Discussion: WOC Long

Running for Team USA tomorrow:

Alex 12:38
Sam 12:54
Eric 14:39

Times are in Sweden (-6 for time in Boston). Follow along at http://live.woc2016.se

Thanks for the reminders Cristina!

Canada

Louise Oram 12:40
Emily Kemp 13:46

Damian Konotopetz 14:23

2 minute start interval. Big train potential.

The men's courses are "forked" in some way but the women's are not. It'll be interesting to see how different the route choice possibilities are, and if that's enough to slow down the train formations.

There's been such a buildup for this on the event website that I can hardly wait. Wonder if I can get a live ticket at the last moment. Maybe I'll just try a scalper and swallow the cost..

how about that 4km leg to #2.

There's a flattish mainly path/track route just N of the red line, looks like Tash is going thataway

... or maybe not - the red-line route ain't bad.

It looks like Alex is going the road (at least to start)... just saw Niamh O'Boyle's spot appearing from the bottom of the map on that road in the SE.

Overshoot #4 and you are in the wrong country :)

Just saw Sam start! And a bunch of loud Americans were yelling. :)

hey we (for now) are getting what we want - Per Forsberg audio! ;-)

5-6 looks a bit ominous with all those cliffs. I would think either pushing the gap between the cliffs just right of the line (after going south of the lake) is the best route. Thoughts?

Team USA blogpost before the long .

Currently hiding out from the rain in the Däftö cabins ruing a terrible internet connection. A lot of refreshing is necessary to keep up with the live stream.

Ditto.

5-6 - I'd go along the track for as long as I could then NW using open marshes and hill sides. MAybe O'Boyle will do that, I guess that the top guys will be more aggressive,

OTOH the first post on the (barely) live blog was 24min after the race started. Second posting another 24min later. I can't find a pulse.

Well maybe all those people standing around at #4 are border guards.

Sam caught Imogene pretty directly... keep it up Sam!

did anybody else just lose video and replaced with countdown?

yes

yes

What is the best way to 2?

Arntraut Goetsch is now in Norway - will she be able to adapt to the different mapping style ?

I like the way Louise took - hitting the olive green house 1/3 of the way south of the line, then across the rough criss-crossed by canals, and attacking from the open south of 2. I think she went to the road a bit too early and could have gone more due east from the criss-cross canal.

Running the road like O'Boyle and Ilieva way to the south seems not that bad so far, but surely it can't be the best route.

Live stream back.

The shot from the first TV control across the rough open to 4 looks amazing.

I think the long road around is not far off the best. There is just nothing great more straight. Perhaps first northeast, jump on the trail past 9, then over to the trail just north of the line... but then the approach from there is ugly.

O'Boyle's avg speed is way faster than anyone else (like 8km/h v 5-6) even now she has spent some time in the bush. It may be the best - Tove would fly on that route and you could plan the whole course.

Best way is probably however Tove does it.

Looks like Ilieva made up a bit of time on Sam by taking the road south.

Goetsch has come back from Norway and it looks like that was the faster route

That's very clever - going through Norway. But doesn't the higher cost of living punish you?

Looks like she saved 4 minutes

Feels like the route through Norway is the best to 6, for sure.

Feels like the route through Norway is the best to 6, for sure.

Great Tash has gone that way too, maybe she can make up some lost time.

Kolosinska taking a very interesting left route to 6; I'm surprised how many people are running left of the line. Nevermind, looks like a mistake running a bit too low for Kolosinska.

This course is actually pretty meh, but my word I wish I was out running in those conditions! I miss soggy Europe!

On the Swedish east coast it's full sun and hot!

Yeah, done with that. Give me fog and rain!

Lots of loud (presumably) Brits cheering for Hollie. Always nice to see some enthusiasm.

Why are so many women overshooting #4 and going into the wrong country?

Lots of polite Swedish applause for Kempster.

Kempster taking a very direct line to 2 so far.

OTOH the first post on the (barely) live blog was 24min after the race started. Second posting another 24min later. I can't find a pulse.

Meanwhile I left this thread for a little over an hour and look what happened.

The Norwegian route just got popular

Gyurko taking the lead at first TV after running the road around to 2.

ah now the commentary starts

Kempster taking the marsh to 2.

Orr had trouble with the uncrossable watercourse half way to #2. Others have crossed it easily though.

Well, the TV coverage was pretty good until they switched over to the English language commentary. Now I have no idea what's going on in the race.

Seriously. Put Per back on.

"So the point there is that the shortest way isn't always the best way?" - gem from the English commentary

Oi - he's definitely Scottish

This second leg has been great at forming packs

Seems like the setters have jumped the gun on the WOC Women's long distance winning time survey.

The English commentary seems to be aimed at the general public, yet those who are watching the internet live stream are almost certainly orienteering enthusiasts.

Yeah, is this the package that goes at to television stations at a later time? I would prefer speaker arena for those of us that paid for WebTV access.

Is that fair?
First 5 controls are the same for M & W!

Why would it not be fair? The men are in an embargo, so it's not like they can get any advance intel.

Some women can have men as friend for a leg...or part of it.
Some others can have just women, not same speed I think...

This second leg has been great at forming packs
No kidding. I just did a quick count and seems like way more than half the runners are in trains after leg 2. It is more unusual to be on your own after #2

#1 is different

General public dont buy online tickets that is for sure. They watch TV or watch football.

What is currently shown on people's live feed, has Emily got to Split 1 yet? #5?

Tove really gaining on Hausken-Nordberg.

Same hole, Mark :)

Just got there, 3rd place

OK, coz I can do a live audio broadcast, but I'd only be able to commentate on what is currently on my live feed, so if the timing is off, it would sound a little weird

3km of road running on the south route would be nasty in o-shoes.

Could you do a live broadcast making fun of the current live broadcast?

Nice pack at #10.

Emily is taking the Norway route - if she gets it right at the end of the track her positions should hold up for a while

plus a bit of commentating on the race, and real pronunciation of names... Eye-sia Basset ...

finky!

Emily dropped a little low and north at the start of the route to #6 though but moving well now.

Good on Louise Oram. A very solid effort.

Anastasias!

The track to 6 looks like it's maybe up to 1 minute slower

Gemperle crushing it on the road run. Wyder not so far behind going straight.

Emma Johansson appears to be running with a couple of the early men's starters.

Tove is crushing it.

Natalia caught up with Judith.

I would tune in to a Nixon live feed for the relay! Unfortunately now I have to go to the dentist, which may be less painful than his pronunciation of most people's surnames.

Nee-av

Ugh.

The Gemperle-Wyder train might be a dangerous one.

She did apologise for mangling Kolosinska, but just because she stuttered.

Gemperle now with Wyder and could set the time for Alexandersson to beat

You'd have thought a Scottish commentator could at least get the Irish names right.

Similarly, Wyder now has a bloke running with her as well as Gemperle.

Is Gemperle firmly on Wyder's tail? (6 sec back at #5) or are they sharing duties?

How did they mess this up?! Starting men simultaneously with women, with the same controls?

Gemperle is in front at the moment. They've dropped Mulder.

Nice to see Nick get some TV time!

oh. my. god.

trafficlighting.... FFS

Eric Bone choosing the road to 2. Not a lot of other guys doing that.

Emma Johansson should have stuck with Yasur - she's lost time trying go more direct.
I wouldn't be surprised if we get a Tove-Johansson-Hausken Nordberg-Fasting train to rival the Wyder-Gemperle one...

Alexandersson over 2 min behind Gemperle at 2.

Congratulations to Judith Wyder on her silver medal

Tove lost time at 2, Nordberg and Johansson losing time to 3

My phone is streaming consistantly 30s behiund my computer, anyone know enough to work out why and how to sort it out?

Hausken Nordberg now running with Gene Beveridge.
This is a bit of a farce - basically one-minute start intervals for the WOC Long Final.

Well, 2 minute

The only option I see, if you want to keep the good long legs, is for the women to do a butterfly at 8, like the men, and have no final loop

How could they not have predicted this???

Control 15 is has caused a lot of damage. Post-spectator-loop syndrome.

And that's not the worst of them.

So, it turns out that the road route was best to #2 even though Tove didn't take it.

Is Germperle wants to win she's going to have to stop following Wyder, as she's missing #8 too

33 minutes of running, and 10th place is already 7 minutes down

The production is awful. Why on earth are we seeing Cat, when unfortunately she is out of the running, when we could be looking at GPS replays to show how the medal race is unfolding?

It's one-minute intervals if you take the men and women together, which seems fair given the first 5km are identical.

If you dropped the spectator loop for the women in favour of a mid-course butterfly, and sent the men around the spectator loop at the start of their course rather than the end, you could preserve the long legs and interleave the start times without them interfering with each other. Might require a little rearranging of the arena.

Undy, I was taught that the long easy route should not be the fastest one. In a Swedish book on course setting.

Again, #2 is a different control. But yes, basically the same.

Have women be able to keep up with men? It's possible, the early men and late women would very well be close on fitness. I certainly struggle to drop Cat in terrain.

@gruver Maybe they never translated that book into Russian ?

There's been bit of a GBR bias to the coverage and (particularly) commentary throughout the week. Perhaps unfortunate given the more international nature of the online audience. Anybody know whether this is the same coverage that's going out on Eurosport Asia-Pacific (or whatever it is that's making WOC available in Vanuatu)?

I think the planner thought straight would be great. Gemperle was 3 minutes faster than anyone going round...

She's only doing 4min/km's on the road though, so not ridiculous.

And yes, that British bias has been quite irritating

What was the EWT for this long?

Gemperle and Wyder losing time at #9

It's really sad that the medals will be decided by which pack goes fastest, not which runner. Very disappointing course planning.

Apologies - I hadn't noticed that the men had #2 slightly further down the hill. I don't think it makes much difference until the final approach to the control.

There definitely have been groups of men and women running together for little periods of time. I imagine the terrain running speed of some of the earlier men's starters is probably similar to that of the top women.

Agreed

Also, hi Scott! Long time, no see. :)

What was the EWT for this long? 80 minutes

So, what commentary haven't wored out, the Gemperle pack has lost 20s to Kinni from 8-11

Very disappointing course planning.
Another reason to have WOC in Canada ;-)

Tove may catch them yet.

Kempster!!

Alexandersson looks like she may close the gap to Gemperle at #9. Probably a 2 horse race now.

Another top 10 for Kemp. But is a podium still possible?

It looks really tough for Emily to make top 6 today. Maybe if some runners make mistakes towards the end.

Yeah, Tove's not out of it yet, and she might yet get Hausken Nordberg to run with round the final loop.
Hi Boris!

