Discussion: Touching Buildings in Sprint Maps

Extract from my M50 map at Brunel (made by gg presumably)

I spent way too long with my ageing eyesight pondering whether I should go to look for a gap between the two touching buildings at the centre of the extract (not from the control shown). Reason is I guess that you end up with a line between the paved and grass areas that is no thicker than a crossable fence and possibly even thinner. If there was a crossable fence here (presumably the buildings don't actually touch?) then it wouldn't be mapped any different would it? So really there should be a thicker diagonal line at the touch point. That's my feeling anyway and I think they ought to put something in about this when they update ISSOM. I'm pretty sure it is not the first time I have come across this sort of scenario.

P.S. Is there some better forum on some other site to discuss mapping issues like this?

Is there some better forum on some other site

Sacrilege!!!

Seems a fine forum for it.

ISSOM lists a minimum separation between uncrossable buildings of 0.4mm. If there isn't this gap, then it's being indicated as one uncrossable building. (Then, that leaves the question of whether an 0.4mm gap is visible at competition speed. Maybe not reliably by my M50 eyes without magnification or stopping. But it is bigger than the typical ISOM minimum feature separation, and usually maps for the older are enlarged.)

One could conceivably interpret that the minimum building size implies that buildings should be no narrower than a half millimeter, and thus that at the corner, there should either be a minimum 0.5mpm of building symbol (one uncrossable structure) or a gap of 0.4mm (two separate uncrossable structures with a passage between). But the specification is not explicit here. And as noted, it's a common pattern. Well worth clarifying in a future edition. Personally, I think that the intention should be obvious at a glance, especially for such a fast pace format as Sprint. So, I think that either a half millimeter wide join, or four tenths of a millimeter gap, would be the way to map. And the next spec should up the gap to half a millimeter (no more trap nonsense). As the spec says, reduce what you map to make the navigational info visible.

Interestingly this was an exact problem with the map brought up by someone that I know that works at Ocad. He commented that it was incorrect mapping because you should offset the buildings slightly or over-exaggerate one of the buildings so that not gap is visible. It may not necessarily be within the specification, but it makes sense that you should not try to trap people. Apparently that particular area had black and yellow tape because there actually was a way through (which was prohibited to use). I didn't see it though, this was what I heard from other people.

Let's no even go into some of the shocking course planning and controlling on the elite course... Uncut lines going across uncrossable barriers is just not good enough in a WRE.

@JimBaker: and usually maps for the older are enlarged.

I wish.

Its clear to me that you shouldn't expect to be able to get through, there is no 0.4mm gap between buildings. The map could be improved, as Jim indicates, by making sure the grey touches.
More problematic is if you can, in fact, get through and someone does. I just don't see how you can DQ them when there is no ISSOM uncrossible feature on the route (thin black line doesn't count).
In general ISSOM should specify that two uncrossible features must never meet at a point. Unfortunately, this happens all the time with converted vector-data basemaps.

Well, actually, ISSOM does say that two uncrossable buildings should be separated by 0.4mm. Thus, two buildings touching at a corner is against ISSOM. The conclusion therefore is that this is one building feature (no gap big enough for an orienteer). To which ISSOM says: "It is forbidden to pass through or over a building!
Competitors violating this rule will be disqualified.". So a DQ seems technically correct. The problem is that this situation is not dealt with more explicitly, nor are the markings mandated to be less ambiguous, and so one has the unsatisfying situation of needing to stop mid race to consult a rule book and a caliper.

One can either make Sprint a treasure hunt puzzle, in which people stand around scratching their heads until they say "eureka, that's where it is" or "oh, that's what it means", or else one can make it a fast paced race of navigating on the run, snap decisions, extreme efficiency. The two are quite different. It's worthwhile for mappers and organizers to decide which they're targeting, and map and set accordingly. If targeting a fast pace Park World Tour style event, then the passibility of that corner of two buildings is far more important to navigation than the little bits and bobs attached to the buildings (which are fine treasure locations). And thus, make that passibility (or impassibility) clear before any fiddling with little fences, stairwells and other miscellany that's often adjacent to buildings, if the latter are not important to through navigation.

