Results and Rankings.

[HOCOLITE]

My ranking points in most events are determined by an M65S but when my course is not linked with these people ie on one of the colour coded specs I should do very short green alongside the W70S etc, in this if I did it I would expect to be near the top so would I get nearly a 1000 like the top M65S or some toned down figure.
When comparing my results with my age class I'm not seeing how well they've done on the course I'm seeing how well they've done against some random man.

I only comment as it seems to be that these older men are very concerned about their points and I wondered how much it would skew my points compared to my age class if I just went to events and did Very Short Green.

[sloaner]

the leading M45 runner has all his ranking points from regional events - none from JK & BOC

Nobody on the rankings list has any points from BOC - points from BOC 2009 (held in February) dropped out months ago and points from BOC 2010 last saturday will be incorporated on Thursday....And those 6 "bog standard" regional events you mention include 1 National event and 1 National Championship...

But Clive came 4th at this years British only 90s behind the winner ( a pretty good run in most peoples world), his ranking points are less than 1300 (the ranking points are on the BOF website), lower than his 6 current scores.... so Jon's statement still stands .......

As does my argument......

[Adventure Racer]

I'm sure that in golf, you would score more points for winning the British Open or Masters than you would the Portuguese Open

A professional sport where "ranking points" are based on prize money (correct me if I'm wrong - I don't really follow golf) doesn't really bear any relation to an amateur one. In the former case of course the best players chase the money, and ranking lists don't really have any meaning beyond the top 20 or 30. In the latter case, people care about whether they're 1004th or 1005th on the list, and there's no such big incentive for them to do one event over another if personal reasons get in the way.

if the ranking list has any meaning there should be a higher weighting on the bigger events.

On the contrary - a ranking list with weighting on the bigger events would have less meaning, unless you're looking for a measure of who does well at big events. The problem being, as argued previously, it then rewards participation - the person who goes to BOC and JK and has mediocre runs gets more points than somebody who can't make either (or even makes a big mistake or mispunches), yet goes round having superb runs at other events. Not the point of an all-event ranking list at all. Of course if you do want to know who does well at big events then just look at the BOC and JK results. I'm really not sure what the big deal is with a BOC win not boosting your ranking points is - I doubt leading the M40 rankings will ever be more important than a BOC win.

[distracted]

You people are arguing like you expect the ranking list to be perfectly accurate and a reliable indicator of everyone's performance at every event. That a fantastic run at a championship event will be rewarded accordingly, and we'd know the expected finishing order even before the competition takes place, irrespective of where it is in the country.

Then you could use it for seeding at major events etc

Oh...

[greywolf]

Sloaner, it has been said before, but, if you want to know who won the championship races, go look at the results of those races. If all the rankings list did was reproduce those results it would truly be pointless.

Instead, it attempts to rank performances, and then to use these to rank individuals on their six best performances over the preceding 12 months. Analogies to other sports are always questionable, but try athletics, where rankings lists, whether world or national or whatever, are based purely on performance, regardless of the status of the event at which they were acheived. Here's the UK Athletics rankings lists for you to check - the fact that Chris Thompson came sixth in some random Californian track meet doesn't diminsh the quality of his performance or make it less impressive than whoever wins the AAA...

Of course in Athletics there is an objective standard - stopwatch or tape - to measure performance, whereas as in orienteering we have to make do with comparing a individual's performance with that of the rest of the field at a given race, hence the complex algorithm and all the messing about with seed data and standard deviations. And that's also why the list ranks individuals on the basis on their best 6 events, to try and level out some of the inherent variability in scores from this sort of system.

Whether the existing algorithm is up to the job, and whether the recognised anomalies in the system will get ironed out over time, are entirely different (and very valid) questions...but the point is that the assumption that winning a given race is automatically the best performance of the year and will automatically earn the most points woudl be false regardless of the algorithm employed.

As for the "only trying hard at major events" argument advanced by Big Jon and others, that might possibly hold true for the elite, and the very best in the respective age classes, but the rankings list is explicitly there to encourage / amuse / inspire / provoke banter amongst the thousands of other less illlustriated orienteers, and let's face it most of us are basically crap at this infuriating sport - it's very rare that mistakes constitute less than 10% of my race time (sprint/urban races maybe 5 - 10%) and this has a far greater effect on my result than whether or not I try and run especially hard because it's a championships...

Btw, I haven't a clue about Golf, which IMHO is less a proper sport than Underwater MTB-O in Drag, but for what it's worth in Tennis it's not unknown for players (e.g. Rios, Safina) to get to Number 1 in the rankings without ever winning one of the 4 Grand Slams, and conversely for multiple championship winners (e.g Mandilikova) to never get to Number 1

[sloaner]

By the way, I think you'll find that in tennis that the grand slams have a higher weighting when it comes to ranking points... and your athletics analogy is very poor... the AAA's is pretty poor standard in quite a few events...... nuff said

[greywolf]

... so Jon's statement still stands ....

well if you think that 2* National events & 1 National Champs are just Regional events then I guess it does...