The commentary seem determined not to address the issue of bunching. They are claiming only that "it can be distracting" and "you have to run your own race" type of things. But it seems to me this bunching is a massive story and can't be ignored...

top two have been together for almost an hour now...

The IOF don't care, they wanted it like this

Looks like Gemperle/Wyder might have lost a little time in the rocks and bushes approaching #15.

This is gonna be really close!

Kempster will have 2nd best ever NorAm WOC placing (2nd only to her awesome performance two days ago). Need a DQ or big mistake to get her on the podium again.

TOVE!

I think Tove will be leading by the run-through

Tove's running a bit more slowly than Hausken Nordberg.
Edit: but faster than Gemperle?

Looks like 8th for Emily. Very impressive once again!

So, did they just not show any footage of Tove coming through with seconds to the lead?!?!?!?!

So my prediction of a Tove-Johansson-Hausken Nordberg-Fasting train was only three-quarters correct - it's Hauswirth instead of Johansson.

They didn't show either Gemperle or Tove through the arena. WTF?

Gemperleeeeeee FFS

Good predicition, lost. Too bad if Wyder misses out on a medal after carrying Gemperle all that way.

No way Wyder was carrying the whole way. Are you kidding me?

Well, she was making mistakes...

Launstein dragged Kkramov to a gold in Japan

I think Tove will edge this though

Again, both Marc and Andrey said they split up and came together and were leading at different points (though Marc was in front more.)

Why on earth is the TV showing us #11?
(Also: "They're running out of map"?!?)

Boris. I didn't say the whole way.

TOVE!!!

What a beast

No one seems to give the Russians any respect around here.

Top 5 all-time NorAm WOC placings

4th Emily Kemp CAN (middle 2016)
8th Emily Kemp CAN (long 2016)
9th Sandy Hott CAN (middle 2005)
10th Ted de St Croix CAN (classic 1985)
11th Emily Kemp CAN (middle 2014)

Congrats Kempster!

This won't get any press, but..

19th - Anna Serrallonga Arques, ESP
20th - Susen Loesch, GER

Both fantastic results

To be fair, it looks like Wyder was in front when they went wrong at #6, and Gemperle was in front when they went wrong at #9.
On the whole I'm pleased that Tove won, but definitely have sympathy for Wyder.

Hey Hammer... are you wondering where US is on that list? Aren't we part of NorAm, too??

Cherry is doing something strange at #8

Ali Crocker has 15th and 18th

If you like stats, you'll enjoy this page: http://www.maprunner.co.uk/wocdb/best/can/woc/wome...

WOC results database

Carol Gas from Norway

Hope he took his anti-bloating meds.

Sorry J-man that should have been top 5 all-time North AND South American WOC placings! ;-)

Canada is certainly on top of the women English speakers with two of the top 4. Congratulations from one who knows the tyranny of distance and a small sporting base. Who's going to shine in the men?

"may well have treated the middle distance as a warm-up for this event"

say what?

@Hammer: Hah!

That control flow from Thierry at #1 is ridiculous!

Sild already onto his 12-minute man!

GOOOOO TIMOOOOO

His route to #1 was brilliant

Timo Sild has no respect for Russians.

Andreas K, the great Swiss...

"checking his map, making sure he keeps in contact with it"... rather than letting it float along behind him

Maybe they are just practising filming Swiss runners.

Bergman caught Basset, Tero only a minute behind.

TG lost a bunch of time down and up those cliffs

4 fastest times to 17 are the Timo Sild pack

And 6th.. Just Coupat breaking in there in 5th

That's right, and they all seem to be running in front of Sild, so I guess they must be "dragging" him around the course :)

If they listened to Per and repeated what he said...

At the moment it seems so, always nice to have a pilot fish. Worked well for Tove

Or if they just broadcasted Per...

In fact, Timo seems to be getting gapped

NICE Nicolas!!!

So, at what point is it strategically superior to wait up for the guy behind you, because you can probably make up the 2 minutes by running faster together (assuming he's not waiting for the guy behind him too)?

We enjoyed the cluster of AP! nations

Nick didn't hang around long :)

Basset has a gap, TG can see Bergman's back.

Eric!

Eric beats Damian by 10 seconds!!!

@Bshields... depends who you have starting after you ;)

Nice!

one person between runners from same country puts them around the philoop in the same order.

Kaas Kyburz Hubmann Lundanes!

For sure, that's true, but I guess what I'm wondering is, if you let any two of the last 10 start together, would they put a >2min gap on the field of solo runners?

Just noticed that with France... but in reality, does it really matter? Are you going to follow someone more because they are from your country? Compare that to people who you could be in the same club with, same city, maybe even same house, but who are from a different country... i.e. Hollie Orr and Lizzie Ingham in the middle

These trains are ridiculous

I can't see anyone beating Lundanes from here, especially with his big pack.

Doubt it, only Thierry is looking like he might medal. Hubman is in 2nd half-way to 17

Is this what the IOF is shooting for? Big packs? I honestly don't know, but it must be intentional?

This Long is a farce and the IOF are a mess.

"head-to-head" racing.

Are we going to get to see the massive pack come through the arena? No, here's Jan Sedivy, coming through the TV control a few minutes down.

Yeah, don't show us the leader in the area, i want to see someone not coming in 3rd out at TV3...

Lind didn't collapse, he's going to be in trouble for disobeying team orders!

If it were part of a well-reasoned, well articulated strategy to improve the visibility of the sport, I might see it. But, if the idea is to ruin the sport without a plan to take 1 step back 2 steps forward, it seems like a bad move, IMO.

Lundanes making a mistake...

@Tim C: posted your meme on Orienteering Memes

oh my, Bergman is a broken man

2014:
1. TG
2. DH
3. OL

2015
1. TG
2. DH
3. OL

2016 ??
1. OL
2. DH
3. TG

What happened with Bergman?

Looks like Kaas got dropped by Lundanes & Co.

Stopped too early at TV2, then had a hissy fit about it, then started walking

WTF, this is the first time they've shown these camera angles!!!

Maybe they've been listening to us...

New discipline Long relay

Looks like Thierry has lost a little time on #23.

Loving the collar on the Russian jersey

Three French and three Norwegians ahead of the best Swede? Ouch.

FFS!!!!! Cut from Lundanes to Thierry

Norwegian 1 & 2! Except Hubmann!

Bravo Lundanes!!!

93 minutes for that is amazing

"Norwegians everywhere."

@ndobbs, that's what happens when you host WOC closer the where the Norwegians live than where the Swedes live

I just got back from the dentist - were Hubmann and Lundanes together pretty much the whole way? Or was that a recent development?

Thierry is taking some time out of Hubmann - it could be close for silver.

Becks>> Whole way, from #2, i think.

Three countries making up top 8....

Becks - they've pretty much been together since #2, except for the phi-loop.

In that case 93 is still spectacular but not surprising...

"this coverage is irrelevant"

Gueorgiou makes his move!

Come on Thierry!!!

Plentry of other high-powered packs, and anything under 100 is impressive

Allez les bleus!

Thierry going for his 7th Long medal in 8 years...

Basset will push Gueorgiou... whathwhathwhathwhthawhtahwthawthawht?

Why are they showing Bergman????

Here he comes... oh wait, no, it's Bertuks... if only we had some way of knowing, like the GPS, his bib, his kit, his face...

credible -> creditable

The calming view of Bergman all alone was like an old school cinema intermission.

Winsplits pack-index is going to be amusing

Last nine years: TG with four wins, OL with three, and DH with two. What an amazing trio of orienteers!

SJ must be regretting not outbidding Nokian

and NickS with a Long PB by 1 place - well done Nick!

"Rollier is a name we're not that familiar with."

Only twelve countries with runners ahead of NickS.

Mark Heikoop still going. On the whole I think I'll be happy if I can still get round a WOC-length M21E course when I'm 52.

one positive from the commentary is that it makes attackpoint more fun

Heikoop is a legend!

Eric beats Damian by 10 seconds?
that's a huge catch up for Damian isn't it? wasn't he far behind early on?

Did Bergman have a problem or did he just give up? 10 minutes lost after TV2.

Brian Porteus being interviewed on the TV commentary now. For our IOF President, the highlight of this World Championships has been... the TV coverage.

Is that seriously what he said?

Also, was there ever a conclusive reason that Hollie was dsq? Wrong #2?

What happened to Denmark?

Someone just posted this on Nope:

https://twitter.com/vcoupat/status/768829287734583...

re: Denmark
If only the race had been in Skane..... ;-)

is that the best Norwegian performance in the men's classic/long race since the Norwegian 1980's podium dominance?

Unfortunatly they disqualified Mark Heikoop, and that after 4 hours of racing!

Mark Heikoop
NED 14:25:00
72:25 (-) +44:14
118:53 (-) +72:43
172:11 (-) +106:11
186:41 (-) +115:42
230:23 (-) +145:40
disq

For my part, I'm wondering whether the French ever managed 3 in the top 8 or even top 10 before.

Too bad they DSQ orienteering's version of Eddie the eagle!

So here is a summary of all of the top 20 placings at WOC from North Americans (individual)

4th Emily Kemp CAN (middle 2016)
8th Emily Kemp CAN (long 2016)
9th Sandy Hott CAN (middle 2005)
10th Ted de St Croix CAN (classic 1985)
11th Emily Kemp CAN (middle 2014)
15th Ali Crocker USA (sprint 2015)
18th Denise Demonte CAN (classic 1985)
18th Ali Crocker USA (long 2013)
20th Ted de St Croix CAN (classic 1983)
20th Pam James CAN (middle 1999)
20th Ali Crocker USA (sprint 2012)
20th Emily Kemp CAN (long 2015)

Because of over the time limit?

I reckon some of those trains (a) had more people on them and (b) were moving faster than the one I took from Gothenburg to Stromstad this morning.

Too bad? He shouldn't even be there, he is a waste of a map!

Thats what they said about Eddie the eagle aswell, You will think differently when Mark Heikoops movie comes out ;)

So what forking method should the IOF implement in the Long that actually works to break up the trains? And would that allow an audience to follow the race and understand who is doing well throughout the race?

3 minutes start interval would already given another podium in both men and women

@jtorrance http://runners.worldofo.com/WOC2011Long.html
and multiple times two on the podium.

Just an idea... longer forks in faster terrain?

There's a time limit of four hours - looks like Heikoop was about 13 minutes over.

18th Denise Demonte CAN (classic 1985)
20th Pam James CAN (middle 1999)

These are two really underrated results in Canadian O history.

