Should night events be ranking events?
Moderators: [nope] cartel, team nopesport
70 posts
• Page 3 of 5 • 1, 2, 3, 4, 5
Re: Should night events be ranking events?
I think I might have voted for no by mistake in this poll, but I can't change it, boo.
Andrew Dalgleish (INT)
Views expressed on Nopesport are my own.
Views expressed on Nopesport are my own.
- andy
- god
- Posts: 2455
- Joined: Thu Nov 06, 2003 11:42 pm
- Location: Edinburgh
Re: Should night events be ranking events?
Good quality night events require more orienteering skills than the average day event. It's not so easy to run hard and relocate, you need to keep navigating.
This will of course lead to "anomalous" results, the better navigators tend to stand out more. Personally, I would have thought this was a good "anomaly" to have "distorting" the rankings.
Anyway, there are only a handful of night events of the requisite grade each year. So nobody is going to get a good ranking based purely on their nocturnal performances.
This will of course lead to "anomalous" results, the better navigators tend to stand out more. Personally, I would have thought this was a good "anomaly" to have "distorting" the rankings.
Anyway, there are only a handful of night events of the requisite grade each year. So nobody is going to get a good ranking based purely on their nocturnal performances.
- IanD
- diehard
- Posts: 632
- Joined: Sat Sep 03, 2005 7:36 am
- Location: Dorking
Re: Should night events be ranking events?
andy wrote:I think I might have voted for no by mistake in this poll, but I can't change it, boo.
You better get some polling training in before 2014 then Andy!
Orienteering - its no walk in the park
- andypat
- god
- Posts: 2856
- Joined: Sat Apr 11, 2009 9:58 pm
- Location: Houston, we have a problem.
Re: Should night events be ranking events?
Scott wrote:
I have to admit that I'm not clear in what way the results from night events are supposed to be "anomalous". If it is simply that some orienteers who are very good in the daytime are rubbish at night and so get beaten by people whom they would normally beat in the day, I don't really see why that's a problem. But if they are anomalous in some way that means that the statistics don't work properly, then it does seem reasonable to exclude them.
"if they are anomalous in some way...." Humph! Of course they will produce some anomalous stats. Question is surely whether they would be more so than all the anomalous stuff that's in there already.
- The Loofa
- light green
- Posts: 243
- Joined: Wed Apr 27, 2005 11:28 pm
Re: Should night events be ranking events?
To give the Swedish view...
Swedish ranking INCUDES night events and EXCLUDES Sprint (and Urban - not that they have any!). They have done their homework and have much more data to work with. Their conclusion is that Sprint events are the ones causing the anomalies. They still rank them however and you can see what you get but are not included in your scoring 6.
Of course there has been debate about the exclusion of Sprint - especially when they seed/decide Sprint Elitseriens/Swedish Champs based on ranking that doesnt include any Sprints!
Also note that it costs 200sek (£20 - £30 if you pay late) a year to have a Swedish ranking - on top of your club fee. (No SOFT fee as it is the club who pays/affiliates to SOFT) (However, you can always look at your Swedish ranking online through a mates login and see where you are. Note you cant run Swedish Champs without paying your ranking fee - but Swedish Champs is only for M/W 18/20/21 classes.)
Other thoughts....
World Cup of the future - sounds a bit like Park World Tour. 20 years ago Jorgen M and guys wanted the IOF to take over and run with the PWT idea/format - bringing O to the people, showcasing O around the world in Sprint/short race form. It has taken them 20+ years but maybe the vision is finally being realised! PWT will be happy!
If the IOF put equal weighting on Sprint Middle Long, then the World Cup programme should have the same weighting, plus a few extras like chase/mass start looped, etc. Plus, bring back the relays. Was a mistake to ditch them. WOC has one, why not a series in WCup rounds...
I think a WOC Night would be interesting and one which would have to have a standard lamp for all to make things fair. The new Sprint Relay could be held at Night.
Maybe DJM is bad at Night events ) I think one example which David will recite is that of Dave Godfree who was UK no.1 for quite a time based on his Scottish Night Champs results. For a year or 2 or 3?, the score he got from that event was so huge - winning, smallish field and some being out for ages - it meant he could stay top ahead of others who were better than him in most other races. However, I think the system is better now. Time for change.
Night events do not change the performance of the top people who still run as hard and navigate as well. However, the slower less stable types!? will have their errors compounded at night so the spread is much bigger. (Maybe Brits are more tenacious and struggle on to massive times?! Maybe Swedes are just better navigators!?)