*my mistake i forgot the CSC Final

[Roger]

... the leading M45 runner has all his ranking points from regional events - none from JK & BOC (where he has performed very well).

Er, no he didn't: Clive missed the JK Sprint, and although he won (by one second) on JK day 2, he was disappointed with his run on day 3 and also at having dropped 4-5 minutes in the quarry at BOC. So even once the BOC results are included in the ranking system, his sub-par run there quite rightly won't be one of his scoring six.

From the perspective of Deeside you may not realise quite how Clive wipes the floor with most of us week in, week out. I was within 3 minutes or so of him at the British, but it's by no means unusual to find myself 10 minutes down when I've not done much wrong. And I wonder how many points he'd have scored for his BOKtrot run if there had been another couple of people on the black course.

[IanD]

your athletics analogy is very poor... the AAA's is pretty poor standard in quite a few events

Whereas of course all the top M21's were at the British Championships rather than the Tio Mila?

[Spookster]

There were 12 ranking events in April (7 Regional plus 5 National). Eleven of these have submitted their results to the results/rankings system, so the remaining one (NOC Nottingham City race) is the only missing ranking event. I'll give them a nudge.

[IanD]

I wondered how much it would skew my points compared to my age class if I just went to events and did Very Short Green.

At the Regional event following the Harvester, I ran Short Green instead of my normal Short Brown. I didn't want anything longer having already run 9.3km earlier that morning.

I did not have a good run. My legs were tired before I started, and I made significant errors on two controls. I do not think I gained any advantage from having run in the same area in the dark, and I only had one easy control repeated.

Therefore I do not think I should have got a high ranking score, but in fact I have, my highest score by a massive 90 points. I earned more points for doing 8.4 mins/km on Short Green than did Andrew Powell for winning the Brown at 6.6 mins/km.

This suggests to me that the ranking system is seriously flawed, either the algorithm is wrong, or the initial values were skewed in favour of people running shorter courses (so that my competitors on Short Green were over-ranked).

[greywolf]

This suggests to me that the ranking system is seriously flawed, either the algorithm is wrong, or the initial values were skewed in favour of people running shorter courses (so that my competitors on Short Green were over-ranked).

The latter has been acknowledged (and mentioned here a number of times), and perhaps the algorithm struggles when one competitor is so far ahead of the field (who in this case were all fairly bunched) - but the algorithm isn't new... Were the majority of competitors scores unreasonable? and does the occasional freak score when someone runs way out of their usual class constitute seriously flawed?

[Roger]

...does the occasional freak score when someone runs way out of their usual class constitute seriously flawed?

Yes, I'd say that it does: that anomalous score, which sounds as though it's at least 120 more than it should be, will be included in IanD's total for the next 12 months. True, the effect of including this on other people's future scores will be moderate (approx +2 points per run if he's one of ten on his usual course, less if there are more competitors), but his own ranking position will be boosted significantly.

As you indicate, it would be interesting to investigate whether the anomaly is due to the algorithm itself or to the seed data.

[greywolf]

The effect on others will be miniscule...especially as Ian is (i think) one of those who does loads of events a year, so this one score is "diluted" (it's average of all scores, not best six, used in the calculations)...and events with < 10 competitors are no longer scored. But it's a fair point that Ian's personal ranking will be significantly affected for 12 months.

We (think we) know there's a seed data effect...although it's difficult to tell whether it is entirely responsible for this size of anomaly...but fixing it would require rewinding and starting again, I'm not sure now much BOF enthusiasm there is for that...

As for the algorithm, some (inc. Graeme) have said all along that using SD as the basis for the rankings was a problem... unfortunately perhaps the Rankings Working Group ToR explicitly excluded "establishing the details of an operational ranking algorithm" - so the old algorithm was adopted...all that's changed is that the data presented to it is being done so by course rather than being split up into loads of classes each of which was calculated separately. Working with bigger populations should make it more less susceptible to fluke results ...however i guess it never had to handle individuals who were 3*SD away from the mean before...

[Big Jon]

[quote="As for the algorithm, some (inc. Graeme) have said all along that using SD as the basis for the rankings was a problem... unfortunately perhaps the Rankings Working Group ToR explicitly excluded "establishing the details of an operational ranking algorithm" - so the old algorithm was adopted...all that's changed is that the data presented to it is being done so by course rather than being split up into loads of classes each of which was calculated separately. Working with bigger populations should make it more less susceptible to fluke results ...however i guess it never had to handle individuals who were 3*SD away from the mean before...[/quote]

I'm not 100% sure what this means in plain English, but I suspect its a case of garbage in, garbage out!