IOF shouldn't be worried about whether the audience can follow along or the time the race takes because the newer disciplines were meant to achieve that TV appeal.

As Wil Smith once told me the sprints are best to watch live at the arena, the middle and forest relay are best to watch live online and the long is best to watch a few hours after it has finished in a post-race produced 60-90 minute summary.

I'm not sure about that. I sure wouldn't want to miss watching the Long live.

@pi - I'd also be curious to know what conditions cause them to form as well as or instead of how to break them up once they've formed. My guess is the more controls you have, the more likely people are to see each other (especially if they happen in quick succession?), so maybe just sticking to fairly long legs and not very many controls would do it. The short legs aren't nearly as interesting to watch, either.

What is the point of having an IOF rule that says "12.17 The start interval in the [WOC] Long distance competition is 3 minutes"? This is the fourth year in a row that a two-minute interval has been used...

I think all alternatives are fairly well understood. Personally I think the best compromise is to stick to 3 min start interval, no forking, great long route choice legs (like today, minus the silly butterfly).

@pi, Wil's comments (which I agree with) were in regards to non-orienteers. ie., don't worry about others not trying to figure it out how a forking system for the long works because the race is too long time-wise for the general public to want to watch live anyway ie., they would rather wait a few hours and watch post-production.

Is the problem with today's race that the pack formed at the end of a long leg with lots of route choice? I didn't seem many (any?) trains going into #2, but after it was rare to see runners on their own (my unscientific observation)

If that is the case, then perhaps part of the solution is to get rid of such (beloved) legs.

As pointed out above, a 3 minute start interval would reduce pack formation greatly.

Funnily enough, I spent a long time looking into this last year.

You might think that Bernt Myrvold's comprehensivereport on separating runners in orienteering would be required reading for a WOC planner.

From that, and from my study of elite races, it's pretty clear that butterflies dont work. In fact, lots of short legs and sharp turns have exactly the opposite effect in creating packs. What's essential is that runners should never be certain that they can follow. A map exchange, or even swapover, is an effective way to do it.
Like pi says, great long route choice legs are the best option, and not having any of those after #6 made it pretty certain that the trains would stick together.

You might think the senior event advisor would know this stuff - it was all explained to him last year ...

But none of this is explaining why the packs formed today after a great long route choice leg. Perhaps it was because the leg came too early in the race? (I have no idea, just guessing)

I think just forget trying to beat it and go with it. Mass start with time bonuses for whoever punches a control first - a bit like cycling.

Well, that second leg really spread the runners out on different routes and if you did not execute well you could lose minutes. So with the 2 min interval you now had groupings. Then there was only one more big route choice to 6 and then the control picking started...

If TV window is 2,5 hrs: first start red group, 3 min interval, then next 10 runners, 2 min; next 10, 1 min and then mass start for others. In other sports work and it is still 30 min time to see at least best 20 in finish.

If you really want no trains set a course that you can't make mistakes on so everyone runs the same time (or close to it), but that's not the sport.

I don't think long legs separate runners, i think they cause groupings. If you have a route choice which has two minutes difference and a two minute start interval you straight away have a group of two.

The start interval needs to be such that if you are caught you shouldn't be getting a medal. Give the red group 3 mins and smaller intervals for earlier starts.

There will be trains no matter what you do, of course, but 3 min will help the impact on medals a lot. I am surprised that the small control-picking butterfly is still used when it was confirmed to be useless more than 5 years ago.

It is useless but not forbiden to use. 5 year learning curve is still very optimistic for IOF.

Not so much the length of leg #2 as 25 minutes of running that caused packs to form.

@pi: yes.

@AZ I looked at winsplits*, and don't see that there were many packs at #2. No more than you'd normally get after half an hour running. So I'm going to say I have no explanation because it didn't happen ;)

Gemperle caught up because it was at least a minute faster for girls to tank round the road than to bother with any of that skogsport stuff (you'd think they might have noticed this since Miranova won the gold in 2014 that way).



* order by passertider at 2. Assuming times are randomised, you have a 25% chance of someone arriving within 15 sec of you. That's about what you see.

If everyone competed at the same rate, then you would have 0% chance.

"I was taught that the long easy route should not be the fastest one. In a Swedish book on course setting."
If this was enforced, then "never take the long easy route" would become an iron clad strategy. So there wouldn't be much point in setting a long easy alternative.

Consider that going into No.4 there was a long section of running in the open, so where the long leg to No.2 brought a lot of runners into range to see the runner in front (probably that's just gaining about 1 min, so the runner in front was in about a 1 min range). During the long run in the open area they could just aim at the runners back in front and make up a lot of time. Probably enough to be in touch and then catch up the rest by No. 5, which was quite visible in the coverage too.

So I think it was a combination of the 2 min start interval and the long open area running to No. 4 combined.

Also, the trains that matter are the ones that include the place getters. If you don't want red group trains, then don't have red starting groups. If you want red starting groups for arena excitement and TV, then accept you will get red group trains. Seems like a clear choice to me, and the choice has already been made.

@graeme.
I only saw the women's race and that's where I thought I saw the bunching at the end of leg two. I looked at winsplits (you can actually press the UK flag and get the English words ;-) and, ordered by actual time of day at control two the following "pairs" had formed ...

4&5 (5 seconds)
6&7 (15 seconds)
11&12 (20 seconds)
14&15 (21 seconds)
16&17 (19 seconds)
18&19 (8 seconds)
20 & 21 (tied)
25 & 26 (5 seconds)

I'd keep going, but I'm starting to think there's some software out there that does this a bit quicker ;-) The pattern seems to continue

Maybe the leg was just too long. Or the problem is that bunching is inevitable and we need to stop break up bunches. Or something else. But I still think that before we can solve the problem we have to understand what is creating it. And I still don't really get it and I haven't seen any explanation of why people starting two minutes apart will be so close after picking different route and running at similar but different speeds.

I think Pauba is starting to get at the reason that trains are formed. It must have something to do with visibility. And long legs bringing runners into visible range.

Maybe they had to group together after the long leg to ensure that between the group they had the whole map - forget the packs, that map that Coupat showed is appalling.

Too much focus on TV, not enough on the basics.

"Too much focus on TV, not enough on the basics."

I don't think you can say this loud enough, but who should this be directed at?
I'm guessing at least one level above the controller(s), and I'm not ruling out IOF Council.
If there is an explanation why the TV tail is allowed to wag the O dog, it better come right way, and better be incredibly convincing.

@Graeme, I thought of your WOC course setting story when I saw the absolutely stupid, pack enhancing, yet NORTHBOUND TV leg that was deemed appropriate here.

If the IOF were really committed to avoiding the formation of trains in the long, how much would the quality of spectating/TV production be diminished by, for instance, alternating starters from the first half of the current start list with starters from the second half of the current start list? I'm assuming that would nearly always result in the higher ranked athletes having to catch up to whoever started two places ahead of them before encountering anyone who would be physically able to keep up with them, let alone contribute to forming a faster train.

Of course, the IOF's commitment to discouraging train formation is open to question or we probably wouldn't be having this discussion.

ETA: Also, the trains that matter are the ones that include the place getters.

Keep in mind, relegation and promotion for the next WOC are also at stake most of the way down the results list.

In Florida JROTC events we have fields about the same size but obviously not the same quality- far from it- as the WOC. But we have the same problem although with the JROTC it is not a problem of two or three great navigators getting together. It is two or three poor navigators jumping on to the tail of a good one and following him around the course. The temptation for the Florida athlete to hang on is great in their minds as they have medals, team points and ranking points at stake, to say nothing of getting out of the wild in one piece.
Perhaps the solution for the JROTC and the WOC problems are the same - a series of referees out on the course who can order time stops for one or more athletes when a group comes in to a control together. The time out would be handled like a road crossing and deducted from the athletes time at the download.

The desire to have the best all start near the end, so that the drama will be preserved, has been the basis of more than one awful decision on the part of the IOF. But at the same time, you have the athletes wanting to start late, when the terrain has been altered by the earlier runners. A fully random start draw, or a start list that would separate similarly-seeded competitors, would help a lot. But I don't hold out much hope of it happening.

You could make the start draw televised - like the FA and FIFA do for soccer competitions. A good 3 minutes package to stick at the front of your TV coverage.

Enough of this. We must make allowances that Sweden has only been organising orienteering for 100 years, the omens are there will be great improvements soon.

Now back to the English speaking league, looks like in the men, what my grandparents would have called "the old country" is at the top. But I note the appearance of Ireland next, we mustn't forget they speak English (of a sort). Australia also had a reasonable showing, NZ prob lulling it into a sense of security ahead of tomorrows' rugby test. Just a thought, due to differing number of runners should we be working on a per capita basis?

why the TV tail is allowed to wag the O dog
Part of the WOC planning remit is to enable great TV. It's an agenda I was signed up to. As a planner, its fun to think about how things will look on TV, but I'm not an expert.

Our TV producer was Karel Jonak. He was excellent. We spent time in the woods, he explained, patiently, why some of my ideas of "good TV" were wrong (too-similar locations, terrain that's tougher than it looks). He's used to covering sports, and totally respects that you cover the actual event, you don't create a TV version.

So its not the TV guy that wags the dog, its the IOF. I think they've got themselves so obsessed with the must-genuflect-to-TV that they actually believe if they aren't buggering up the course, then they aren't trying hard enough. It's so sad because its so unnecessary. Presented with the course as a fait accompli, and given sufficient time, Karel has the skills to make great, yet unobtrusive, TV coverage.

This is the same reason that the swimming coverage from the Olympics was scheduled so late - to cater to the (US) TV audience. Perhaps this explains why so many of the swimmers went on rampages following their meet (lack of sleep) and so we can justly assume that the orienteers will do the same (due to lack of good terrain to put cameras in).

ordered by actual time of day at control two the following "pairs" had formed ...

How can you justify some of those as "pairs" forming, just because they happened to be there at the same time? Such is the nature of orienteering. The pair that you say was 21 seconds apart at CP2 for instance started four minutes apart and were over a minute apart at the following control and that gap got bigger.

Also 30+ minute leg ftw :-)

Kemp was the top finisher that was either alone at controls or leading a group into controls. All others that placed ahead of her had at least a few controls where they were led into the control by somebody else. Gemperle was led into more controls than any others in the top 10.

https://twitter.com/worldofo/status/76890554247552...

But tRicky are you saying that the US tv audience likes to watch grass growing, ie swimming races? They should try rowing and canoeing as well.

Hammer are you Russian bashing? I've always assumed that orienteering is too minor a sport for Putin to include in his drug program, but you never know.