I think in urban races, some will run the same speed as their forest speed, as they cant run any faster (lack of running training), whilst the 'runners' will see their pace increase. So if you can run fast, your performance is relatively better on tarmac than in the forest where the veg, harder map reading, hills will slow you down. It is no surprise to me that scores are better for Urban races.
Swedish ranking INCUDES night events and EXCLUDES Sprint (and Urban - not that they have any!). They have done their homework and have much more data to work with. Their conclusion is that Sprint events are the ones causing the anomalies. They still rank them however and you can see what you get but are not included in your scoring 6.
Of course there has been debate about the exclusion of Sprint - especially when they seed/decide Sprint Elitseriens/Swedish Champs based on ranking that doesnt include any Sprints!
Also note that it costs 200sek (£20 - £30 if you pay late) a year to have a Swedish ranking - on top of your club fee. (No SOFT fee as it is the club who pays/affiliates to SOFT) (However, you can always look at your Swedish ranking online through a mates login and see where you are. Note you cant run Swedish Champs without paying your ranking fee - but Swedish Champs is only for M/W 18/20/21 classes.)
Other thoughts....
World Cup of the future - sounds a bit like Park World Tour. 20 years ago Jorgen M and guys wanted the IOF to take over and run with the PWT idea/format - bringing O to the people, showcasing O around the world in Sprint/short race form. It has taken them 20+ years but maybe the vision is finally being realised! PWT will be happy!
If the IOF put equal weighting on Sprint Middle Long, then the World Cup programme should have the same weighting, plus a few extras like chase/mass start looped, etc. Plus, bring back the relays. Was a mistake to ditch them. WOC has one, why not a series in WCup rounds...
I think a WOC Night would be interesting and one which would have to have a standard lamp for all to make things fair. The new Sprint Relay could be held at Night.
Maybe DJM is bad at Night events ) I think one example which David will recite is that of Dave Godfree who was UK no.1 for quite a time based on his Scottish Night Champs results. For a year or 2 or 3?, the score he got from that event was so huge - winning, smallish field and some being out for ages - it meant he could stay top ahead of others who were better than him in most other races. However, I think the system is better now. Time for change.
Night events do not change the performance of the top people who still run as hard and navigate as well. However, the slower less stable types!? will have their errors compounded at night so the spread is much bigger. (Maybe Brits are more tenacious and struggle on to massive times?! Maybe Swedes are just better navigators!?)
I think in urban races, some will run the same speed as their forest speed, as they cant run any faster (lack of running training), whilst the 'runners' will see their pace increase. So if you can run fast, your performance is relatively better on tarmac than in the forest where the veg, harder map reading, hills will slow you down. It is no surprise to me that scores are better for Urban races.
-
Ravinous - light green
- Posts: 278
- Joined: Fri May 13, 2005 9:48 pm
- Location: Just by Monty's Bunkers
Re: Should night events be ranking events?
Ravinous wrote:It is no surprise to me that scores are better for Urban races.
It should be, because it's impossible If an event has twenty finishers (who are all 16+, BOF members & have at least one previous ranking score etc) then the total number of points scored by the 20 finishers is exactly the same regardless of whether they've run sprint / urban / middle / long (and whether they've run at in daylight or at night).
What might change between these event types is the distribution of those points between the 20 competitors. if different event types have a different characteristic distribution of finishing times, then you'd expect the winner of some event types to get more points, all other things being equal, than the winners of other event types. But it would be balanced up by the losers of the first event type getting fewer points.
In practice this effect happens between courses in "normal" forest racing too - if you look at the graph that Spookster posted on another thread all the talk has been about how many extra points Colin Dickson got but the tailenders on Blue clearly got even less than they might have expected.
Ravinous wrote:Dave Godfree who was UK no.1 for quite a time based on his Scottish Night Champs results. the score he got from that event was so huge - winning, smallish field and some being out for ages -
Won't speak for DJM but that was the "anomaly" identified - there were a few examples where in very small fields the winners scored huge points because a couple of competitors had got massively lost. As has been said ad nauseam, the 10 ranking competitors rule would reduce the impact of that now
-
greywolf - addict
- Posts: 1416
- Joined: Sat Apr 07, 2007 12:45 pm
- Location: far far away
Re: Should night events be ranking events?
