Register | Login
Attackpoint - performance and training tools for orienteering athletes

Discussion: I enjoyed reading your comm...

in: CleverSky; CleverSky > 2017-07-19

Jul 19, 2017 11:19 PM # 
I enjoyed reading your comments, very much so.

I've only been to the Manitoba sand hills once, 1982, Canadian Champs one weekend, NA Champs the next, a couple of low-key events during the week between. I loved it. Though I wasn't sure what percent of the field liked it, as it's not easy orienteering. And when you're lost, it can be a big problem.

We were in an area without the bigger dunes you had here. I'm not sure what differences there might have been in the vegetation, but I suspect it was not so different. If anything, the bigger dunes should have made the navigating easier?

As far as the mapping, back then the maps were obviously much less detailed. 1:15,000 and I don't remember any difficulty reading them on the run.

Here's what I wrote about the second day of the COC some years after the fact. Clicking on the title gets to the map.

So with you and Charlie going, and my fond memories from the area, I followed along over the weekend to see how things went. And I was left with two conclusions.

1. Once again, a map that is way too busy, way to hard to read.

I even went and looked up the current IOF map specs. A bunch of times it talks about how important it is that the map be legible, such as:

"The mapper must always take into consideration the special conditions for orienteering map reading. Firstly, running makes reading a map more difficult. Secondly, orienteering often takes place in forests, and in all kinds of weather. The light in forests with dense canopies is dimmed even in the middle of the day, and there are numerous other factors that impact map reading, such as rain, dirt and damages to the map or plastic bag caused by rough handling. Therefore, it is obvious that legibility is of utmost importance for orienteering maps. Minimum graphical dimensions must be respected and unnecessary detail must be avoided."

To just choose a single example, think how much better the map would be if you removed every single form line. Every single one. I doubt you'd have a single complaint. The course setter might have a couple of locations that he/she might want to change -- using more distinct control features -- but that too would probably be an improvement.

Of course, then you could also go after the other aspects of this "fill up the space with details" mentality, which could also be called "map every bush." Forget the tiny patches of green, the little clearings in the green. Respect the IOF minimum dimensions.

Think of the amount of work that went into this, and how much better the map could have been with less work.

I think one of the reasons it happens is that everything is done on a computer, and it's so easy to blow up a map segment to fit in more details, and that's considered cool.

But it's really a shame.

2. I'm still somewhat baffled by comments from you and Charlie that it was way too hard. You guys are both really good and really experienced. Realize that I wasn't there and therefore it may have been quite different from the way it looks on my computer.

My experience in such negative terrain is that three skills are essential. The first is using handrails. And the map shows all sorts of handrails out there. And certainly when you're in the open areas, they should be very visible.

The second is using a compass very very accurately. Particularly in two cases -- when you are following a handrail, to make sure it is going the right direction (and therefore you are on the handrail you think you are); and also when going through areas of limited visibility to make sure you go exactly the intended direction. I see a lot of cases where I can't imagine a compass was being used with much care.

The third is skill in reading contours, particularly in understanding the particular characteristics of features. Knolls, for example, are not just high points. Each has a certain size and shape and orientation in the terrain. One should be able to not just check off the various features you are passing, but also be absolutely sure you have identified the correct feature.

With regard to the third item, I looked for reports on the event on AP to see how folks were managing and found the following -- "Everything really does look the same. Most every depression and hilltop is the same shape and size of flattish kidney-looking thing, and there are quite literally zero stark features, whether linear, round, or point. No trails, no cliffs, no ponds, no creeks, no boulders, nothing man-made. And few reentrants or spurs in the usual sense, i.e. larger than one or two 2.5m contours with maybe a form line thrown in just for fun. "

To which all I can say is Nope. Things are different, you just have to look.

Again, I was just really surprised at the struggles of the two of you. And Charlie wasn't even on M35... :-)

Thanks again for the write-up. If I was still orienteering and they offered another event in the sand hills, I wouldn't go back unless I had my eyes (and legs) of 30 or 40 years ago, and certainly unless the map was really legible.