Hammer - is there a fuller analysis of these charts? (I can only see the top 10 listed)


The way I read the chart is bad news for the early long leg #2.
For the men three of the top ten were in a group at the end of the long leg and two of those stayed in groups the entire way.

For the women it was much worse with six of the top ten in a group at the end of leg 2, and three of those staying in a group the entire way.


And also bad news for the TV Zone
For the men four additional runners joined groups at control 5 (the end of the TV zone), two of them remaining grouped until the finish

The paper referenced above says that groups happen due to time loss (errors usually, or perhaps different route choices) or differences in running speed. And that groups are quite difficult to break up as it is difficult to entice runners into taking different route choices when they are together.

This all points to leg 2 as being a bad leg, does it not??? Contrary to the common belief that long legs with route choice is good for avoiding trains, in this case it actually created them.

So, as much as common wisdom says long legs with route choice are good for avoiding trains, I think there is more to it than that. Certainly in this race they appear to have helped make trains.

No. The interesting long leg caused time differences. The start interval made trains.

The problem is that the leg came too early. It looks like trains were forming after 30-40 minutes, due to running speeds, mistakes, only a 2 minute start interval, and an open leg for TV. So if the long leg came after 25 minutes, then it would have done a much better job of splitting people up. However, the plan has gone for the classic mega-short leg into mega-long leg. I do this as a planner myself, as it doesn't give people enough time to plan ahead properly. However, in this instance it's bad for a few reasons. There were 3 legs with signifcant route-choice, and this was one of them, and the longest, and so it did nothing to break up runners. But also, the 1st control ended up being a re-start, and it was filmed, and a lot of people had to stop to make their route-choice. Is that really the image we want to portray on TV, people getting to the first control and then stopping? We understand what is going on, but the non-orienteer, who it seems at TV is aimed at, is just going to think we aren't very good.

@AZ, you need Winsplits Pro, which I have on my other computer, I can post the full analysis later maybe.

@Bshields, Interesting that Sild did indeed get dragged round that course, he got to 1 on his own, and maybe 5 others as the leader of the pack. Everything else he was behind. Bakkman and Nykodym do most of the work in the pack.

If the beginning had been short legs, that would be getting blamed for trains forming. It's the start list/start interval that's a problem.

@Nixon/bshields. Doesn't mean he got dragged. It was fitness, not nav, that was holding Timo back, by the looks of things. Well... he gains some seconds like that, so okay...

I realize many people are blaming the start list/start interval for forming the packs - but I think it is worth looking at accepting those as given and trying to figure out if there is a way to reduce packs forming within those constraints.

To that end I agree with Nixon that having the long leg a bit later would have been better - the long leg might have been used to stop trains forming just before they were expected to form.

Sure, it's not going to be "pure" - but it would have been much better than it was without requiring any thing more than (IMHO) better use of a good tool.

Some photos from the long.

Unbelievable that something like this can happen to a world championship and especially in a country like Sweden ... More disappointingly, even the slightest excuse or explanation by athletes or their delegates from the organizers.

https://twitter.com/vcoupat/status/768829287734583...

Vincent Coupat :"Oui... J'étais vraiment énervé là bas. Et totalement impuissant.
Le pire est qu'ils connaissaient le problème, et que leur seule solution a été de nous proposer 2 changements de cartes durant la course pour qu'on puisse remplacer nos cartes abîmées au cas où... Foutage de gueule au finale."

Pasi Jokelainen wrote in twitter that same happenned to them couple of months ago (Kainuun rastiviikko?), ink just did not stick. Same paper brand and type they have used a years, quality of one batch just was different. Possibly same paper here, very unlucky. Apparently you need to test it even if you use same paper as before, paper known to be good..

Statistically, after 30-40 mins of running (whether it is a long leg like here or a number of shorter legs), people with 2 mins intervals WILL get close to each other.
Simple to think through: people will make mistakes and/or choose better/worse routes, so you'll have the situation that two subsequent runners would have covered that section with one having at least 1min better time than the other. This will bring them to relatively close proximity (because we have people in front and behind, time difference will bring people closer to either the person in front or the one behind). This is natural and simple statistics will confirm that it is inevitable (at least with 2 min intervals), and not necessarily a problem in and of itself as we do not yet have people running in bunches. The question is what do you do at that point. The issue here was (as I was writing before) that at this point of the race there were two large areas of open with good visibility ahead: the last bit of leg 2, where most runners have attacked the control from below, previously running along the open area for a while (giving the opportunity for the runner behind to see the back of the runner ahead and close the gap somewhat even if not fully), then the large semi-open area to control 4. What this meant is that there was enough time/distance for people to close the gaps and then give the opportunity for either to choose to run together.

If there was less open area before the second longish route choice leg (5-6), it would have been more likely that people would have made differing route choices as they could not see where the runner ahead was heading out to, and that would have meant that sometimes the faster runner could have overtaken the person in front (thus no longer bunching), or alternately, the person who was almost caught up, to stretch the distance back up.

There will always be "jitter" in the speed of competitors. You gain a little here, you lose a little there. These variations will cause people to encounter one another, no matter the lengths of the legs, and if they are of similar abilities, they will stick together like magnets. True forking can spilt them up, but you can really only do that in a relay, butterfly loops don't work very well. Allowing a control skip can work sometimes, but I have trouble imagining that in a major international race.

Long map - autogenerated comparison animation positioned on Google imagery. Use opacity slider at top right corder to see aerial imagery.

http://routegadget.net/kartta/?mapurl=http%3A%2F%2...

If we think out of the box maybe instead of the butterfly a section with a few controls where you need to choose a different route choice. It is a pressure on the athlete to read a map and decide where to leave the pack. If not, you get + 3mins. Apply this to CP 2-3-4-5-6

jjcote, sure, once they converge and get close enough they can stick together.
However we can look at the circumstances where we make it more likely that those that got relatively close would be more likely to close the gap - and that is open areas.
Having (long) route choice legs BEFORE they have a chance to see each other reduces the likelihood that they end up "within closing range" because there are more variations on how they can open up gaps (lateral as well). Like I described, the runner catching up can overtake on a different route, or the one just in front can gain back time. The former possibility does not really exist on shorter legs because passing close by (i.e. seeing one another) gets us again to the magnet effect.
I think the key is decreasing the chances of bunches forming - you can't avoid them completely, but breaking them up is much more difficult.

kofols - like the Berkley Marathons! First person back gets to choose which way they do the final lap, next person has to go the other way.

I don't like it in orienteering - if you know that section is starting, you hang behind, take it easy, then sprint past the navigator just before the control - you get the choice, they lose.

@jagge - nice image - do you reckon going off the map on 1-2 is fastest :)
How did you like the form lines?
Why don't we get to see the north lines?

I wonder did they have base map for that part or did Lantmäteriet scan Norway too.
I feel like doing the Vincent Coupat trick to them.
Maps looks better on TV without them?

In relation to splitting packs, is the forking which was used in WUOC long this year a good idea? Usually the loops in a butterrfly are done one after the other but these were done very far apart, effectively having the power to split packs up twice in the race.
Didnt run the race myself or anything, just speculating.

Greater visibility requires more runner separation at the start in order to reduce pack formation. I'd guess that this is true irrespective of the leg length. It can be accomplished with a bigger interval, or with seeding.

(Ever orienteered in Wyoming? Hoo, boy!)

If placing the long legs carefully in the Long reduces trains, then that raises the question of whether predictable location of long legs (within the course) is a good thing. I suspect that the course designers wanted to challenge the competitors by having a long leg so very early. Is unpredictability a good or bad thing?

Also, for TV, tape delay is a possibility, as is so often done for Olympics, making start interval perhaps less of an issue. Someone will be in the wrong timezone to watch live in a global sport (which orienteering is slowly continuing to become, such as with a fourth place Canadian finish this WOC).

Have the WOC 2016 organisers published details on the paper type and printing press that were used?

On predictability of long leg - I can see the argument, but not sure it means that much if orienteers know up front that longest legs would come after, let's say, 20-40 minutes running. There are still many more variables, surely - not convinced that that knowledge would change ones preparation/race tactic.

The only thing I can think of is the knowledge that one would not have the need to plan ahead on the long leg immediately, as it would be possible to defer that planning until there is a reasonably simple to navigate section within the first 15 minutes and then dedicate some effort into planning that leg properly. However, this knowledge is something that can be gained in about 20 seconds (max) of the start anyway, so not sure it matters that much.

The most important is to fix 3 min interval at least for the red group and then trying to implement other solutions to minimize pack formation. The problem is that most people see this as a main problem and we also written this into the rules but somehow - FOC, IOF council and organizer always find underthetable agreement why this WOC can be exception from the rules. Do we have such examples also in other sports? Do people in IOF able to play with any rule if they wish?

Another reason for trains were random blank areas on the map, runners heading to the general area and waiting for someone with a legible map to follow. I hope they can organise the banquet better!

The problem with the maps must have been picked up by the organisers beforehand, as competitors had opportunities to take fresh maps at various controls around the courses.

At a major event in Australia some years ago competitors turned over their map to find that the first control was 3.5kms away, with multiple route choices. Despite many competitors staying at the Start point for a minute or two to study the map, I don't think that many trains were formed. It doesn't matter where the long leg is in the course, and competitors should not know it in advance.

Agree with jj - at WOC this year the cause of the trains forming was mainly the 2 minute start interval, then the good visibility at the end and just after the long leg, and this was exacerbated by having a red group.

Agree with JimB, both tv and gps can be replayed, so there really is no need for a red group, or the red group can be bigger. For example, in a field of 70 you could have the 35 or 40 lowest ranked runners starting first, followed by the 30-35 highest ranked runners drawn randomly. I think this would add interest to the spectators at the course. Seems to me that in the countries where orienteering is followed seriously on tv, the audiences won't mind the longer program, and in other countries the broadcast can be edited so that only the top runners are shown with those starting earlier in delayed gps/viewing mode.

@Simmo - no disrespect to Austrialian runners, but I don't think comparing an Austrialian relay to a WOC relay is particularly valid. The spread of running speeds and navigational abilities is way different. Heck, in Canada we just had a sprint relay with zero forking and pretty much zero trains formed. I don't think it would be valid to say that lots of short legs ensures no trains will be formed ;-)

Watching the relay is way more fun than watching the long. Just drop the long from WOC and embrace trains. The solution is pretty simple ;-)

AZ - it wasn't a relay, and the comparison to WOC Long is valid because I was just making the point that if you are going to have a very long leg of 3km+ it doesn't matter where in the course it comes from the train-forming point of view.