There are examples where in very large fields winners scored huge points because everyone else was very average. I think Matt Crane had a massive score from a Harvester middle a few years ago. And also a top junior, can't recall who, also had a massive score from a CSC round blue course. These were both in forest based day time orienteering events.
-
mharky - team nopesport
- Posts: 4541
- Joined: Fri Oct 24, 2003 3:39 pm
Re: Should night events be ranking events?
Not sure what a "Harvester middle" is and the info about the other example is just a tad vague ...
But, if you go back into rankings correspondence of a little while ago, you will see that it was riddled with examples of anomalies, many of which were caused by incorrect data entry and others by errors in the rankings algorithm which were subsequently corrected.
I'm not sure we can deduce anything from mharky's post.
But, if you go back into rankings correspondence of a little while ago, you will see that it was riddled with examples of anomalies, many of which were caused by incorrect data entry and others by errors in the rankings algorithm which were subsequently corrected.
I'm not sure we can deduce anything from mharky's post.
- DJM
- diehard
- Posts: 978
- Joined: Sun Sep 30, 2007 8:19 pm
- Location: Wye Valley
Re: Should night events be ranking events?
Wrong stuff deleted, what Dave says below is right. They had some weird idea that a spread based on other events elsewhere would be better than, e.g. using other courses at the same event.
I'm going to run a student project on the ranking list next year. The student has no background in orienteering, so can treat the data without prejudice. We'll look at whether night or urban results really are anomalous. I think it is possible to figure out for each event how anomalous the results are, which data could be used for weighting. I had thought that would make it too difficult to understand the system, but it seems that actually hardly anyone understands it anyway.
I'm going to run a student project on the ranking list next year. The student has no background in orienteering, so can treat the data without prejudice. We'll look at whether night or urban results really are anomalous. I think it is possible to figure out for each event how anomalous the results are, which data could be used for weighting. I had thought that would make it too difficult to understand the system, but it seems that actually hardly anyone understands it anyway.
Last edited by graeme on Sun Oct 28, 2012 6:55 pm, edited 1 time in total.
WOC2024 Edinburgh
Test races at SprintScotland (Alloa/Falkirk) and Euromeeting (near Stirling).
Test races at SprintScotland (Alloa/Falkirk) and Euromeeting (near Stirling).
-
graeme - god
- Posts: 4723
- Joined: Thu Feb 12, 2004 6:04 pm
- Location: struggling with an pɹɐɔ ʇıɯǝ
Re: Should night events be ranking events?
The anomaly was actually due to the previous incarnation of the ranking list disregarding the statistics for small fields (which I think was less than 20) and instead having the standard deviation as a fixed % of the mean (20%?).
This made it easy to score high points in events in technical or physical terrain where the spread tends to be larger. As pretty much every event in Scotland had less than 20 ranked runners then anyone like me who went to 6 in a year got a good ranking score.
There was nothing specifically anomalous about night events - they just normally had less than 20 runners and the added difficulty at night tends to give a higher spread.
The current ranking system no longer has this anomaly as it uses the actual statistics for the standard deviation. Instead it doesn't give points at all for less than 10 ranked runners. (Which has resulted in Black courses in Scotland not getting ranking points.) With this change there is no longer any reason to exclude night events.
This made it easy to score high points in events in technical or physical terrain where the spread tends to be larger. As pretty much every event in Scotland had less than 20 ranked runners then anyone like me who went to 6 in a year got a good ranking score.
There was nothing specifically anomalous about night events - they just normally had less than 20 runners and the added difficulty at night tends to give a higher spread.
The current ranking system no longer has this anomaly as it uses the actual statistics for the standard deviation. Instead it doesn't give points at all for less than 10 ranked runners. (Which has resulted in Black courses in Scotland not getting ranking points.) With this change there is no longer any reason to exclude night events.
-
Godders - blue
- Posts: 416
- Joined: Tue Nov 04, 2003 4:37 pm
- Location: Swanston
Re: Should night events be ranking events?
DJM wrote:Not sure what a "Harvester middle" is and the info about the other example is just a tad vague ...
But, if you go back into rankings correspondence of a little while ago, you will see that it was riddled with examples of anomalies, many of which were caused by incorrect data entry and others by errors in the rankings algorithm which were subsequently corrected.
I'm not sure we can deduce anything from mharky's post.