And that means legible without using a magnifier, a crutch that seems to have become common at the elite level in recent years. I never thought that was, or should be, any part of orienteering for anyone of reasonably normal vision.
Jul 20, 2017 12:47 AM # 
Some agreerment, and some clarification...

I haven't been too critical of the map publicly, because Rob seems like a nice guy, and I haven't yet decided how to do so tactfully. I think the computer does indeed come into play in terms of filling the map with unnecessary detail, but not just because it makes it easy to draw. Have you looked at the basemap that was posted? I'm not positive, but I think the contour interval is 0.5 m [edit: I reread the notes, and the interval for the smallest contours was supposedly .0625 m, but I'm sure he meant 0.625 m]. That means that the mapper already has all this superfluous detail at his fingertips. I imagine it's very tempting to just throw on a form line when the shape is already there for you -- in fact, you have your choice of several! It probably takes a lot of restraint to not include a lot of that, especially when there's a piece of the map that would otherwise be blank...

But instead the result was a map that was very hard to read. That's what made it so hard. I had a magnifier, but I don't have a lot of experience using it, and I have trouble getting my eyes to converge when I use it (my own problem). There were a couple of other things I forgot to mention, that are just things to deal with, but they did complicate matters. One was that it was really bright and sunny. That's normally good, but if the map was in the sun, there was too much glare off of the map case, and if it was in the shadow of my head, then it was too dim for my tiny pupils. And a problem just for me was colorblindness. In this case, it manifested itself as difficulty distinguishing between yellow and light green. I've been on other maps where the problem was worse, but I definitely had issues where I'd be in a big open area, and some of the reentrants were full of trees, but I couldn't tell from the map which ones they were. Or I'd want to go somewhere that looked okay to me on the map, nice and open, but when I looked up all I'd see would be trees. That's also my own problem. But the overcomplicated map was the big factor that was vexing me. I came into this prepared for what I remembered from four years ago, and if the map had looked like that one, I think things would have worked out a lot better.

I learned about the importance of compass use in sand terrain from you, at a course review in 1990 in Minnesota, and most of the time I was doing just what you describe. From #3 to #10 on the Long course I had my act together, being very careful with the compass and making my own handrails through the terrain by stringing features together. In fact, you can see two places on leg #9 where I caught myself after having contoured around to a point where my heading was off, and stopping to correct. A couple of times I got in trouble (#1 and #7 on the Middle, for example, and #6 on the Sprint) when my heading was off by a fairly subtle amount in a vague area (most of M#1 and S#6 was through vegetation that was just sort of rough-semi-thick). I dealt with it okay on M#7, but not on the other two. S#5, yeah, I really didn't use the compass well, but the first 2/3 of that route really was my plan, I just screwed up the passage through the green, which I can now see was due to one of my green/yellow mixups.

Importantly, though, is the comments about everything looking the same. I totally agree with that comment, and yet I don't disagree with you. It's all a matter of context. When you're going from feature to feature, keeping track of where you are, the features are all unique and identifiable. If you're even reasonably close, you can conceivably figure it out. But imagine being dropped in the middle of that map, taking off your blindfold, and trying to figure out where you are. A lot harder in that circumstance. And due to human failings, a lot of people made small errors that blew up, because their bearings weren't exactly right, and the features they were checking off looked like they matched up, and then that stopped being the case, and suddenly they've just taken off the blindfold. At that point they might be a few hundred meters from where they thought they were, and on this map, that's a really large search space. It's not that the features are all the same shape, but there are a lot of subtle variations. And it's not exactly that you can look at a feature and have an objective notion of what it should look like. In a flattish area, you might see something that could be a sort of triangular depression, or could be the area between three knolls, depending on exactly where the contour level happens to land. It's obvious when you're maintaining contact, but less so if you're haplessly trying to relocate.

