Discussion: Nav technique

Our last club meet was held in some very technical sand dune terrain covered over with underbrush and forest. There are a number of roads and trails bisecting the area and on some of the legs where a road or trail passed through the leg perpendicular to my route, I chose to run as fast as I could on a compass bearing to the road, then relocate my position on the map. This worked well early in the race, but as I grew more tired, I made some significant parallel errors when relocating on the road.

Is this strategy of "run fast to the road, then relocate" recommendable for skilled orienteers? Do top orienteers ever use this techique?

I don't think you should ever be out of contact with the map, even if running to an obvious collecting feature. The reasons are just what you noted: you spend more time relocating and you leave yourself open to parallel errors.

I can't entirely agree with Eric since much depends on how hard it will be to relocate on the road (and, in Todd's case, apparently, how fatigued he was, assuming the earlier and later legs presented relocation problems of comparable difficulty) . The time savings by skipping doing a lot of slow, detailed navigation can be very big, after all. They just have to be weighed against the risk of making a large error in consequence (also, against the risk that losing map contact on the way to the road will make it impossible for you to efficiently avoid areas of poor runnability or significant climb). In any case, skilled/top orienteers certainly do use this technique in some situations.

i agree with ebuckley that you should always maintain contact with the map, but with linear features (trails, stone walls, streams, etc) i often use the "aim off" technique to minimize time spent trying to pin point my exact location along the l feature.

what it entails is picking a point along the trail (i.e junction, distinctive rock/contour feature, etc) that will be noticeable and will clearly inidicate your location and "aim off" or take a bearing a bit to the left or the right of that feature. this way when you arrive at the trail all you will have to know is whether to run right or left until you hit the feature that will indicate your location. done right you should be able to hit the trail running and relocate (or more appropriately reassure yourself of your location) with minimal lost time. as useful as this technique is, it is only helpful with the proper planning and execution. be sure to choose your feature on the trail wisely (i.e. a boulder in rocky terrain is probalby not the best choice) the more distinctive it is the easier and faster it will be to find. also, you want to continue to watch your map and compass because you don't want to "aim off" too much.

one thing you also might want to consider is ignoring trials that bisect your leg. trails that run perpendicular to the direction that you are running are generally not the best feature to navigate by. they should be noted as a collecting feature of course but odds are good that there are stronger features that will be faster and easier to navigate by.

I would never be tempted to use this technique if a linear feature parallel to my route was available. Interesting enough, I aimed off on the roads early in the race and so arrived knowing which way to turn and go, but after I was tired, I just ran for the road with out planning, on faith that I could relocate once there.

Far be it from me to say there's nothing in the whole universe of tasks for which faith-based methods work consistently, but the set of such tasks would, even if not null, definitely not include orienteering in very technical sand dune terrain.

There are no atheists in sand dunes.

Yer missin' his point there, pardner (and I can't blame you, with all that fancy talk he used). Prayin' ain't going to get you through the sand.

Perhaps I should clarify my position. Running a rough bearing and keeping track of how far you've gone (usually by pace counting) towards a collecting feature is NOT the same thing as losing contact with the map. Such care, potentially combined with aiming off is a perfectly sound strategy. What I am recommending against is the "run until you hit the road" strategy. In that case, if things go wrong, you're really in trouble because you've got no way to reliably get back in contact with the map.

Perhaps I should have whipped out my ;-) emoticon after that atheist joke.

I do get the point. My orienteering is completely pragmatic and the fact that I casted off across the dunes for a linear collection feature of a road was merely a calculated risk based on the fact that I can usually relocate just about anywhere I end up. I was trading security and accuracy for speed, nothing more.

I am leaning towards the conclusion that I should just stay in contact with the map all the time. That's what I enjoy about orienteering anyhow. Careening through the woods on a rough bearing aimed off to a large collecting feature may get me to the next control faster sometimes, but I am inclined to reduce the risk in my orienteering technique rather increase it. And I learned last weekend that once fatigued, the risks of "running for the road" and sacrificing contact with the map, go way, way up.