Sorry - I have relay on my mind for some reason ;-)
But my point is that the calibre and depth of the field must be taken into account. The variation in travelling speed in Australian (and Canadian) events is much larger so, in particular, trains are difficult to stay in. Well, maybe I should only speak about in Canada - but I assume it is not that different in Australia. So even if trains do form, they split apart quite easily (that was my point about the Canadian Sprint relay, for example - we had a mass start with no forking and trains didn't last long). But at WOC breaking up trains is very difficult.

AZ, I'm pretty sure that the event Simmo is referring to was the World Cup a few years ago so yes, it was a world class field.

As for the point about where the long leg is... it seems to me that it depends for more on the quality of the terrain then the length of the legs. As mentioned above high visibility less technical terrain is going to bring runners together more than dense / technical areas so if you're long leg has some route choice options with easy running and high visibility then there's a decent chance of packs forming. If all the route choices are more technical then it likely won't make as much difference.

Finally, I think it does make a difference if we know where the long legs are likely to be in a course. The amount of prep work WOC athletes and national teams are doing these days means that any restrictions on course setting make it exponentially easier to guess what the course is going to be like and where those route choices will be.

Okay - I stand corrected (again!!) as to the calibre at the Tasmania event.

But never let it be said that I won't stop flogging a dead horse ;-) (and I perhaps still don't have the right race!)

At WOC in Sweden in the women's race from position 1 to 13 there was a time difference of 5.5 minutes. In Tasmania the first 13 were divided in time by only 1:42 and you have to go to position 26 until you get a time difference on the leg of 5.5 minutes.

So, I would theorize that the legs aren't that comparable. The Swedish leg definitely separated the women better in time - which ironically is one factor that brings them together in trains.

In other words - can this be true - the better the long leg is at inducing time differential the worse it is at creating trains?

The Tassie event in Jan 2015 had pretty much the A teams from Sweden, Denmark and Switzerland along with some other top European orienteers. And of course the Australian and Kiwi national teams.

@AZ@Simmo
I think you're missing the point. After 30 mins of a classic trains will form. Maybe less so if there's been a long leg. The real question is whether you can use a long leg to BREAK UP trains once they are formed. Myrvold's study provides evidence for this. If you start the race with a long leg, before the trains have formed, it can't break up what's not there, so as a planner you are wasting your best trick for train-breaking. Nobody started with a butterfly yet, but I suspect something like that will appear as the new urban mass-start format.

The Long race at the World Cup in Tassie had the long leg from 2-3, but not that far out of the start.

The event I was referring to was back about 10 years ago, and the long leg from the start triangle (it had a few variants) was set for most age classes from 16 to 60. At the time my class (M55 or 60, can't remember) was very competitive, with over 50 entries and at least 20 very good runners, and we had a two-minute start interval. The terrain was steep, typical Australian gully/spur, rather similar to the map in the link above except less rock but more complex contours. I think I ended up about 5th, I saw about 5 people from other classes, but no-one from mine on the whole 3km first leg. In fact I didn't see anyone from my class until near the finish of the whole course.

AZ perhaps your theory is correct, but if so isn't it one of the aims of the course planner that the long legs on a course should be the most decisive? In the case of WOC 2016 it appears that the second long leg created most of the time differential between Olav and Thierry in the Men, while the 1-2 leg was certainly the major factor in Gemperle's silver medal run, and the 3rd long leg from 8-9 gave Tove the gold.

Let me go back to an idea I posted about 150-200 comments ago. Call it a train. Call it a wagon. It is the same problem in WOC since a world class marathoner followed a world class orienteer around the castle grounds in Scotland forty years ago.
It is the same problem we are having with high school competitions in Florida where we have upwards of 80 participants in a class and must often go with one-minute start intervals. And the kids soon get to know who the good ones are. It is the same problem whether it is a pack of two or half a dozen or more. How do we break up packs of orienteers where an inferior navigator is gaining a better place than he/she deserves based on merit and where two good/ very good navigators are improving their placing by colluding, intentionally or not.
I don't know if this solution will create other problems but I would like to see tried the placing of referees at random controls in the forest where when a pack arrives one is sent on; the others are held back to create a space. It is easy enough with existing technology to deduct the hold back time. It would be easy enough to create a protocol as to who gets held back. It would be easy enough to block map study during the time out period. I'd like your thoughts but I'd also like to try it and let you know how it works.

Stopping for a minute and getting cold, losing one's rhythm, having heart rate change could be detrimental. Having a minute to read the map could be beneficial. It's less equivalent a race then. And some may learn how to game the system to their advantage. It makes the sport more complex; the simplicity of "navigate to these points in order, fastest wins" is nice. A mass start Motala (individual relay) might be better (but could reduce the opportunity for very long legs). Or accept that trains are part of Long, and increase the starting interval to 3 minutes to reduce their prevalence. Maybe a campaign to IOF to make that the new actual starting interval, at least for the top orienteers. ;-)

Gordhun, that is impossible to referee. Who do you hold back Iin Hubmann Lundanes? Yeah, Hubmann lost time early on, but he has more medals than Lundanes. Plenty of other people lost time on that leg - should Hubmann be punished just because his time loss meant he hit the control at the same time as Lundanes? When others go unpunished? How are you supposed to recover from a error if some guy yanks you off course and holds you for a minute?

The only thing to do is try to minimize train impact with planning, and I agree with Jim and many others, a longer start interval.

For your school stuff, you could do something like a butterfly that starts with a map exchange, and the maps in the box alternate forking. That guarantees that someone hanging onto a superior navigator will at least need to hold his own for a bit. If he succeeds, then I guess he's not so bad after all.

I don't see this as a useful approach for international elite races, though.

@becks You hold back whoever got caught. Just like in bicycle timetrialling. As mharky and the 95-minute-women keep reminding us WOC is about finding the World Champion, not the best losers. Nobody ever won WOC after getting caught (he asserts without checking).

Not sure it would work though, what if you're held back a minute, and someone else arrives? And how close is "together".

Longer start interval would help too, obvs.

IOF have experimented with the start interval & start list to have a better TV coverage. I think TV producers and IOF should do something based on the statements from the first three athletes at the press conference after the race. It would be appropriate from the officials to say something or at least to acknowledge the athlete's frustrations. No one wants to hear bad news but we need to say a big thanks to Jan to publish these statements. Official IOF media makes only propaganda. They need to learn how to deal also with the bad news.
http://news.worldofo.com/2016/08/25/woc-2016-long-...

Nobody ever won WOC after getting caught

We should also ask has anyone ever won after being 2 minutes behind leader at some point of the race. And the ansver is yes. That happend three days ago (Alexandersson).

In addition, someone (leader) might get caught for someone accdetaly skipping a control or making a mistake and hitting a control too early by accident.

@Graeme, I think Khramov won JWOC long having been caught by a Swiss, who then blew up in the heat.

@Jagge, It is of course possible to make back 2 minutes in a race, but much harder to drop someone 2 minutes. Bakkman got caught by Sild, but managed to put the 2 minutes into him on the last loop.

At the JK long in Scotland I caught Oli Johnson 3 minutes quite quickly, hit the wall, had a sit down, and crawled over he line 2:58 after Oli for a glorious 2s victory!* :D

That said, in general, now that the standard of WOC is so high, once you are caught it's probably game over, tuck in for bronze.

*I think I was actually 3rd on the day though, but they way I felt for the last 30 minutes, just finishing was amazing :)

@Nixon "tuck in for ...?"

Gold
I decided to not use [Tove] as a reference. So I pushed and was in front from the 10th control until [Alexandersson passed at] the second last control. AMH Nordberg

Silver
Judith was leading quite a lot in the end. We worked together. A few times I saw that she was taking another route, but then I thought that it was faster to go where she went. N Gemperle

No disrespect to any of the athletes, indeed more power to them for honesty. We've all done it. They can only run under the conditions put in front of them.

Also - what kofols said +1

Metaphorical ;)

Thanks for that link Kofols, I hadn't found that article on WorldofO before you posted it here.

I used to be a proponent of qualifiers and allowing more small countries at WOC. I'm starting to wonder if the solution isn't, in fact, the opposite. Only have 40 people start the WOC long distance. Keep things 'short and sweet' and limit participation to those that truly belong amongst the world's best. You could then have a 3 minute start window without making it worse for TV and not needed to worry about those athletes the TV folks don't care about anyway.

Of course for this to not be disastrous for the develop of high level orienteering nations in small countries you would need to compensate with other international competitions. For one, the World Cup would need to be beefed up and made more 'World' and there would need to be qualifiers re-instated for the Middle distance (which, thankfully, there will be from 2021 on). The World Ranking system should be revisited and turned into something the athletes care about (used for number of runner allocations at WOCs and World Cups) to give meaning to these events for athletes and national teams.

If there are other meaningful competitions that smaller nations can participate in then it's far less important to have the WOC long distance open to them and would give smaller bite-sized goals and targets for them. It would probably go farther to helping these nations develop then sending athletes to come in 70th at WOC as happens now.

@Canadian, the qualification and final system did allow more athletes from small nations to compete at WOC, but you had to qualify for the final by right and earn your start spot (rather than using a non-statistical WR score from a championship 2 years prior), and it was 45 people, short and sweet.

When I ran the middle final in 2013, I finished 10th in my heat, and started 17th in the final. If the current system was used at that time, I calculated that I would have started first because everyone who I out qualified had better WR than me. It's BS.

To get a good result in 2017 you're going to want a late start as it's going to track up loads, and you'll be in better trains if they form. But to do that you need to spend a lot of money playing the WR game, which itself has startlists based upon WR. How do you break in?

Just playing a little with the TV angle here.

So if starting interval/field size etc are all adjusted to cater for "better TV production", there are still arguments for keeping start windows 3 min. Although, to be honest, I think there are a lot of ways the TV production can be made much better before we start considering start windows (and let me graciously "skip over" - haha does anyone get the reference? - the quality of the commentary here).

Nevertheless, let me think aloud a bit here on the pros/cons (strictly from TV coverage perspective, just to try and see if that pressure is even real or not).
Longer starting windows clearly limit how much of the field "fits" into the coverage window live. But, edited coverage (as someone already said) probably would be better for general consumption anyway. It would present a challenge of covering a specific control and sometimes having to wait quite a lot before the following orienteer arrives - heck this year Cat was probably only covered in her start and maybe once more, but as she sadly did not have the best run, the coverage basically forgot about her. As the field thins out towards the end, getting back to a control to cover those not having a good run but being out late (and not running in a train...) becomes less interesting - but this then introduces longer "wait time" in between areas of coverage for those that are interesting from a result perspective.

