Results and Rankings.

[Big Jon]

Just had a look at my ranking - I scored more points in coming 5th in the JK sprint than in winning Day 2 at the JK (both in M45) and some people wonder why the new ranking system is held in ridicule by many orienteers....

[DJM]

Jon,

Hold on a bit ...

It's been made clear several times recently that errors and teething problems have been spotted in the system and that these will be corrected soon. It's been a massively complex task to effect the new scheme too. It really isn't fair to be so critical under such circumstances.

We have been entering uncharted territory in going from a multiple scale system to a single scale one and we now know a good deal more about this new terrain after the first few months. I'm confident we'll do better when the corrections have been made!

But each change to the system has to be tested thoroughly, and these things take time ...

[madmike]

Jon,
unless I fundamentally misunderstand the new system(and I confess that's a possibility), your ranking points are more dependent on realitive performance against all classes rather than just your position against us other M45s - I think ridicule is a bit strong - give it time.

Personally I was exceptionally rubbish at the sprints and got a very low score, average for me on day 2 and got an average score, and pretty good (exempt control 2) on day 3 and got my highest score so far. The difference between exceptionally rubbish and good was 700 pts to 1100 pts with average at around 1000 - so ranking points seems about right for me.

I'd be interested to know whether your post race analysis tells you that you'd done exceptionally well when you won and OK when you were 5th?

[Big Jon]

David MAy wrote:
"It's been made clear several times recently that errors and teething problems have been spotted in the system and that these will be corrected soon."

How long is long enough to clear up "teething problems"? - the system has had 4 months and as far as I can tell (from my own results/ranking) and other's comments it is pretty close to jobby!

[Big Jon]

"jobby" in the above posting was not my phrase - I used the word "jobby" as that is what I meant.

[Big Jon]

obviously an automatic word change - I used (and meant) the word "sh1t"

[graeme]

If you're interested in fixing the system, why be so hostile to Jon reporting problems?

Jon's problem arises because the M45S runners are overrated relative to the M45L runners. So he gets loads of points for beating them in the JK sprint, then less in the 45L when they're missing. It's consistent with other statistical problems people have reported. It comes from treating the seed data assuming a "normal" distribution, and it isn't going away in a hurry.

And you know this because I've been saying it for years.

The problem now is that the whole list needs to be restarted, which will be temporarily embarrassing. But thats better than being permanently embarrassing.

Isn't it?

[DJM]

If you're interested in fixing the system, why be so hostile to Jon reporting problems?

Who is "you" and where is the hostility?? I see none at all ...

[NFKleanne]

I was amazed to find myself in the ranking list! And was quite chuffed not be last in my club, and 200 odd from the bottom in the whole table.

Haven't got a clue how its worked out, and don't really care - its just encouraging that I'm higher up the table than I thought I was.

Keep up the good work

[madmike]

Jon's problem arises because the M45S runners are overrated relative to the M45L runners. So he gets loads of points for beating them in the JK sprint, then less in the 45L when they're missing.

I'm now totally confused, is it just us overrated M45S specialists Big Jon's scoring against or is is all overrated short(or shorter) course runners in the new system

If the latter then surely it shouldn't matter whether the likes of me are included or not. Or is the whole thing my fault for being so utterly rubbish at the sprint and inflating the points of all the M45Ls

If the former then how the hell does the single list work?

PS Jon was genuinely interesed in your feel on the realitive quality of your runs.

[greywolf]

I'm now totally confused, is it just us overrated M45S specialists Big Jon's scoring against or is is all overrated short(or shorter) course runners in the new system

Your points are calculated against everyone (18+) who is running the same course - so whereas the M45L class at the JK had its own course on Days 2 & 3, at the JK sprint M45L ran together with M45S (and M40L & M40S)

The seed data used to initiate the rankings list (almost certainly) didn't differentiate classes sufficiently, so that if you share your course with a slower class you might expect to score some extra points, if you run against a faster class you might expect to score less.

At the JK this effect was quite marked in M45 & M40 - Dave Rollins 1355pts & Martin Ward 1345 won the sprint races, whilst the leading scores on Days 2 & 3 were 1306, 1300, 1304 & 1298...( so this suggests an effect worth up to 50 points - ~4% )

BUT, the effect was nowhere near so marked with most other courses: e.g. on Course 7 at the Sprint the winners were Mike Murray (M65) 1201 pts, Sara Campbell (W45), 1173 pts, the following day at Cookworthy they both won their respective races and scored 1211 & 1161 pts respectively... likewise W55 winners scored 1130, 1139 & 1145 points (and no ranking system is ever going to be perfect in a sport like orienteering...IMHO anything within 1% might as well be regarded as noise)

As for fixing it... there's a number of stages: identifying the problem, suggesting solutions, checking them, then implementing the solutions...and they aren't all done by the same people...the first couple have been done, but it is down to BOF to accept and implement solutions, my impression is that they are loathe to commit significant effort to rebooting the rankings system (especially ahead of the competitions review)...but if anyone really has a problem with the (acknowledged) shortcomings of the list then i'd suggest lobbying BOF to get it fixed. Of course if your assessment of a ranking system is based on how closely the output corresponds to the pre-existing ranking list in your head then the whole thing's a fool's errand...

and incidentally, IIRC the rankings list was launched mid - January, i.e. ~ 3 months ago rather than 4...against that standard of accuracy it's not doing so bad...

[madmike]

Thanks Wolf that's much clearer and it probably was my fault

[Spookster]

As for fixing it... there's a number of stages: identifying the problem, suggesting solutions, checking them, then implementing the solutions...and they aren't all done by the same people...the first couple have been done, but it is down to BOF to accept and implement solutions, my impression is that they are loathe to commit significant effort to rebooting the rankings system (especially ahead of the competitions review)...

Good summary, thanks Greywolf. We want the emerging problems fixed soon, and I think they will be.

[DaveK]

I do wonder if there should not be an option to 'opt out' of the rankings (as you can with the electoral register) so as to avoid embarrassment with one's peers?
Anyway, on the same general theme, to beat those post-election blues do come to the Harvester weekend.

[Scott]

Well, it seems that a couple of changes have now been made. I've not had time to look at the effects in detail yet, although one of the most glaring anomalies on the old list appears to have sorted itself out under the amended scheme - Cat Taylor's 1423 points for winning W20S at Greythwaite last year has become a rather more reasonable 1195.