I have a navigational concept that I call "The Zone of Uncertainty". The basic idea is that I am always in contact with the map, but that this contact is not always in the form of a point feature. In other words I can't always say "I am exactly here", instead I think of my position on the map more as an area feature (ie "I am exactly somewhere inside this area"). This area is what I call my Zone of Uncertainty - I'm sure I'm in there, but not certain exactly where in there.

As I run a race I constantly vary and control the size of the ZofU according to the challenges of the particular section of the course. For example, by using pace counting and careful compass work I can keep the zone very small. If I'm running toward a highly visible feature I may allow the zone to get quite big as I will be able to spot the feature from far away and then reduce the size of the zone. And from time to time I will need to collapse the zone to a specific point, for example if I am using an attack point on a leg, then I will need to know exactly where I am before attacking the control. And of course, the zone is collapsed automatically when I reach a control.

I believe (though haven't tested the belief) that most experienced orienteers navigate this way. Some, like ebuckley perhaps, will keep their zone very small. It helps no doubt the ebuckely is probably an excellent map reader, because that is the very best way to keep a small ZofU. Though mechanical techniques (pace counting & running on a compass) are also valuable techniques, especially for those with less map reading skills.

I have seen once a small article written perhaps by Thierri G. (sp?) that kind of talks in a similar way. He is always thinking ahead about how visible will features be, and then adjusts his orienteering to get within the "Zone of Visibility". He doesn't care exactly where he is, so long as he can be sure to get inside the Zone.

So to answer the original question I would say that you are letting your Zone of Uncertainty get way too big late in the race. You need to understand why it is getting so big and then fix that problem. The technique you are using isn't bad in itself.

AZ: very interesting way of putting it.

When I trained with Peter Oberg in Sweden (national team member) I reckon he definitely used something like this. i always used to ask him how come he was so fast in the terrain or we would talk over our courses and he would often state that i didnt need to know exactly where i was all the time. he used to say he didnt know where he was but i dont think that is completely true because the speed at which he picked things up "relocating" was incredible. he also used to use strong direction (sometimes compass sometimes map) and just go like the clappers for a while mid-leg.

AZ: That's very similar to how I've heard people describe Kalman Filters. {And having scanned that link, I can say it makes a lot more sense when you don't have all that math in the way. My head just about exploded.}

It's a lot about anticipication: to put it in another way, you dont have to know exactly where you are all the time as long as you have a clear picture of where you are going and what to expect/look for in the terrain.

ebuckely is probably an excellent map reader

Hee, hee, nice to end the week on a good joke! Seriously, I would insert the phrase "tries to" in front of most of those assertions regarding technique.

One thing I've been working on lately is identifying and using subtle handrails. Even in very complex terrain, there are often things that can be viewed as linear features (a series of knolls, two point marshes close enough together that you can see both, etc.). Basically, what I'm trying to do is apply Orange-level techniques to advanced feature recognition. This is not entirely my idea, it's based off an article I read that describes Tero's method (which was called something like "shortest full-speed route" - I lost the link, but it's worth searching for).

In the nordic countries, where I've done most of my orienteering, you have to "filter" most of the stuff on the map because otherwise reading the map "all the time" would slow you down too much. This also depends on the accuracy of the map. Nowadays in Finland most of the maps are so detailed that you really have to know how to find the most important things and ignore the rest. But this doesn't mean that you don't know where you are, you have to know where you are and aspecially where you are going all the time.

The good old traffic light technique describes it pretty well: in the green zone you run as fast as you can relying on the big features (like hills and paths), in the yellow zone you slow down a bit and make sure you know exactly where you are and in the red zone you orienteer from point to point to the control.

Thierry Gueorgiou describes his route choice technique for the distillation of complex terrain down to the most visible and distinctive features.

Full speed - no mistakes

Golden Route

Great thread, really interesting. Got me all inspired about orienteering again and now I can't help but throw in my 2 cents.