Thanks for the discussion. Good to see. I heard about the tape in the gap too, but I didn't see it or go to look. If there was tape (and therefore a gap to tape) an uncrossable fence thickness of black should presumably be there at the join, or overprint indicating a barred route.

I discovered this exact problem in a map recently (made by graeme!) where I was going to change it to join the buildings together to make sure there was no doubt.

I've made some complex urban ISSOM maps and I try to focus on legibility above absolute accuracy.

I've looked at a bunch of satellite images of the buildings in Brunel, and I can't 100% tell if there's a gap there or not. It looks like there might be. The vector shapes in Google Maps show two separate buildings with a gap.

If there is a gap, then I'd scoot the southern wall of the northern building north until the minimum gap (0.4 mm) was achieved.

If there isn't a gap, then I'd scoot that same wall south until the minimum building size (0.5 mm) was achieved.

This map snippet shows to me that it's extremely well drafted. But, I think the mapper failed to follow the minimum gap rules in ISSOM here, and the course designer abused this part of the map.

---

I'll also mention that a plausible scenario could be something like: there *is* a narrow gap between the buildings, but it's *not* crossable. For example, there's a dumpster there. Or a bike rack. Or an unsafe stack of discarded building supplies or office furniture. In these cases, I'd map the gap, but then add an uncrossable purple line to cover the gap.

Also, if such a gap exists but it mustn't be used, it should be flagged off according to ISSOM.

"... the mapper failed to follow the minimum gap rules in ISSOM here, and the course designer abused this part of the map. "

Agreed, but is this newsworthy?
Sounds like a normal urban Sprint O event, although we probably haven't yet fulfilled the discussion requirement.

Sounds like a normal urban Sprint O event

I wish this weren't normal.

I prefer urban sprints to be a little tricky. But I don't think they should be deceptive. There's a difference.

Well it was the JK sprint and a WRE event. Don't we expect a higher standard than "a normal urban Sprint O event"?

I hold myself to a higher standard when I make sprint maps and design sprint courses. But I'm a perfectionist.

Can vouch, Socks' urban Sprint O events are the shit!

It may be that the map was drawn from aerial photos, and the building overhangs touch (or almost touch) but the actual corners at ground level don't touch. The mapper may have not actually checked, or overlooked it when drawing.

For a WRE I'd be making sure to check minor details like that, particularly if you set the leg shown on the map.

Compared to the little bits of crossable fence and tiny indentations ​into buildings, the contact or not of those two buildings is a very major detail. It's​ far more important to check that than to draw in the tiny indentation for the depicted control.

To me, this looks like a reasonable compromise way of showing two buildings that nearly touch, with no easy passage in between. A suggestion by Pink Socks to show the gap and then add an uncrossable purple line sounds like the absolute worst possibility and I don't understand how someone with so much experience could have made that suggestion. How wide would that gap need to be for the purple line to actually be visible at running speed and interpreted correctly? I'd say 2 mm at least, 3 mm is better, which would lead to massive distortions. Joining the buildings so the grey touches is a possibility, but imagine someone making a mistake, ending up in that place, expecting to see a single building based on the map and seeing two buildings instead.

I've run that sprint, by the way (M40), and I don't recall having any problems with map interpretation at this point. No gap means no passage, period. Just common sense. To those who say the course setter abused that part of the map, keep in mind this isn't a very technical area, so there isn't that much else to use. Why they used that area for a WRE, I have no idea, but for someone like me, a non-competitive guy who needed an excuse to visit London, it worked :)

@andrewd. Why did I know that was going to happen? - gg is a better mapper than me. Hopefully you are talking about the new Dalkeith map which is still in progress, and highlighting how well the ISSOM system of planner and mapper working together works? Aye, that'll be it.

Sure graeme, yes, absolutely.

Using a purple line doesn't need to create any distortion. It can overlap on the buildings as much as is needed to be visible. Make it 10 mm long if you need to.

