Rankings

[bendover]

so would it be better if there was a separate category on the ranking lists for the short classes?

[guest]

Well it already is isn't it?

[awk]

However, to me it seems unfair to those who are 'competitive' on their short course and, if they had a good/excellent run, would stand a chance of winning. But then they are denied by someone who could be doing well on a long course, and in cases have shown this to be true.

As a regular short course runner, I have absolutely no problem with anybody running the same class, even when they stuff the rest of us out of sight (and I'm reasonably competitive on it). Winning the S class is pretty trivial anyway, as whatever the technical standard, it IS the B course in all but name. S class is for anybody who, for whatever reason, decides that this the right distance/time for them. My competition is with my peers, not with somebody who is way better (although it's fun measuring against them) or worse, and where I finish in the field overall is not that important. Different if I was running the Ls, but not yet!

What I would like to see are more events where the shorter courses are the premier races.

On the rankings scheme bit, David and I have long agreed to differ, and his post is a model of logical argument (and in fact I agree with much of it!). I did however shudder at the possibility that any "ranking" scheme be used for selections (even seeding would be dodgy).

[HOCOLITE]

Can we call the Short Course Ranking Lists, if we must have them, something other than National Ranking Lists, which they are not, e.g., Short Course League? There can surely only logically be one 'National Ranking List for each age class.

Some people strive and even train to do well at a 'National' level. These 'second class citizens' deserve recognition in a single 'National Ranking List'.

WE've been down this alley before please do not refer to short course runners as These 'second class citizens'. I don't think anyone in the short rankings has any belief they are better than the long course ones and the person at the top does not bragg about being top of the National rankings. As AWK says it gives short course runners something to gauge their performance on. I believe it is you who has a chip on your shoulder if you think they do, are you worried your status is being eroded, by people who run short courses?
AWK short courses should not be 'B' by technical standard they might be by competitor standard.

[Gnitworp]

I'm afraid you've misunderstood me, Hocolite, I was referring to Long Course runners as 'these second class citizens'! Their rights are being eroded.
Apologies for the over-subtlety.

[graeme]

Back at the original question.

The trouble with ranking lists is ... they try to do two incompatible things: Measure how good people are, and be a "fair" competition.

For a stats project, the sensible thing to look at is how to get a ranking list which best measures how good people are. "Fairness" is too subjective. If you're interested in a ranking list which gives a good competition, or encourages people to go orienteering, or rewards the deserving, or uninterested in statistics stop reading now.

If you handed in a stats project which purported to "rank" orienteers by some number, and failed to assign a statistical error to that number, then no respectible university should give you a stats. degree. A statistical measure of anything includes some notion of how accurate its likely to be. So for example...

How many races you run should not affect a measure of how good you are. It does
affect the uncertainty of the measure. An unbiassed statistical measure shouldn't discard poor results and keep good ones (to deal with heteroskedasticity you might ignore outliers at both ends).

Suppose you ignore my previous post, and use the BOF scheme which measures
"number of standard deviations from the mean". Multiplying this measure by different factors from different events is a statistical nonsense. Being average at the JK isn't equivalent to being above average at the White Rose. People (rightly) have the sense that the JK result "means more" than the White Rose because people are taking it seriously. So the JK may be a more accurate measure, in which case the way to incorporate it is by weighting the averaging (e.g. JK counts as two runs, White Rose as 1/2).

Hope this helps,

Graeme

[Gnitworp]

[quote="graeme"] So the JK may be a more accurate measure, in which case the way to incorporate it is by weighting the averaging (e.g. JK counts as two runs, White Rose as 1/2).




Sounds a better way of weighting than multiplying by a factor of 1.1 with all Regional Events multiplied by 1, as used to be the case.

[awk]

As AWK says it gives short course runners something to gauge their performance on.

Do I? If you mean when I talk about 'measuring against', I was referring to the good runners who run S classes. I don't advocate using ranking lists at all for gauging performance - don't believe in them, as I think my posts show!

AWK short courses should not be 'B' by technical standard they might be by competitor standard.

Never said otherwise! I said they were 'B' classes in all but name, which they are, whatever the technical standard of the course, because of the standard of the competitors (which includes me!).

[Simple Soul]

Having established that Short and Long courses are to the same technical standard, why is the ranking for a particular Event not based on Mins / Km for each age class regardless of whether Long or Short.
The Navigation element remains the same. One arguement may be that Long course competitors may get tired so their performance deteriorates. The Short course competitors may have more controls /Km to compensate. Has any work been done on this?

[Simple Soul]

Have just scrolled back and found David Mays comments about combining Long and Shot course under the title One list for everybody. The point there was that fudge factors would have to be introduced and comparing mins/km was very dodgy. Surely there are fudge factors at work at the moment if at a Regional Event there are only two people on a course. Any ranking points there would have a fudge factor.

[CHS]

Another issue with mins/km is the different parts of an area different courses visit, eg on Graythwaite two weeks ago some of the shorter courses may have stayed mostly on the relativly flatter open beech woodland on one side of the road where as the longest courses had a significant ammount in the more physical green and steeper bit on the other side. Its impossible to quantify mins/km for diferent bits of an area.

[distracted]

which brings up the separate issue of why they do so when allocating badge standards...

[NeilC]

I'm not one who is too bothered by the ranking lists personally, but one question has just sprung to mind. As an M40 I often end up running the same course as an M35. When the ranking points are calculated for M35 are the M40 runners taken into account? I'm guessing not but statistically it would seem to be sensible. It would also mean M40 runners getting two ranking scores for an event, 1 for M35 and 1 for M40.

This thought was prompted by Saturday's race where my daughter (a W16) wanted to run the women's open course and could therefore have run as W18S, W18L, W20S, W20L, W20E, W21N, W21V, W21S, W21L or W21E. Presumably the need to award badge standards / ranking points drove the splitting of a single course results into so many classes. Madness.

[Ant W]

I agree Neil. I entered Vet Men on Saturday, had an OK run for my level of fitness, and then had mixed feelings to find I was placed 1 out of 5 on M40S. Of course I can piece together from the other results where I finished on the course (very average!), but it's a shame that the ranking points appeared to have dictated how the results appeared. I often find there's much more value in finding myself say 15th out of 40 on Blue than 4th of 10 on M40S, as the level of competition is more consistent.

I've now spotted that Course Results are available on the website, so that kind of makes a nonsense of the above! The sentiment holds, though. There are too many regional events where classes have less than 10 competitors, and course-based results (and, if possible, rankings) would be more appropriate.

I have a kind of morbid fascination for where I am ranked, but I'm not sure how much it means...

[PG]

II don't know whether Jayne has been discouraged from her original proposal to

just have one long list rather than all the different classes

As CHS and others have pointed out, minutes per kilometre aren't adequate to compare different courses, but it must be possible to find a way to do it.

The current ranking list uses the mean score of ranked runners and the mean time of ranked runners competing as the point calculation base to try to ensure that high points aren't awarded because of a relatively weak field.

Once a single ranking list was available, it should be possible (though I'm no statistician) to use these figures to provide a comparison between different courses in the same event that does reflect differences in terrain, number of controls, percentage of path running etc.

There may be some difficulty in building the list to start with, but hopefully it could be checked using sprint or middle distance races where more classes run the same course.

Incidently - please don't add district events. Many people use them to focus on sharpening techniques (or trying to), not as races.