New BOF Website

[graeme]

Lets hope its OK. I didn't get the code to check , and only found the one implementation bug by doing the calculations and reverse engineering it . Life is short and there are still corners I couldn't get at.

[DJM]

Updated ranking algorithm has gone live today, together with extra facilities which give runners' full points histories (pictorially too), and direct links to the events at which points were awarded, etc. More detail here.

Any feedback on the new look and feel??

[Spookster]

Updated ranking algorithm has gone live today, together with extra facilities which give runners' full points histories (pictorially too), and direct links to the events at which points were awarded, etc. More detail here.

Any feedback on the new look and feel??

Two things I immediately like are (1) the ability to show total ranking points graphically over time, with or without other runners also on the graph, and (2) the ability to see a long list of all the ranked races I've done, or any other runner has done.

It seems strange to also display the formula that each course's points are based on. I guess that won't mean much to most people who look at the rankings.

I also see that GG now tops the rankings, so I'm guessing he'll appreciate the changes that have been made (to remove the anomalies) as well!

Thanks.

[greywolf]

The new algorithm has only been run on events since 1/1/2017 (not sure why...) so there are still some silly scores from last autumn, but otherwise seems to have produced a more sensible list, e.g. reducing the general over-ranking of MVs, etc.
Plus what Spookster said.

[graeme]

When you go to your individual events listings (a very useful addition), the numbered link at the bottom goes to some random events listing instead of your older results. AFAIK you still can search for a person by name.

The new algorithm is designed to automatically catch probable data-entry issues and eliminate them from the algorithm.

They didn't implement the algorithm correctly, and can't fix it for "technical reasons". Too technical to bother my pretty little head with the details, apparently. It will only generate serious problems when a small event includes a fast-improving runner with only one or two previous results, and someone winning by a large margin.

I assume technical reasons which they won't explain = can't be arsed

(don't know why)

The database seems to be stored by calendar year: when I got a download, it came in one year chunks, so assuming the rerun was done offline it would be fiddly to back-propagate the data. It should be OK, e.g. Ben Mitchell's silly score still has him too high in the rankings, but he has a lot of races so in calculating averages its washed out.

[Simon]

When you go to your individual events listings (a very useful addition), the numbered link at the bottom goes to some random events listing instead of your older results. AFAIK you still can search for a person by name.

If you add &page=1 to the end of the member results url you'll get the next page of individual results.

[roadrunner]

When you go to your individual events listings (a very useful addition), the numbered link at the bottom goes to some random events listing instead of your older results. AFAIK you still can search for a person by name.

If you add &page=1 to the end of the member results url you'll get the next page of individual results.

I seem to get what I expect - older results - so this appears to have been fixed.

I'm a bit puzzled by the "incentive scheme" column, though: I'm nearly always faster than 12.5min/km (and I think always faster than 20min/km), yet apparently only two of my results have counted for the "racing challenge": the Saltaire and Shipley event and the Ainsdale Dunes one (both in 2016, and both apparently at TD5??). Most of my events have been urban ones; I don't know what TD they would have been registered as but I would guess they would be TD3.

[gg]

I also see that GG now tops the rankings, so I'm guessing he'll appreciate the changes that have been made (to remove the anomalies) as well!

Yes, much better

[Sunlit Forres]

yet apparently only two of my results have counted for the "racing challenge"

I'd guess this is probably because the course naming in the results upload for those non-counting events wasn't in the format needed by the incentive scheme. The last Mike's e-news explains:
https://www.britishorienteering.org.uk/document/3970e6ee247970051a879421263bb3d5/eNews%202017-08-18.pdf

[King Penguin]

Two things I immediately like are (1) the ability to show total ranking points graphically over time, with or without other runners also on the graph

What am I missing ? I cannot see how to do this.

[King Penguin]

OK, just found the graph

[roadrunner]

yet apparently only two of my results have counted for the "racing challenge"

I'd guess this is probably because the course naming in the results upload for those non-counting events wasn't in the format needed by the incentive scheme. The last Mike's e-news explains:
https://www.britishorienteering.org.uk/document/3970e6ee247970051a879421263bb3d5/eNews%202017-08-18.pdf

Yes, that seems to be the explanation, though it makes a mockery of the system: the "blue" course at the Saltaire and Shipley event could hardly be TD5 as it was a standard urban course! The incentive scheme doesn't bother me, but if it's going to be fair to those who run urban events then I think it needs some revision.

There's another, different, anomaly that I can see as well: among my events are "NWO Urban Event" on 25/9/16 (which was at Royal Wootton Basset) and "NWO Four Colour Event" on 22/10/16 (which presumably was Hens Wood). I'm shown as running the same course, in the same time, at both (but with ranking points only for the first), whereas in fact I never ran the second event! I presume this is some sort of data upload issue (and perhaps it was there before the update; with the old system, I wouldn't have seen the second event).

[Sunlit Forres]

another, different, anomaly

Looks like NWO (I'd never heard of them before) uploaded the wrong results file for the Hens Wood event.

I don't think the practicalities of results submission was properly thought through when the Incentive Scheme was established in 2012. I contacted BO a while ago to see if there could be a box to select the TD on each course when doing results uploads rather than relying on results secs renaming courses to a colour. It would avoid this silly situation when a big proportion of the results submitted contain invalid names for the Incentive Scheme. I guess it won't be quite as easy as that.

[roadrunner]

I don't think the practicalities of results submission was properly thought through when the Incentive Scheme was established in 2012. I contacted BO a while ago to see if there could be a box to select the TD on each course when doing results uploads rather than relying on results secs renaming courses to a colour. It would avoid this silly situation when a big proportion of the results submitted contain invalid names for the Incentive Scheme. I guess it won't be quite as easy as that.

It's certainly a bit crazy, because the most prestigious events - British Champs, JK, etc - don't use the colours and so don't count! And, as has been said before, some areas of the country don't support true TD5 courses, so some Green/Blue/Brown etc courses are really not the TD applicable to their colour. Surely it would be simple enough to map the usual age-class courses at both forest and urban events to the intended TD level, e.g. for forest events M21L = TD5, M10A = TD2, and for urban events MV = TD3 and so on.

[canol]

Thought the ranking list is supposed to identify the best orienteer, and I still get something out of it as a plodder comparing myself to other plodders. (My definition being on a good day might get 1000 points). But now I find my scores have improved often by over 100 points per event and now I am bobbing along just over a 1000 points per event. Going to have to think of a new term/definition (? bobber)