So, can we propose to do something interesting in those wait times?
- It could be used to switch to proper quick GPS analysis on the interesting developments or route choices.
- It could be used to cut to short clips of the "life stories", or "route to get here" on the top orienteers in the forest (would have to be pre-prepared, of course), or of useful introductory clips of aspects of orienteering, such as what is electronic punching, basics of running on compass bearing (especially in legs where we'd expect runners to do that), route choice and other techniques, map basics etc.
- It could be used to purposely cover short aspects or summaries of the runs of some of the runners slightly off the pace. At the moment, the coverage is basically random at the beginning (whoever gets to the controls with cameras) initially, then only covers those beating the best splits so far, then exclusively the top 2-5. Not a lot of coverage on those finishing maybe 8th-15th who ran late. (I understand why it is, just suggesting how they can be covered in a differently edited production with 3 min starting intervals. Surely this would be welcome especially in markets with large potential viewing audiences but without orienteers who could win)

I am sure there are other options. All of the above would, however, require more effort and work to be put into the production.
Without the additional effort, is the 2 min interval better for TV? It might be. In fact, for those not into orienteering, even the coverage of trains may be more exciting: easier to understand (head to head racing), seen on the screen, as opposed to the time trial prospect. It would be interesting to see what the viewing figures are of the Tour De France time trial stages vs., let's say, the mountain stages.

Another question - can we do more to cover the "time trial" nature of the sport better? It would be interesting if we could get a couple of running camera man cover the same leg(s) with a few subsequent runners, then put it on a screen side by side, a combination of how, for example a Formula 1 qualifying comparison is shown side by side for the top 2 runs from the onboard cameras. Combine that with an overlay of the synced route coverage (as it is done in some of the introductory leg(s) covered at the beginning of the coverage). I am sure I would be glued to the screen to see and compare some of how the micro route choice decisions are made by the top guys, compare their map reading routines etc...

commentary was excellent. As pretty much always. (in finnish by Janne Salmi & co, Janne knows pretty much everything and everyone).

The more challenging the navigation, the more spread in the athletes' performance, the higher the potential for trains forming. Long legs or short legs.

The potential advantage of a long route choice leg vs short control-picking is not primarily that it breaks up packs, it's rather that if each athlete has to solve it themselves, then runners spread out on completely different routes through the terrain, opening up the chance to pass another athlete without ever seeing them. This is much more unlikely with short legs in most forest terrains where athletes are forced into the same running corridor.

One nice thing about the 2nd leg at WOC was that it ensured that each athlete had to make the decision alone. Exactly what we are looking for in this discussion, right? And what great theatre the first 30 min of that race was, especially when you had the GPS tracking going in parallel with the TV. Later the race got predictable and much less fun.

Of course, there are trivial ways to ensure at least some breakup of trains. A loop system/butterfly of some kind with very unequal length of the loops, where one loop should have a long route choice leg. Say at the half way point of the course or so when significant trains have typically formed. But one argument against is that it makes it harder to follow for the audience. Thus my question hundreds of posts ago. Is this why IOF is not willing to do it?

One nice thing about the 2nd leg at WOC was that it ensured that each athlete had to make the decision alone. Exactly what we are looking for in this discussion, right?
I'm not sure I agree. What I want to see almost all of the decisions made by themselves. The trouble with the 2nd WOC leg is that they runners made one (very cool) decision by themselves and then almost no more.

I'm not clear in my own mind, but I tend to think the suggestion earlier to use long legs to prevent trains forming is very good. That the long legs should come just before runners are expected to converge on each other. Then perhaps some butterflies or something to try to break up any trains that do form in spite of that. If that does work, then it gives us lots more independent decision making.

Another problem I'd say is that the women's leg was simply too long - it almost guaranteed trains would form because the running / navigation differential is relatively high among the women (compared to the men). I suspect that if the women's leg had been shorter then fewer trains would have formed (way fewer formed in the men's race at the end of that leg).

I think it is time for those guys to write a second paper analysing long legs ;-)

AZ, the greater variance among women typically leads to more/longer trains regardless, after 30 minutes of running time.

OK, so maybe it's possible to delay the inevitable by having the long leg 15 to 20 minutes into the race? There will still be the occasional train already, a few more for the women, but that might give slightly more time with independent navigation.

Again, long route choice legs don't break up trains that have already formed, at least at the WOC level, as it's very rare that an athlete has the guts to try some other route alone.

So I'm still curious to understand a bit why the proven-to-be-useless mini-butterfly keeps showing up in the WOC Long courses? From the course planner or the IOF advisor? Especially since it's trivial to actually guarantee some break-up of trains with very unequal length loops? Why is that not used?

I don't see much difference between having two lops with equal distance versus two loops with same distance. Pair members will run that forked part independently and only if either of those two is >30sec faster they will continue independently. So what matters is how long those loops are in total together and how challenging is the forked section, not the length difference of those loops. But of course having one long and one very short loop makes it possible to make that long one more challenging. Like first trying to design one as challenging leg as possible on race map and then designing long loop and rest of the course around that. But simply having two loops with different length will not make any difference.

Of course it will make a difference. Olav and Daniel are in a train half way through the course. They come to a butterfly where one loop takes 3 minutes to run and the other 30 minutes to run, including a challenging 2 km route choice. Olav has to do the short loop first, so now Daniel is forced to make an independent decision how to solve and execute the long route choice. 3 minutes later Olav is also forced to make that decision alone. Perhaps they are so equal that they come out of that within 30 seconds 35 minutes later, but now they were forced to navigate independently for a significant part of the course.

You are describing what I wrote above. 30 minutes is quite long forked section, 30% of the course. You would get same effect with two challenging 15 min loops. Only difference is it is easier to design challenges when legs can be longer.

OK, I understand what you are saying now. My point was that this is the Long and we want athletes to solve long route choice legs independently, where a 15 min loops is a bit short for that.

So again, why are long loops not used?

@Jagge - in a phi loop, equal forks mean you run the central section together, including all visits to the common controls, so typically it's about half as effective.
@pi - they did use unequal leg lengths (>1min) this year. But the whole section was ineffective.

The women had no forking. Try planning a course with arena start, the given TV controls and an arena pass through and you'll see why.

@AZ - very good point that the number of important, independent decisions should be large. Interesting that you think you shouldn't have a long leg early because it makes trains - I think you should have it late because it breaks trains: we completely disagree but come to the same conclusion about what should be done!
Now you throw in a completely different argument which I do agree with - and still the conclusion is the same!

On the women's second long leg, only two pairs stayed together Gvildyte/Bassat and Gemperle/Wyder. Even the men's loop breaks up almost half the pairs - but create many new ones!

Try planning a course with arena start, the given TV controls and an arena pass through

I did that some days ago.
https://www.facebook.com/177518995597572/photos/a....

@graeme - you think long legs break trains that have already formed at WOC? Really?

@Jagge - the 3 min/30 min butterfly is the same effect as the streamered delay, but I agree it restricts course planning a lot because you have to come back to hub. But I think you can still do it with good TV and good course with some creativity.

@Jagge - wrong link? That's just the WOC course without loops.
[edit] - thanks Cristina.
That should work, though you may want more than two dead-runs since a lot of pairings are +-4 mins.

It's the WOC course with two streamered bits at controls 6 and 10.

@pi in addition to it being less restrictive there are other advantages. Every third starter has same spreading (in 3/30 butterfly every second has same). Also in 3/30 loop later starter may join the previous starter for the short loop being too close to the start interval.And it is hard to make short enough loop (40 sec loop) to make it not happen. In this streamer method the later starter never gets more than one streamering run closer to the previous starter (~40 sec) so to make them meet the later starter must have closed most of the gap on it s own simply by running better. And because delays are quite constant and known in advance GPS tracking can easily be set to compensate it in sync mode.

But as noted, this spreading is illegal and must not be used anywhere without special permission.

So again, why are long loops not used?

Perhaps TV - they like to know who is leading at any one point in the course. During the loops you cannot tell who is winning.. The longer the loops the greater the % of the course for which that is true.

Also as Jagge says the best way to include long legs is to have 1 short 1 long loop. So two 15 min loops doesn't give much scope for long legs but 5 and 25 does. However unequal loops exacerbates the problem of knowing who is leading and can cause extra confusion.

A couple of examples:

The loops were unequal in Ukraine 07 and the runners passed through the arena during the loops. Arena commentary team was unaware of the different lengths and was saying things like 'A is 5 min behind B at the spectator control'. Some runners heard this misinformation and it affected how they approached the rest of the course. I think one runner sort of gave up trying to win (or place or something?) thinking she was way off the pace when in fact in she still had a good chance.

In Finland 2013 the loops were different lengths - the TV commentators didn't realise the implications and were playing a head to head GPS tracking clip of two women who appeared to be neck and neck when in fact one of them was actually well ahead, having already done the longer loop.

Another thing: They did short and long loops in France 2011 - seem to recall Thierry was critical of that - something about one loop being more technical and it was unfair to get the more technical loop last when you were more tired.

UKR and FIN is poor commentary; how can anyone commentate a long race without having studied the course?

FRA, I have to agree, they came very late and were quite different

For commentary, they need to switch to a golf-style idea of par time.

The liveresults software already has a good algorithm for that and the idea of "leader at the finish" and "leader on the course" is familiar to most TV sports watchers. This can easily cope with legs taken in a different order.

@pi Long legs can break up trains, yes. Only two pairs survived the women's second long leg. Even the common first leg of the sprint relay broke things up. The main advantage of following comes near the controls, and it's very difficult to break away if you have to keep slowing to find a control (this is especially important in relays).

Oh come on. The 2011 Long was a 5 min delay loop to enforce break-up of trains, the 2013 Long was a 3 min delay loop to break up trains. I don't see how such a short delay loop is unfair because the athlete is so much more tired a few minutes later. I don't think you can say FRA was bad but FIN was good.

Anyway, totally agree that it's simply poor commentary.

That should work, though you may want more than two dead-runs since a lot of pairings are +-4 mins.

Two is enough to break groups of 3 consecutive starters. One runs it twice, second once and third none. So your 6 minute man can keep on running together. I ahven't checked was there any such groups that also did stay together for longer than just couple of legs. But 6 min is too much to get to medals and with 3 min interval it would be 9 minutes, that should take you out of medals.

Good commentators can handle loops just fine. Past bad commentary is not a good reason to not use loops. And we can take splits from gps tracking and compare in sync gps with short loop split subtracted.