But (and this is meant with a smile, not a sneer), you may have less experience with this (relocation) than most orienteers. I've known you long enough to know that you are one of the folks I know whose brain works differently than other people with regard to navigation, partly due to training and experience and partly due to something innate. As I noted in the writeup, many people in this world would not enjoy, or maybe evenbe capable of, ordinary orienteering. Many orienteers whom I know would not have been able to handle this past weekend. I'm a second-tier navigator, and am barely competent enough for this stuff. But as I was thinking while I was out there, PG would just eat this stuff up.

I should also mention hat I've scoped out some sandhill terrain in eastern Colorado that is completely free of trees, and the poison ivy (where it is) is only about an inch high. I think that stuff would be a blast.
Jul 20, 2017 1:11 AM # 
but I think the contour interval is 0.5 m
That is normal practice for a lidar basemap, but it does require careful consideration from the mapper for use.

When I do my basemaps, I run a set of 0.5m contours with very little smoothing. They show small gullies and benches very well and I use them as reference for position. I also run a set of contours at the final contour interval (2.0, 2.5 or 5.0) with a very heavy level of smoothing. I want something that is close to what I would draw, and shows clearly the general shape of the land. The important step in this though is that I redraw every single contour by hand with input from the fieldwork, the 0.5m contours for fine features I may want to exaggerate, and the general shape from the 2.0m set. I have little respect for maps that have auto generated contours as their final data layer. I've put a huge amount of development time into my lidar surfacing and contouring software, but its still nothing like a human with an eye towards what a runner really wants.
Jul 20, 2017 1:52 AM # 
Well, I said something on the event thread.

I agree, having the extra contours there is fine, provided you have a way to limit your use of them, which is a happy side effect of redrawing all the contours that you're going to use. Rob may have actually done that, I have no idea and I'm not going to analyze the base to figure it out, but if he did then he still succumbed to the temptation to fill up all of the empty space. In the days when the basemap interval was the same as the final map interval, the mapper (or at least the photogrammetrist) had to notice that there was something important that was worth showing. Sometimes it would be shown by a form line, sometimes by tweaking a regular contour.

So I lied, I did just compare a bit of map to a bit of base. There's a long flat spur I'm looking at that has four bumps shown on the base with these 0.625 m contours, and three of them show up on the map. Nearby, there's an irregular shaped piece of high ground connected to the adjacent hillside by a narrow isthmus, surrounded by three of these very shallow depressions. He apparently did redraw the contours (here, at least), because that fat spur has been pinched off so that it's a knoll. And all three form-line depressions show up. Plus he added a U depression. Ack.
Jul 20, 2017 2:43 AM # 
On other footnote: the map was ISOM-spec, with one special symbol, a green line indicating "Fallen tree (difficult to cross)". Completely bizarre. I have no idea if there actually are any on the map, but nobody I spoke to seemed to have found any. Plenty in the terrain, of course, but it's hard to imagine how a mapper would choose a few from the tens of thousands that are out there, and decide they needed to be shown. A fallen tree out in the open is no big deal to go around (I guess it would be an additional navigational feature, not that the map is lacking in those, but the description implies that it's a passability warning). A fallen tree in the white or green is... nothing special. A fallen tree blocking a narrow yellow passage might be useful, but it would be indistinguishable from a narrow bit of green, either on the map or as a practical matter. But maybe that would be so narrow as to be a standards violation, and declaring it to be a special symbol is a cheesy way around that.
Jul 20, 2017 2:49 PM # 
Looks like I got a response to my comment, which in turn elicited a more detailed reply from me. From my non-objective perspective, his response reads like, "I'm glad to hear from you, but I don't like your tone, and this is the correct way to map and it's what people want, and you're stuck in the past. And the mapping standards should be changed so that we can go even more in this direction". Not too surprising, and maybe it sounds different to other readers.
Jul 20, 2017 3:00 PM # 
Sounded the same to me. I'm sure he (and others) feel that they are making more accurate maps, and they are. The fact that they are less useful doesn't register.
Jul 20, 2017 7:14 PM # 
From my (more objective?) perspective, I didn't read his response as dogmatic, saying "my way is correct, yours is not", though of course he has a strong opinion on the matter (and does claim your opinion is in the minority). The crux of the issue is when he says: "I feel that there are a lot of very significant features that did not get mapped on Hog's Back that simply needed to be on the map to make it fair and consistent." I have no way to evaluate that claim. And I wonder if a mapper walking through the terrain might think some feature is needed that an orienteer running at speed would not. But, if the claim is true, it would go some way to counter, what is just obvious, that the old Hogs Back map is much more readable.
Jul 20, 2017 9:37 PM # 
I was able to do sub-12 min/km and sub-15 min/km on Hog's Back four years ago, as opposed to 20+ for all three races this time. Granted, I was four years younger and not dealing with a knee problem back then, but I think the map was also a factor in that. Part of it may have been because the simpler map didn't make it so easy to hide the controls, and part because it was possible to read the map without standing still, despite the fact that the Long map was 1:15000. I didn't think there was anything unfailr about it.
Jul 20, 2017 9:53 PM # 
Agree with your position on this. These maps all seem way too brown - too many contours, way too many form lines and when did tag lines for every down slope become the fashion?