16/07/11, Harvester Weekend - Middle Distance Event, Bowden Housteads, SYO
1 - Matt Crane, 1394 pts
http://www.britishorienteering.org.uk/index.php?pg=results&eday=60029&results=60029&course=10&
17/10/10, CompassSport Cup Final, Burbage Moor, SYO
1 - Matt Halliday, 1379 pts
http://www.britishorienteering.org.uk/index.php?pg=results&eday=53912&results=53912&course=9&
The actual CSC round I was referring to seems to have been corrected in some way, it was won by Peter Bray with Jonathan Crickmore very close behind. They both had scores similar to above, and I can recall them being counting scores for a significant period of time, if not the whole 12 months they were eligible.
I highlighted the above examples in response to the statement made by Greywolf about very small fields creating anomalies. I was trying to show that in very large fields winners can also score huge points.
greywolf wrote:Won't speak for DJM but that was the "anomaly" identified - there were a few examples where in very small fields the winners scored huge points because a couple of competitors had got massively lost. As has been said ad nauseam, the 10 ranking competitors rule would reduce the impact of that now
That is what you can deduce from my post.
-
mharky - team nopesport
- Posts: 4541
- Joined: Fri Oct 24, 2003 3:39 pm
Re: Should night events be ranking events?
graeme wrote:I'm going to run a student project on the ranking list next year. ... We'll look at whether night or urban results really are anomalous.
I think that just as great an issue is how different courses at the same event score. Today's sprint at Holt Park finally put the nail in the coffin for me: in terms of mins/kms, and in terms of the split times of common legs, I (on Course A) was running to within 15 seconds or so of the two leading M55s on Course B. Yet I score 1075, and they score 1148 and 1146.
This is not the first time by any means where it appears easier to score points on shorter courses. It may not appear much, but in terms of rankings it makes a pretty big difference.
TBH, I'm not that bothered (believe it or not!): I was initially interested, but it has become increasingly clear to me that the rankings are so full of holes as not to be worthy of taking serious note. My only concern is the odd occasion when somebody proposes basing something important on them.
-
awk - god
- Posts: 3224
- Joined: Sun Mar 21, 2004 5:29 pm
- Location: Bradford
Re: Should night events be ranking events?
awk wrote:This is not the first time by any means where it appears easier to score points on shorter courses. It may not appear much, but in terms of rankings it makes a pretty big difference.
Nor the last it seems. Today, I ran 39:02 for 6.2k on the B course at the SROC Urban Race 'earning' 1047 points. The same time on the C course (i.e. if I'd run 39:02 for 5.6k) would have 'earned' me almost exactly the same, whilst someone who ran the same mins/km would have scored 80 points or so more. Indeed, Karen Nash (C winner, 1135pts) and I found ourselves running together for a decent chunk of the course and reckon that we were running very much the same pace (as our results confirmed).
These sorts of disparities happen week after week. IMO, the rankings list are simply not fit for purpose.
-
awk - god
- Posts: 3224
- Joined: Sun Mar 21, 2004 5:29 pm
- Location: Bradford
Re: Should night events be ranking events?
Thats odd because in Scotland you often find that the winner of the shorter courses gets pretty short shrift when it comes to points and if you were a ranking points tart, you'd often be better finishing last on the long course. I guess it all depends on the points people were allocated when the rankings came out and how/if they are evening out.
I guess part of the problem is that most of the events where we all run the same courses are level D and (rightly) dont attract ranking points.
I guess part of the problem is that most of the events where we all run the same courses are level D and (rightly) dont attract ranking points.
Orienteering - its no walk in the park
- andypat
- god
- Posts: 2856
- Joined: Sat Apr 11, 2009 9:58 pm
- Location: Houston, we have a problem.
Re: Should night events be ranking events?
awk wrote:I think that just as great an issue is how different courses at the same event score.
That is an issue, but the cause of and remedy for this when viewed as a statistics problem are so well understood I don't imagine it will detain a smart student for long. Not many marks for pointing out the bleedin' obvious I'm afraid.
We might look at a totally different system based on who-beat-who.
WOC2024 Edinburgh
Test races at SprintScotland (Alloa/Falkirk) and Euromeeting (near Stirling).
Test races at SprintScotland (Alloa/Falkirk) and Euromeeting (near Stirling).
-
graeme - god
- Posts: 4723
- Joined: Thu Feb 12, 2004 6:04 pm
- Location: struggling with an pɹɐɔ ʇıɯǝ
70 posts
• Page 3 of 5 • 1, 2, 3, 4, 5
Who is online
Users browsing this forum: No registered users and 109 guests