Firstly, (and I'm sorry to state the obvious) the road is not your objective. The road is only useful if it can help you find your attack point faster than the alternatives. If you treat it like Gueorgiou might, its just another feature and you'll flow right through it (not "to" it) or you may even ignore it! But if you merely run to the road with a mind to relocate once you get there, I think you're probably adding a navigational process into your leg which is unnecessary and slower. IMO its better not to think of reaching the road - instead, think beyond the road - think about the leg.

More stating the obvious: each leg is different and is to be treated on its merits. The run/relocate technique may well be appropriate in some situations, but blanket application would be ill advised IMO ...for what its worth...

If you *are* going to use the road as a navigational feature... with all due respect, I don't see any speed-gain in pace counting when you know you're gonna cross the road midway through the leg. In this situation I'd prefer to focus on lateral postion, rather than distance travelled. Pace counting can pay off when you need to know distance travelled and there is no other reliable indicator.

thanks again for all the links and your interesting thoughts - really like that Zone of Uncertainty!

with all due respect, I don't see any speed-gain in pace counting when you know you're gonna cross the road midway through the leg.

Sure, there's no speed gain. However, there also isn't any speed loss, and that's what we're trying to minimize, and if you screw up, knowing that you went a certain distance in a vague direction will help you minimize time loss while you're relocating.

Well... perhaps I should apologise. I didn't intend to turn this interesting thread into another boring argument over the merits or otherwise of pace-counting. My bad, people - someone threw the bait and I took it...

Suffice to say we've each got our own toolkit and choose from it the technique(s) we think will get us to our destination the fastest... race ya!

Don't apologize Nick. This is interesting stuff. Have at it!

On the merits of pace-counting (sorry guys)
If you're gonna cross a road midway through a leg that's perfect for pace-counting: just count to the road and then from it.

Cheers toddp.

Still got your map? Have you reviewed those legs where you used the run/relocate approach, and looked for alternative approaches?

I'd be interested to hear your conclusions; have you developed a new policy to apply next time you're faced with a similar leg? Or would you use the same approach again?

Why not post the map for all to see. We are still discussing an imaginary leg for tactics and approaches that are highly specific to the particular situation.

My conclusion is that I gained a little time early in the race by running fast to the road on a bearing exclusively and then relocating. However, I lost quite a bit of time later in the race by executing that strategy poorly because I relocated incorrectly on the roads with parallel errors. I think the the strategy may well work in some situations, especially super complex and repetitive terrain, but even in the difficult sand dunes I was navigating, the terrain still offered enough features to stay on the map. I think next time I will stay on the map MORE and run rough bearings to catching features LESS.

I think that is a sound policy. (But, are you sure you weren't tempted by anything in Thierry's "zone of visibility" concept?) Good luck out there!

I consider those articles the best I have ever read on advanced orienteering. I used the his "zone of visibility" techniques to good affect during the sand dunes race. I am excited to use them more in the future.

phew - same!

I am thinking about a simulation exercise to train this "zone of visibility" concept. Take a map with interesting features (micro contours for example), put a circle on a feature of the map and in the vicinity put a point or an arrow pointing to the target feature. The idea is to answer "yes, the feature is visible from the point" or "no it is not" with a punch card. Put 50 pairs of those on the map and go for a jog. A very good mental exercise.

Some more about the Zone of Uncertainty model of orienteering...

First, note that the ZofU is just a model - an attempt to explain how (at least some good) orienteers actually navigate, and to help identify ways to improve your own methods.

The ZofU tends to grow in length and in width and so you need to control its size along those two dimensions.

Handrails are the very best way to control the width of the ZofU. The better you are at recognizing handrails the narrower will be your ZofU. ebuckley talks about developing his ability to recognize & follow some subtle handrails - an excellent way to improve navigation.

Compass bearings are another good way to control the width of the ZofU. Thierry G is, I've heard, superb at running on a compass. toddp on the other hand is not so good ;-) My hunch is that late in a race toddp is getting fatigued and his concentration falters so he is not able to use his compass to keep a narrow ZofU. His ZofU gets wide and then he suffers from the common symptom of a wide ZofU - the parallel error!