Must be disappointing for a major event to be so poorly supported, but maybe the others knew the nature of the maps.

As for the PI, was it throughout the area - fields as well as forest and green?
Jul 20, 2017 10:41 PM # 
Yeah, poison ivy was really everywhere. Definitely everywhere in the woods, except for a few deep dark places where there maybe wasn't enough light for it (and you really didn't want to go there). In the fields, there were a few spots where I noticed that there wasn't any, but they were rare.

The slope tags were probably necessary, but would have been less so if the map were not so overcrowded with details.
Jul 21, 2017 12:55 AM # 
Just was able to get the comparison map view. The older method would be much preferred. Too much brown on this "new and modern preferred " mode of map. The tags help but should not really be needed.
Jul 21, 2017 8:58 PM # 
Tags are needed if there aren't any water features to give you a clue of up and down. And sometimes, even if there are.

I've orienteered on a southern Ontario map that was rolling like the sand dune terrain, and had a heck of a time there because depressions/downs didn't have tag lines. (This was some time ago; I don't remember which map but think it was in the Hamilton area.)
Jul 22, 2017 12:42 AM # 
Tags should be used into depressions not off knolls or hills. With all the contours and form lines, it is really hard to discern up and down as this was mapped.
Jul 22, 2017 1:40 AM # 
Partially agree. I think there are circumstances where it makes sense to put a slope tag on a convex-down feature (spur or, more rarely, knoll) as opposed to a concave-down feature (depression, or ocassionally, reentrant), but they have to be used judiciously. You're quite right about the up versus down, though. A frustrating recurring thing for me was looking at a long, high ridge, and having a lot of trouble finding the top of the ridge on the map. Eventually I realized that it was (usually) going to be a couple of lines before it turned dark green, but that's an odd, indirect way to read contours. I don't recall having that problem on Hog's Back, even though it has similar dune ridges.

Depressions are always supposed to have tags. One argument I've had with a noted mapper whose opinion I otherwise respect has to do with those tags on a multi-line depression. I maintain that every closed loop that's downhill on the inside has to have tags. He says that if you have a deep depression, only the contour in the middle/bottom has to have the tags. As far as I'm concerned, that's flat-out wrong, because that's how you map a hill with a dent in the top, like a volcano. This came up because of a map of his that we were discussing that's in karst terrain, where there are no hills, just numerous huge holes. I can understand how the map is workable in that special context, but I still say it's wrong. (And this paragraph has nothing to do with the discussion at hand.)

This discussion thread is closed.