For those who wonder, the loop counts athletes needs to run at the first dead running section is based on start number and the scheme is 2,1,0,2,1,0,2,1,0 ... so in 66% cases thelater starter runs one loop less and gets ~40 sec ahead, and in 33% cases ends up 80 sec behind. (this spreading has its own disadvantages and and this all is of course just speculation how things could be in a parallel universe)

Actually, 2013 was approx a 3/20 min loop and 2011 a 5/30 min loop, in the last 3rd of the race, so basically what I'm arguing for.

2012 and 2014 did not have any forking at all. Anyone have a train analysis of these 4 years? ;)

Sorry 2014 had forking for the Men... looked at the women's...

I don't think you can say FRA was bad but FIN was good.

I wasn't saying that - just trying to guess at reasons why WOC organisers/IOF might be adverse to these things. I get the feeling that not many WOC organisers/IOF read discussions like this on AP because of they did they . . .

I suspect in Ukraine (pre GPS tracking?) the commentators might not have been allowed to see the course beforehand. Certainly that happened at a WC in Australia 2000. There was a mass start long race with 3 loops. The commentary team asked to see the courses beforehand but the organisers refused. The 3 loops were unequal lengths and they would not even tell the commentators that. Not surprisingly the commentators were often wrong about who was leading.

Jagge, the out and back loops look like an interesting idea.

The only concern I have is that IF there isn't a train, in a lot of cases this loop would actually probably increase the chances of a train forming between the two controls with the streamer section.

I guess it would however work with some small adjustments:
- The (first) streamered section would be at a control before a middle-style section (i.e. not at the start of a route choice leg which would more likely separate runners, whereas a middle-style section would favour the train to stay together)
- The streamered separation in the first control would only be applied to people within trains, i.e. if people arrive within 10sec to the control, the 2nd runner gets 1 loop, the 3rd runner gets 2 loops.

The above would require a communication link between people manning the two controls so that the relevant equalisation of runs would happen, and anyone not getting a delay loop in the first control would be asked to do 2 at the later control.

I still have a bit of concern though because I feel if people have 2 loops that gives them more time to plan ahead as opposed to the people not doing loops. I know it equalises over the two streamered section, but planning would need to be really excellent to enable people to take advantage of being able to plan both upcoming sections with equal value. And even then, some people would have planned much further ahead than others, so having this additional loop would not be of equal value for them...?

@robplow: Not being allowed to see the courses...

The Swedish Mapping Committee told us at the conference that they had offered their help in making sure the WOC maps were of good quality, but the organizer (SEA probably?) had veto'ed that: Nobody outside the designated mapping team were allowed to give any input at all.

The solution could be to always include one commentary support person and a country/IOF MC member in the organization?

It is well known that runners do mistakes towards the end of the course where the pressure and tiredness are the two most decisive factors. If nothing works as it should (long legs, butterfly,...) we can have a biathlon 250/300m loop within the arena to break trains.

The later starter have a privilege to run the loop at the end of the course just before the finish and the other runner or runners have to do it when they are passing the arena. In this case TV should follow only the leading runner to the finish and then skip to other runner with sync GPS routes for the last loop or just LIVE to see the last meters towards the finish. I think this could be better TV picture also for commentators. They would have time to focus on each athlete separately when they are coming to the finish.

@Psuba this idea is already an old idea. It was thought to be bad bad idea to have marshals saying on the fly who should run loops and when. Much better if competitor and TV/GPS team knows it in advance. Human interactions / communication during race is asking for trouble. For the reading ahead advantage there has been several suggestions, like having map exchange or not allowing map reading during the streamer section. Or allocating this advantage to qualification heat winners, so this advantage could be earned (in parallel universe we still have qual race). Or not having route choice leg right from the separation making it less big advantage. No idea what would be best. This type of spreading has been used only once and field was there so uneven there would have been no packs anyway.

What it comes to forming packs, it is easy to study what would have happened by studying splits of various races, just pick a control and imagine that's the spreading control and add the delays and see who are together and which packs are separated.

Terje
as a mapper I think I would prefer not to have the national mapping committee getting in the act as well - getting caught between conflicting opinions is a recipe for time wasting and frustration. There is already the IOF SEA and a national controller - who is subordinate to the SEA - that should be enough oversight.

Sounds to me like in this case maybe a bit of politics/power plays/personal rivalries... might have been involved???

I would assume the TV producers are involved in the course setting process - and have access to maps and the terrain to check out camera locations, etc. The commentary team don't need to be involved in all that but i would have thought giving them access to the course maps a few days before the event would be useful to help them deliver a better performance on the day helps if they know what to expect and have planned ahead for likely scenarios. Also suggestions like Jagges to subtract the shorter loop split needs some planning and programming I imagine.

Yes, I agree that comms / human interaction during the race is asking for trouble. On the other hand, map exchanges are some form of human interaction too (although admittedly nowhere near the level we are talking here).

If we don't do it though then I am just going back to my concern that if there aren't packs when they get to the first such control, this is likely to increase the chance that they do form for the middle section. So in fact is breaks them up if they existed before , it creates them if they didn't - net effect: there is more likelihood that trains will exist, but for a shorter duration than before. Not sure if that is an acceptable target.

How on earth are any of the schemes above going to lead to a race that is: (a) fairer, and (b) understandable to the casual spectator? Answer: not at all. It's pretty clear that the bunch of O-geeks on here can't agree on which ideas are sensible, so the chances of getting buy-in from elite runners - let alone the public at large - is zero. Bear in mind that international orienteering can't even agree a progressive and clear cut proposal supported by many athletes (gender parity).

Anyway, the key solutions were all mentioned at the start of the thread:
1. 3 min start interval (at least)
2. ditch the 'red group' system that ensures runners are capable of sticking with the person catching them

To which I would add:
3. don't put epic route choice legs at the start of the course where runners have insufficient knowledge of the terrain and mapping style to make an informed decision

In support of which, here is Daniel Hubmann (via Google Translate):
"The site turned out rougher than expected and that's why I found the extremely long and selective route already for 2. Lots bit unfair. You had to decide on a variation, without being able to adapt previously little-site. Accordingly, the loss of time and the great group formations were enormous."

@Jagge commentary was excellent. As pretty much always. (in finnish by Janne Salmi & co, Janne knows pretty much everything).
Yes. I met Janne at control 20 when he was running the WOC middle course the day before the race last year. Nobody on the organising team knew he had a map. So much for security 8(

@Psuba The effect you describe is same in butterfly/phi loop methods. At common control someone may come from an other forking and join you for the next part of the course. It is not any different here, so difficult to see it as a deal breaker.

@graeme Yes, he made it clear in tv year ago he ran there (I think he said he ran women's course). He did the same now in Strömstad. It was intresting to here how he describes what the terrains is like also in places without cameras. Like now in relay how soft and wet some marshes are after the rain. Gives lots of depth to commentary. I think he often runs it just before the race and enters studio straight from the forest and sometimes a bit late.

Well, a lot of ideas were tested in the last 10 years but how many times we have seen a 3 min start interval?

If 3 min start interval is not acceptable for all, I think this should be at least compulsory for the red group. Elite runners suggested this a long time ago. IOF
officials are able to spend hours to talk about media requirements to make orienteering more visible but in the same time they are not able to speak about fairness and 3 min start interval. What IOF did in the past, any open discussion?

Holger Hott 2005:
3 min start intervals and forking methods, together with high ethical integrity from the runners should be sufficient for keeping the sport as we want it: The lone man against the elements.

Swisscheese 2005:
If you want my opinion, this was a very unfortunate way of putting on a long distance race. There should be 3 min intervals and especially some butterflies.

3 min start interval was used 2011. Together with the the loop spreading pi was after, very short loop and a long one. Bronze medalist ran together with the winner most of the course, from 8th control to the finish. 6 min behind was good enough for bronze. No red groups, there was a qualificaltion race.

Like pi said
"The more challenging the navigation, the more spread in the athletes' performance, the higher the potential for trains forming"

Even with 3 min this was a predictable outcome. They could easily had a larger start interval for the best without losing interest from ordinary TV viewer. So much was happening all the time.

Here's a manual translation of the Hubmann quote paw mentioned, in the hope it is useful:

"The terrain turned out to be rougher than expected, and that's why I thought the extremely long and selective route already to the second control was a bit unfair. You had to choose one variation without being able to adapt a bit to the terrain first. Therefore, much time was lost, and an enormous number of groups appeared."

Note: I guess he uses "route" as a synonym for "leg" here (I am not aware of a simple, elegant German word -- Postenverbindung is lengthy and stilted; the meaning of "selective", however, is lost to me)

If trains tend to form near controls or on short legs, and not as often in the middle of long legs (at least if there are many route choices), then why not set the Long to be almost entirely long legs? Route choice is one of the key characteristics of the format, and there are formats like Middle and Sprint with shorter and medium length legs. The Long would be even more compelling that way, chock full of route choice.

I do like how the IOF definition of a control site is that it's merely the means to end a leg.

robplow: In Sweden it's quite common for professional mappers to see map control as a punishment and not as a way of improving map quality. I simply don't get it at all. To have someone else scrutinizing your map should only be a problem if they disagree completely with you on basic principles, and in that case either the wrong person is mapper or the wrong person is controller.

As for the WOC maps, it has been known for a long time that the Swedish federation map commission would not be involved at all in the process of controlling them and it has nothing to do with "confidentiality". To what extent politics/power plays/personal rivalries has been involved, I don't know. But there is definitely a problem when some of the WOC maps have been made in a way that the mappers must know would not have passed the normal swedish controlling procedure for high-level events.

@graeme in a phi loop, equal forks mean you run the central section together, including all visits to the common controls, so typically it's about half as effective.

No. The other runner runs central section first and then two loops. And the other runs two loops first and then the central section. So having equal or not equal forks makes no difference.

@paw - yes we are all aware that these AP discussions have zero influence on reality ;)

@graeme - OK interesting that the later long legs broke up some trains. Sure didn't look like it on TV where all the potential medal trains were together the whole way.

The conclusion from the train wreck in Japan was that "the first part of the course should not contain many short legs with difficult orienteering". The conclusion from this year's train wreck is that "the first part of the course should not contain a long route choice leg with difficult orienteering". Perhaps we should remove the navigation element completely from our sport? That would sure help a lot!