(As an aside, I think that toddp & me are in the same boat - we need to either improve our fitness so our concentration lasts longer, or improve our concentration skills so that we can remain focused when fatigued. I doubt we're the only two in that boat.)

My goal is to control the size of my ZofU so that I am certain it is contained completely within a feature's Zone of Visbility - then I am able to say that although I don't really know where I am, I'm sure I will be able to see the feature.

My goal is to control the size of my ZofU so that I am certain it is contained completely within a feature's Zone of Visbility - then I am able to say that although I don't really know where I am, I'm sure I will be able to see the feature.

This sounds like a good strategy, but with one caveat: you need to be sure you have a way to uniquely identify the feature once you spot it. Otherwise, you may be able to see the correct feature, but still get suckered by a parallel feature.

I think the key here is simply to recognize that the weaker your map contact, the more distinct/visible your collecting feature needs to be. In particular, using a linear feature to relocate is fraught with difficulty unless there is something else to tell you where you are along that line (in which case, it's really a point feature, just with a nice handrail into it).

This is why the "run to the road" strategy is so dangerous. Sure, you can find the road easy enough, but you still don't know EXACTLY where you are. Attacking a technical control from a hazy location is begging for disaster.

On a related note, I personally have found that I do much better in complex terrain if I make a point of going ALL THE WAY to my attackpoint as opposed to simply checking it off, even if it's a ways off to one side or the other. I believe there are two reasons for this: 1) being AT the attackpoint is more precise than "about 30m roughly SE" of the attackpoint, 2) I can then follow in the approach that I've already planned out rather than adjust for the fact that I'm left or right of my intended line.

I have found the article I referred to earlier that describes the "Zone of Visibility" concept used by Gueorigou. It is written by Martin Lerjen and is called Analyze Your Map Reading in Orienteering.

And to further try to advance the "ZofU" lingo I will restate ebuckley's comment about why the "run to the road" strategy is so dangerous, using ZofU terminology ;-)...

The "run to the road" strategy is dangerous because reaching the road only makes your Zone of Uncertainty short - but you have not necessarily controlled its width. If you have allowed your ZofU to get very wide you may not have adequate map contact (i.e. your ZofU may be too big) to accurately and safely continue on your route.

Two strategies for controlling the width of the ZofU when running at a linear catching feature are using careful compass, and/or aiming off so that when you hit the linear feature your ZofU is entirely to one side of a given point on that linear feature. Another way of course is map reading to keep the ZofU small all the way to the catching feature. And my favorite is to find an intersecting linear feature that I can follow until they meet - at which point my ZofU is 'collapsed' to a single specific point.

Anyway, this probably sounds a bit silly. I just want to say again that I'm trying to validate (at least in my own mind) a new (?) model that better describes how orienteers navigate. If this ZofU model is 'good' then it can give us insights and help us become better orienteers.

Yeah, I like this idea of ZoV/ZoU area, and how you should deal differently with restraining the width of the ZoU (with compass or linear feature) as well as its length (by map reading and selection of visible elements). Way to go!

I think the terminology is useful. I think most competent navigators already have some sense of this, but it helps to be able to articulate it, both internally and externally.

I also think that the far greater danger is in letting the ZofU get too wide. For one thing, it's a pretty rare leg that doesn't cross any linear features, so unless you really have your head turned off, you should be able to "reset" the depth to zero a few times during the leg. Secondly, because your forward vision is typically the furthest, you can afford a greater uncertainty in depth because you will still see the features in the distance; they just might be a bit closer or further than you expect.

So, despite these two useful concepts I don't feel as if I can relax my attention to the CARE system.

Instead they seem to work as underlying policies; ZoU - relating to the actual/desirable degree of map contact one has at any given time; ZoV relating to selection of features at any level of CARE, filtering out all but the most obvious features (ie: high end simplification)

Or, am I missing something?

CARE system?

Control
Attackpoint
Route
Exit