  StatNav Posted: 14 May 2013, 9:29 PM  
Alister, thanks for the offer but I have automated the process of getting the results off WinSplits so I probably wouldn't get too much extra benefit from OE2010 result files.

The one thing that Clubs could do to make things easier is to clearly indicate, either in the WinSplits results or in timely OY Points tables, which new or infrequent runners are Club members and which ones are just casual entries.

Show Profile  Bryan Posted: 15 May 2013, 3:24 PM  
I agree - very cool - and I know as Statistician the amount of work that has gone into creating this (after collating results for 25 years for the NZOF).

An apology for not updating the 'offical' ranking system for quite a while now - for the last 2 years I have been too busy mapping the world cups and I've just come back from mapping in Cape Town and I'm currently involved with drawing three South African maps. I have not been able to spend as much time as Statistician as I have wanted.

As Statisician, rankings are important but I am more keen on storing all vital results information for major events like event details, distance, climb, club, organisers, location, map details and a scanned image of all orienteering maps.

The website is currently missing results for the last 2 years - I have saved results on my PC but have not got around to loading the information yet. I want to rectify this and your methods of automating the loading of results may be useful to me.

Winsplits Pro (even free demo version) allows export as a txt file - it is a good way to get consistent standardised results from the last few years. Keen to known what database you store the results in, what software you use and what data you store and whether you are willing to share the data/software you have created.

The current 'official' ranking system (which hasn't been updated for quite a while now) is based on the Swedish system and is only on major national events (including sprint events and excluding regional events). In my elite days, the Auckland OYs were treated as training events by the top elites so rankings for these events would have been skewed. I feel this also occurs today and to get a more meaningful 'national' ranking, ranking events have to where competitors are trying hard. The current 'offical' system allows a ranking if a competitor has only 4 counting events.

Your 'unofficial' rankings show a slightly different picture and is perfectly valid from a Statician's point of view. All rankings are artifical and many times statistics can sometimes not show the real picture (as quipped by theoman - 'Real orienteering is done in the forest').

I see no reason why both ranking lists can't exist side by side and your 'unofficial' renamed to 'official' - eg a national major events list and a national regional list. Also if people really love your system, why it can't replace the official.

I don't know why you don't won't your real name known - you should be applauded for the amount of work you have done.

Show Profile  theoman Posted: 15 May 2013, 3:29 PM  
Statnav is like batman. He's the hero we need, not the hero we deserve.

Show Profile  onemanfanclub Posted: 15 May 2013, 3:49 PM  
"who was that masked statistician?"

Show Profile  pete s Posted: 15 May 2013, 8:19 PM  
That is fantastic!!

Show Profile  mark Posted: 16 May 2013, 8:59 AM  
I know who statnav is.
Or at least I know who the owner of the domain name is.

Show Profile  magnus Posted: 16 May 2013, 12:00 PM

Show Profile  hughff Posted: 16 May 2013, 7:39 PM  
When the NZOF was doing its review for strategic planning, I advocated that up to date rankings would be helpful. If nothing else, they serve as motivation for people wishing to improve but I also think they are very useful for selectors. Therefore I'm delighted that someone has filled this void. (Bryan, according to his post above, has been more archivist than statistician in the last couple of years, a priority that I also heartily endorse.) However, I have three qustions that I wish to ask Statnav:

1. Do you ensure that classes who run the same course are mixed together as a single race or do you treat them as different races? Obviously if they ran the same course on the same day they were effectively contesting against each other, even though their results were published in seperate classes. For example, at the Nationals Long, course six covered the classes M16A, M21AS, M55A, W40A and W45A. Only two of those classes (M16, M55) had 10+ finishers so were the winners of the other races (M21 Bryn D, W40 Antonia W, W45 Marquita G) given points for those races despite fields of 7, 5 and 9? Similarly, in the HB OY series, the red short, red short vet and red short super vet all run the same course so I believe they should all be lumped together.

2. Why do you exclude the bottom 10% of results? Surely it's not because they're outliers? Because if that's the case, then logically the top 10% are just as outlying and should also be excluded - which is silly. If someone runs the race and has a disasterous time, it's still their time and it's still a valid race result. I think it should be calculated. I think the effect of discarding the bottom 10% would also be exaggerated if you treated each class as a different race. (See question 1 above.) In a perfect world there would even be a way to take mispunches into account when calculating rankings - someone who frequently mps but is very fast might have a very high ranking despite only finishing half his entered races - but I cannot see how to include them.

3. Why not publish your sprint rankings too? I wonder about the logic of not including them on the site. The IOF has made it fairly clearly that they consider sprint ori to be significant and, indeed, if ori were to crack the Olympics and public notice generally, sprint ori will be the way. (The analogy I would draw is to one day cricket, which was the pathway to increase cricket's mainstream spectator appeal. The shorter form is usually more accessible to John Q. Public.) I also don't see STB as being the only valid sprint results. There are sprint courses at NZ Champs, Auckland, Wellington, North Island, South Island champs, NZ schools, Oceania champs - all of which would be valid to track, I would have thought. The HB OY series includes a Sprint event as well. I'm not suggesting it should be mixed in with the other rankings but I believe you should be able to publish both sets of rankings.

Show Profile  StatNav Posted: 16 May 2013, 8:34 PM  
Hugh, some brief answers for you:

1/ Yes, all classes running the same course are considered a single race. The two examples you give (Nationals Long Course 6 and HB OY Red Short) are good ones to help visualise this.

2/ I don't think I've explained this very well and I'm not sure how to do it any better. Basically, excluding the top 10% as well as the bottom 10% would be appropriate if the times were a Bell Curve but they aren't. There is a long tail I'm trying to avoid. Note that the bottom 10% still get allocated points, they're just excluded from having their existing score taken into consideration when calculating the points.

3/ I'm with Brian on this one. I've got Sprint rankings going back to Sprint The Bay 2011 sitting on my computer but I haven't got around to publishing them. I'm open to feedback on this one - should Sprints be included in the "Classic" rankings or should they stand alone? And how do we deal with the dominance of STB, do all six races count or should just one, two or three? If all six, then given the geographic spread of the other five or six major sprints during the year the Sprint rankings would probably mirror the STB results.

Show Profile  Michael Posted: 17 May 2013, 12:38 PM  
3/ It's worth doing sprint (or I would suggest urban) rankings separately if they would produce significantly different results. To find out, we'll have to do it, won't we?

1   2   3   4   5  

Ruffneck Productions © Ruffneck Productions