Ranking List

[greywolf]

Presumably, though, this could happen at the moment?

It could, but the effect is much smaller, and hidden.
Take a runner with 6 ranking scores, average 1100 points, who has a crap run for whatever reason, which is only worth 1000 points. Under the current system it doesn't affect his/her ranking, which is still based on the best 6 runs, but does reduce their average from 1100 to 1086 (rounding). Which means next time they compete, the mean score of competing runners will be slightly lower than if they'd deliberately mispunched, but only by 14/n where there are n runners (so, crudely, if there's 14 runners, everyone, including your original runner, loses 1 point on average, if there are 56, everyone loses a quarter of a point, etc).
There's also an effect on the SD of competing runners scores but thats harder to quantify as it depends on the distribuion of other runners scores

If however, you base the ranking list directly on averages then the consquences of the single bad run are immediate and much more sizeable - probably a drop of over 100 places in the list - which looks like a much greater incentive to mp

[awk]

What about the average of the best X% of your runs with mps, dnfs etc. counting towards your total number of runs, and a minimum number of runs to contribute? No point in deliberately dnfing then.
e.g. best 50% to count, with a minimum of 5 to contribute, so that if less 10 races run, will have to include more than 50% races (and if less than 5, don't get ranked).

[Simon]

it uses the average of your non-zero scores, so that if you have 5 1200 point runs and a mispunch your average is 1200, not 1000 (if it doesn't do it this way then it can't assess the strength of the field properly, which is essential for determining how points are distributed)

When greywolf mentions non-zero in this averaging how are zero point scores due to taking a long time treated?

I ask because:

old scores of zero are now being given as 298.

So, if a group of people had taken a long time on a course (due to missing starts at the JK) then they should have scores of 0 for those races but would now have 298. And if these scores were previously ignored in the average calculation, for being 0, then they would now be included at 298 - deflating the average of those people.


I am also wondering how one person on a course can get points?

[AndyC]

There is an advantage of allowing people with one ranking run to be listed as it makes a newcomer/slow runner feel part of the sport as a whole and enourages them to go to more events to move up the list.
When I presented the club awards last year I added a "Lantern Rouge" prize for our lowest ranked member (he isn't any more there are two below him at the moment) -it was well received and helped make the club more inclusive IMHO

[dorien]

And I think you overestimate the issue anyway; I'd say this was akin to quick M21s running say, short green courses against the W60s just to get potentially massive scores, but it doesn't seem to happen that much.

In fact attempts to do this actually show that the system sort of works. Nick Barrable blasted round Short Green at Barossa a couple of years ago, came home nearly three times as fast as the country's best W70s, and scored... slightly fewer points than he was scoring for Browns and Blacks at the time.

Perhaps we should encourage Elites to do this in order to cross-fertilize the scores and wash out any remaining old-man bias?

[NeilC]

I tried to test this hypothesis that easy ranking points could be gained on shorter courses at the JK. I ran M50S at Cold Ash (since I then had to go and run round all the Hambledon Relay controls in the afternoon). Didn't have a clean run but won the course easily, however I ended up with a similar number of ranking points that I'd expect from a run of that quality.

[Marian]

Perhaps we should encourage Elites to do this in order to cross-fertilize the scores and wash out any remaining old-man bias?

or you could encourage old women to complete longer courses. One of my top six ranking scores is from the Black at the Boktrot - something which would not have happened in the early days of the current ranking system.

Marian, W50, doing her bit for cross fertilisation.

[Glucosamine]

In fact attempts to do this actually show that the system sort of works. Nick Barrable blasted round Short Green at Barossa a couple of years ago, came home nearly three times as fast as the country's best W70s, and scored... slightly fewer points than he was scoring for Browns and Blacks at the time. Perhaps we should encourage Elites to do this in order to cross-fertilize the scores and wash out any remaining old-man bias?

But seems rather hard on the W70s...

As to the cross-fertilisation, I suspect there are age limits whether for W70s or old men (though Charlie Chaplin managed it in his mid seventies, and medical technology is making rapid progress).

[Oldman]

Perhaps we should encourage Elites to do this in order to cross-fertilize the scores and wash out any remaining old-man bias?

Please note there is no hyphen in my name

[King Penguin]

Trouble is, when I run up to Black, on a significant number of occasions it has ended up non-quorate and no-one gets any points. If there are less than 4 EODs, or even then if just one does not finish, this could happen again tomorrow.

[greywolf]

For example w.r.t Clive Hallet ... His top score comes from a BOK level C event where he beat..... no-one.

Looking at those results, it certainly would have helped that a v good M55 appears to have jogged the first few controls. But also it seems curious that three of the top five finishers apparently aren't BOF members (they don't have clubs listed and haven't been given any ranking points) ... but in the version on the BOK site two two of the three do have a club attribution. Without knowing their previous ranking scores etc it's impossible to tell what difference it might have made if they'd been properly included in the calculations, but it wouldn't be the first time that folk have been on here moaning about anomalous scores and how crap the ranking system is when in fact the problem has been casued by errors in the uploading of results.

[m4rk]

For example w.r.t Clive Hallet ... His top score comes from a BOK level C event where he beat..... no-one.

Not quite sure what your problem is here - I agree that he got a lot of points in that event partly because he already has a lot of points. But also because the area is extremely physical (pretty much either vertical or bramble-filled; see routegadget) then the spread of times is much wider than you might ordinarily expect, so the standard deviation is higher. Clive clearly not only coped with the conditions better than everyone else, but also better than he'd have have been expected to given his prior ranking scores. Surely that's exactly the kind of circumstances that high scores should be allocated?

[awk]

Having said all that, it's what happens in the forest that matters, so believe it or not, I'm actually not that worked up about it, especially if nothing of any importance is based on it.

Just seen that start times and heats at the British Sprints will depend on rankings, so some things are starting to rely on it. Oh dear.

[Angry Haggis]

Ollie's stats have not updated this morning for the old events (before approx. mid-March), which makes me suspect he might have to do something to force a full reload.

Stats now resync'd from BOF going back to April last year, and now working back to April 2011. N.B. If the new system means all the ranking points get recalculated every week, then, in the absence of an API, it's probably not going to be practical for me to keep resync'ing on a regular basis. (Although this new process will make the results better overall by removing the downwards drift, so if my site's accuracy is a casualty of it then so be it!)

[greywolf]

Just seen that start times and heats at the British Sprints will depend on rankings, so some things are starting to rely on it. Oh dear.

If you’ve got 6 ranking scores then the list is accurate enough for this (unless you’re really bothered that your ranking of 32nd means you’ll start 1 minute earlier than if they used the list in your head).

For those with fewer than 6 counting scores, or not in the ranking list at all, it's rather more significant...