So clearly, difficult navigation form trains, yet that's what we all want, spectators as well as athletes. I still argue that 3 min start interval, plus an effective forking in the last half of the course (like in 2007, 2011 and 2013) is the most realistic compromise that can be accepted by all. Sure, the 6-minute-man can still ride along the whole course, but hopefully it's very unusual that 6 min behind will give a medal? Otherwise we have to move to 4 min start interval, or forking with 3 loops. The mini-butterfly is proven ineffective and no forking at all would probably require abandoning the red group concept, which probably can't be accepted for live TV broadcast? Yet, 2012 had no forking without becoming a train wreck (for the men at least), is it fair to say this long had unusually easy navigation?

Here is the recent history of the Men's Long at WOC.

2004 - mini-butterfly
2005 - no forking
2006 - mini-butterfly
2007 - short "delay" fork before longer loop
2008 - mini-butterfly
2009 - mini-butterfly
2010 - mini-butterfly
2011 - short "delay" fork before longer loop
2012 - no forking
2013 - short "delay" fork before longer loop
2014 - mini-butterfly
2015 - mini-butterfly
2016 - mini-butterfly

Another issue is that TV sections (typically 3 controls - prewarning, start and finish of TV leg) aren't allowed in the forked sections: so with two TV sections and forking you end up with a lot of short legs.

@pi@paw. At least in part due to what I learnt on AP, there was no looping in the middle last year WOC2015: AP has at least as much influence as elite runners ;) Both winners ran in a group for most of the race. But the relay forking split everything apart rather nicely ;)

is it fair to say this long had unusually easy navigation?
Probably, but it's not just navigation. Rougher terrain leads to greater variance too as people stumble through heather or take different routes around a tree. So also remove terrain running. Or, you know, three minute start interval.

2012 pace was almost one minute per km faster than 2016. So with same amount of seconds behind you were further behind in meters. Same effect as visibility. Fast terrain and low visibility may indicate less trains (hard ground, trails, green). Or at least slow terrain (open forest, soft ground, no trails) with good visibility may indicate more trains.

@pi, not sure that there is a conclusion that hard navigation early (whether long of short legs) is causing trains. I and a few others in this long thread have written many times the opposite: it is a statistical certainty that after 30-40 min running (depending on terrain as ndobbs wrote above) runners will have sufficient variation in their run times (either because of route choice, or mistakes, or the speed with which they tackle the terrain) that they will come close together. What I (and others) have the went on to contemplate is two things: a.) what, if anything is effective after this to break up trains that formed, and b.) can we adjust course design roughly at this point when people are statistically close together but maybe not quite joined up into groups, to avoid the final closing of the gap.

Which is where my comment of introducing long route choice legs roughly at this point (regardless of whether there was a long leg before or not) came in, as that gives a higher chance of passing without seeing each other, or pulling away due to better route choice. This was ALMOST the case in this years course, it just came slightly late, as the open areas before control 4 (and somewhat at the end of 2) allowed the closing of the gap just enough to make the long leg to 6 slightly less effective in avoiding trains as it would otherwise have been.

Graeme's observation on leg 6 breaking up trains belongs to the other argument (how do we break up trains once they have formed).

But, we are getting repetitive now. Just wanted to reiterate that the issue isn't at the start of the course (except for the start interval, which of course IS at the start of the course :) )

@Jagge - the difference is the length of the section where the group can take advantage of running together.

In case of a butterfly/phi loop (at least in how it is implemented currently), the groups that are created due to the randomising effect of time delay is carried through a relatively short section. But even regardless of this, it may be part of the reason of explaining why they aren't very effective in making a difference...

In case of the solution you propose, the section in between the two controls with the extra loops is fairly long (at least how you designed them).
But I take your point, one can play with the distance between those two controls. The question is what is the right balance between that section being long enough to allow significant enough difference to develop between two runners grouped before the first such control to not be back in group after the second (if it is too short then the effect is same as with the current implementation of the butterfly loop), but not long enough to allow much advantage of the artificially formed groups in between.

So having thought it through now, I am not sure that this solution has any advantages over a properly planned, long enough phi/butterfly section. Other than the ability to separate 3 runners rather than 2 (which you could equally do with a modified phi loop with 3 loops).

@Psuba - the conclusion from Japan didn't come from me, but the investigation at the time. The conclusion about this year seems pretty obvious and well described by Daniel Hubmann. Not sure what you are trying to say actually, what do you mean by "opposite"? You are saying the same thing as me that due to runners variation they will start to group significantly around 30 min. This variation will obviously be higher when the navigation is difficult (wherever that is in the course), are you claiming that is not the case? Of course the terrain also affects the variation, that's another factor.

But yes, I agree that it looks like the optimal is to wait 20 minutes or so before the epic long leg, to get a bit more time with independent navigation overall.

@pi, I was contesting your inference that people believe that the problem stems from hard navigation challenges at the beginning of the course. So I agree with your conclusion, I just wasn't convinced that the opposite opinion is a widely accepted stance.

Did anyone see the WUOC Long Distance course?
Cracking effort at forking. Unequal gaffles. Packing was probably starting between the end of leg 3 and about leg 5, and then the butterfly started at 6.
And what was cool was because of the arena map exchange halfway through the butterfly, you didn't even realise you were in a butterfly.
I haven't checked at how effective it was, but I imagine it was quite good.
There's probably also a wider spread in abilities at wuoc then at WOC, which would be a factor as well.

@Psuba - sorry, what does that mean? ;)

@ColmM, reminds me of WOC 2007 forking, just longer and much earlier.
http://www.arua.ch/packfigures/woc-men-langdistanz...

because of ... map exchange ... you didn't even realise you were in a butterfly.
I wonder how often this happens? I used a map exchange (flip) in the WOC relay for exactly this reason, but it doesn't show up on routegadget on online tracking.

One way to stop people following is to have forking, but its often enough that they don't know whether a leg is forked.

One way to actually break up packs is skips, as used in many American "goats" (a format created to encourage Americans to train for WOC length Longs (at the time the only WOC format other than relay, I think), in which competitors can skip any one control except the last (or, say, any one control 1 through 9, any one control 10 through 19,etc., or some such rule, preferably clear and simple). If some orienteers in a train have skipped and others have not, then the train will at some point split. Since everyone gets the same course and has the same options, it's fully fair.

But, I still prefer the simplicity of three minute starts and lots of long legs to reduce formation of trains (as in, nearly all legs long or longish with lots of route choice). Simple is more compelling. Enough long legs, few enough short legs, and a reasonable start interval, and few trains should form.

Re WUOC, I was at the arena (later running in the accompanying public race) and did not see much bunching of people coming through it, although that is not proof and of course only a spot check at one place anyway. And yes, there was much more spreading of abilities, and I think even the startlist was not bunching similar abilities people as much as the red group in WOC does.

@pi: don't worry, it's not that important anyway :)

To put theory in practice, which course setting rules should prevail and would best suits to WOC Long 2017 in general? What elite athletes could suggest to IOF?

Easy or difficult navigation - 2 or 3 minute start interval - Forking, no forking, butterfly -...

Looking at this, I would have to say most important is 3min start intervals need to be brought back.

Difficult navigation is always priority, even if it can increase packing to some extent.

And forking is preferable too, but should be thought out in such a way that it is likely to reduce picking, i.e. The short butterfly at WOC this year clearly didn't work, something would need to be bigger, and preferably unequal lengths to the 2 butterfly sides.

Long legs just need to be more thought through in terms of timing of when they occur etc to reduce following.

No red group.

which course setting rules should prevail and would best suits to WOC Long 2017

I'd say Estonian planner team knows these things well and also what suits best the terrain and circumstances and what should be done and what not. Far better than any of us writing here (well graeme and nixon might be able to give some valueable input). Let the locals show what they can do.

graeme might be able to give some valuable input :)
Yup - my advice is that Estonians will know better what suits Estonian terrain than non-Estonians ;)

It's a shame they won't be able to do exactly what the want though... ;)

In my youth I might have run to my computer and created a simulation of an orienteering course.

Each leg (and in fact various sections of each route choice of each leg) would be given parameters such as:
* visibility
* speed of terrain
* likelihood of being chosen by a runner
* how long runners will stay still (or slow down) to make their route choice


Each simulated runner would be given parameters such as:
* speed in various terrain types
* navigation skill (how accurate/mistake prone - range of likely error)
* navigation skill (how fast in last section of the leg)
* navigation skill (how fast they leave the control)
* boost from being in a train (with like-speed runner, slower runner, faster runner)


And the race will be given parameters - most notably:
* start interval

Then monte carol simulations could be run millions of times to see if this model would in fact predict what has been observed in previous events.

And once the model is sufficiently robust, then it would be made available to all course planners - especially those for WRE events (including World Cups & World Champs) - to allow them to test their course design ideas in advance.

I can imagine many will cringe at such an idea. But I find it very cool and a potential way to anticipate potential problems. And a pretty cool way to try out new course planning ideas.

Maybe it already exists??

I wrote something very similar when I took part in a two-day Project Management course: The speaker used Microsoft's PM software and I realized that it had no form of delay modelling so I spent part of the evening writing code that took the Gant chart data and matched it with an additional file where I specified the expected distributions and ranges for each of the sub-tasks in the project.

At this point I ran 10K to 100K Monte Carlo simulations and accumulated the statistics, which of course showed (among many other things) that projects were far more likely to overrun in time & cost than the opposite. :-)

The code you ask for is of course very similar, you just need significantly more modelling data for each competitor along with the terrain and length of each leg.

I am feeling very sorry for the poor sod teaching MS Project the next day. No, I would have loved to watch the report back. ;-)

@AZ
This? Official Scientific Publication of the International Orienteering Federation IOF
The editor (not me!) wrote


Pack formation in orienteering cannot be eliminated. Because start intervals in international competitions have been shortened in order to increase the spectators’ enjoyment, world class orienteers have got even more used to this phenomenon over the last few years. Yet still, discussions arise after every important competition. And the more important the competition, the bigger the discussion. Ackland
JG shows in his article that one can analyse final results very precisely by including pack formation in calculations, and demonstrates these effects for the long-distance final competition in WOC 2005. Maybe in future it will even be possible to consider this formula in course planning and defining the starting intervals, so as to reduce pack formation?

The editor (not Graeme) asserts "start intervals in international competitions have been shortened in order to increase the spectators’ enjoyment". Really? Is that the official IOF rationale? Is there any evidence?

"increase spectators’ enjoyment" -
434 AP comments prove we're enjoying it :)

-IOF: 'It's not a big it's a feature'

@graeme, it sounds like that is exactly what Adrian is getting at except that he suggests taking it one step further and giving individual parameter values for each leg on a course and therefore using it in a predictive way while course setting (a possibility the editor previously hinted at)

I wonder how much splitting it up leg by leg would affect the accuracy of the model.