Comments on: Separating runners in Orienteering: Part III

By: Steffen

Steffen — Mon, 25 Jan 2010 18:26:12 +0000

One simple possibility to reduce forking would be to change the starting order. If you begin with the winners of the Q-heats and let start the 15th placed you get less grouping. At the actual system it happens automatically that a runner catches earlier started competitors.

By: Simmo

Simmo — Mon, 25 Jan 2010 13:29:13 +0000

Why make what should be a simple race over-complicated just for the sake of tv?

Long Distance should be a classic race without any fancy splitting systems. Just rely on route choice and long legs, together with a good ranking system and longer start interval. And the race should be a true test of stamina (think Czech Republic 1991)

Do away with qualification races. Update the ranking system so that it gives rankings for the different disciplines. For long distance, allow the top 70 ranked runners, plus 10 ‘wild cards’ for runners with recent good form or coming back from injury, with start interval of 3 minutes in reverse rank order.

It won’t be possible for so-called ‘weaker’ runners to be helped by stronger runners because they won’t be near them in the draw. If a runner makes a mistake of 3 minutes its most likely that they cannot be placed even if they manage to stay with the runner who catches them.

For now, let tv have the sprint and middle. If orienteering becomes more popular, tv will want to include long and will find other ways than splitting to make it attractive to the audience.

By: BOM

BOM — Sat, 23 Jan 2010 07:56:37 +0000

As I see it there are just two principally different ways of using short loops as a means of separating runners.
1) The runners do all the loops in one pass (typically butterflies an phi-loops).
2) The runners do one loop at an early stage of the race and the other(s) at later stages.

The advantages of 1:
- easy to implement
- relatively easy for TV and spectators to follow
The disadvantages of 1:
- not very effective (the runners are only forced apart during the loops, after the loops they have ran exactly the same distance and are probably not much separated

The advantages of 2:
- different running times for the two loops forces the runners apart also after the first loop

The disadvantages of 2:
- almost impossible for TV and spectators to follow (you would have to compare two different sets of intermediate times depending and which loops the runners have done)
- severly compromises the course setting as you would have to get back to the forking loops at least one more time

By: Jan Kocbach

Jan Kocbach — Fri, 22 Jan 2010 16:03:21 +0000

But, what about the fairness of running the legs in different time along the race? If there are legs in the phi-loop that are phisiclally or technically much more demanding than others, it’s not the same to face them earlier or later in the race. Therefore, phi-loops must be even along themselves and this reduce the quality of the course. (Big butterflys have the same problem)

That is an extra added unfairness, yes, Luis. If the course is set in a good way, this unfairness should be far less than the unfairness due to the cooperation (or following) between two runners in a large percentage of the course.

By: Chiu

Chiu — Fri, 22 Jan 2010 15:56:57 +0000

Thanks for your nice articles. I learned a lot!
Here is a course I set on Dec 2008 for a local event in Hong Kong:
http://www.hkoc.org/forum/attachments/forumid_9/20081214_58af16fb6eebb4a6a657Hj8BieM9V8WG.jpg
It was a mass start for 16 runners. Everyone’s map is different. I added 2 extra loops at each ends of the phi-loop.

By: Martin

Martin — Fri, 22 Jan 2010 10:39:09 +0000

Where is my mind?
The forked Looping has 8 alternatives.
4 alternatives = every fifth runner has the same combination = 10 minutes start intervall

By: Jan Kocbach

Jan Kocbach — Fri, 22 Jan 2010 10:06:51 +0000

@Martin: Thanks! Better than mine in some respects – especially regarding having two forks. I tried to avoid having too many short legs (as this often increases grouping), and that’s part of the reason for mine being with only one fork. But I’m sure you would have used longer legs if all of the course would have been planned for this :-)

By: Martin

Martin — Fri, 22 Jan 2010 09:00:22 +0000

:) A really ain't a course-setter but ... here

By: Jan Kocbach

Jan Kocbach — Fri, 22 Jan 2010 07:06:58 +0000

:) I’d surely have made my WOC-09 example more like your “forked loop” sample with 8 possibilities if I had some more time for it… I invite you to make a better alternative :-)

By: Martin

Martin — Fri, 22 Jan 2010 00:05:47 +0000

Uuhm. A misunderstanding. I meant the one a little higher above. The one with ‘equal running time for both arms of the loop’. Your WOC 09-Phi looks pretty good.

In your example, the runner taking both wings first can easily identify the control his competitor is cuting to (his 14th Control). Therefore the central leg should maybe not be straight, but contain at least one control.

Even if you lets say manage to create a certain uncertainity, two weak points of the Phi remain:
1) The time the timegap is opened is relatively short to separate two runners, regarding the fact, that both runners are pretty good.
2) The problem is still that you have only two alternatives here, so even with this more elaborated phi Mamleev as Gristwood would just know, that their according mate would run the same way trough the separation.

Uhmm. The forked Looping has 8 alternatives. :P

By: Jan

Jan — Thu, 21 Jan 2010 14:28:47 +0000

What I like on the forked Looping is, that it does not especially try to hide but it just introduces a time gap between two runners keeping it up for some time before resolving it. That way it is highly likely that two runners do not stick together longer than Start-Looping or Looping or Looping-Finish. The classical phi-loop does the same but for a shorter while. The phi-loop Jan introduced above does not even that. :)

I think the phi-loop introduced here introduces a time-gap which is present between control 12 and 19? That was at least the intention. In addition to the fact that you have to pay a little bit of attention to the map reading as you are not 100% sure if the other runner has the same next control.

Regarding phi-loops / partially forked loops, these methods are very similar when going down to the basics. Maybe I should stop calling them phi-loops, but I don’t like calling them forked loops either.

By: Martin

Martin — Thu, 21 Jan 2010 12:33:14 +0000

Correction: Partially forked loops (p.37)

By: Martin

Martin — Thu, 21 Jan 2010 12:30:01 +0000

:) A detail. It seems me and Jan have a disagreement about the WOC 2007 Men Long Final spreading method. I don’t to see it as a phi and the report, Jan talks about, doesn’t file this spreading method under phi but FORKED LOOPS AND INTERVENED LOOPS (p.35) I see it more like a Looping with a forked base… Maybe it can be named a phi too, but somehow one might think about giving these kinds of phi more specific names. :)

Some Points
1) We know that most of the athletes make tactical considerations about the spreading method, while approaching it in contact with an other runner.
2) As far as I know, the spreading method used is always explained before the competition.
3) Especially bad is, when runners know beforehand that they will have the same wing first (e.g. started 4min apart, like Gristwood and Gueorgiou WOC09).
4) Remember, that the case we want to solve is to separate two runners. Even if the separation method is not obvious, the knowlege of its existence, the behaviour of the rival, ‘unexpected’ contact with other rivals plus the information one the map deliver plenty of information, that they are now entering the separation method.

By: Luis

Luis — Thu, 21 Jan 2010 09:22:45 +0000