Apr 27, 2017 6:59 AM# MChub:
To me, this looks like a reasonable compromise way of showing two buildings that nearly touch, with no easy passage in between. Joining the buildings so the grey touches is a possibility, but imagine someone making a mistake, ending up in that place, expecting to see a single building based on the map and seeing two buildings instead.

That's an interesting three state solution:
1) Buildings touch clearly, no gap
2) Buildings barely touch, gap that can't or may not be passed
3). Minimum 0.4mm between buildings, passable gap

But it's not clearly specified in the specification, and it might be hard to discern at a run.

this looks like a reasonable compromise way
Compromise really isn't the way to go. Runners don't care if the buildings "touch clearly" or "barely touch". All they care about is "Can I get through" *

Why they used that area for a WRE...
It doesn't work like that. They used the JK-event for a WRE because it's the biggest race in the UK. Then the JK chooses an area subject to its 2000 person constraints of access, location, and (especially) parking. The WRE has to use the area its given. (which was perfectly fine IMO)

* That would be a quote from the aforementioned mapper, gg.

Passability is so, so, so, so important in urban sprint terrain. So getting it right should be important to mappers (and course designers).

The mapping spec is for minimum gap of 0.4mm and a minimum width of a building of 0.5mm. This location in question is between those two, so it's not true to spec, either one way or the other.

If you were at this race and didn't have a problem here, then goodie for you. I figure that's like a basketball player who plays a game with a hoop that's two inches too high and says, "I still made my shots, so what's the difference?" That doesn't change that the hoop should be 10 feet high.

There's a spec for a reason, and I believe that when it comes to passability, it should be followed. The recent urban sprint at the Flying Pig had several out-of-spec crossable and uncrossable symbols. The courses were fun, well-designed, and well-set, and most people "still made their shots", but some people were at least confused.

With sprint maps in particular, I think that there's a trust between the competitor and the mapper. The competitor needs to know that the mapper got things right. When you're out there running and you see something that's not right, then that trust level goes down, especially when it has to do with passability. "Hmm, if this gap between buildings isn't quite right, how many other buildings can I or can't I get between?" Or, "Hmm, the passable fence symbol is too thick and the impassable walls and fences are too thin... I think I can get through, but am I sure?"

Because I'm both a sprint competitor and a sprint mapper, I'm keenly aware of this trust. When I'm mapping complex situations, I'm constantly asking myself, "Can the competitors trust me here?"

(This isn't the only WRE that's had mapping issues. The 2016 NAOC Sprint had some ISSOM issues that impacted legibility and some things flat out wrong inside of control circles and along ideal routes.)

I suspect that if Sprint maps did a better job of clear indication of passibility, then fewer people would experience the frustration of "traps", which I suspect are often due to people not being able to confidently perceive passability on the map while running.

The gap being discussed here was one UK brick wide - i.e. around 215mm / 8.5inches. Children may have been able to squeeze through but most adults weren't - hence the hazard tape to make sure anyone didn't try.

I would always assume from such a map that there is no gap and as such it is uncrossable. Further, if there was a gap to cross, surely the circle would have been cut to show the corner detail?

surely the circle would have been cut

Hilarious!

Assume, guess, speculate, postulate, theorize, scratch head. The nicest is to simply read map and run. The less of the former, the more of the latter, the better. That should be an objective of mapping, especially for a fast paced format in which a few seconds of uncertainty, or of going the wrong way because of misinterpretation, is so significant. Sprints would be less frustrating if this were emphasized over minor details that have less effect on navigation.

No matter how you map it, someone will be unhappy. I think the way it was mapped was the best possible compromise. Others disagree. Another time it will be the other way around. Standards cannot take all possible situations into account, so mappers and runners will have to use common sense and yes, assume and guess - or only map and run in boring terrains.

There will always be disagreements on how something should be mapped. And for minor details like passable features, this is not so significant. But to accept that something that can easily determine placing like passability is a matter of guesswork and assumption is to make orienteering quite a bit less than a sport. It's pretty central. There are plenty of conundrums that we should give mappers and standards a break about. But here, the solution is simple... Overlap the buildings by half a millimeter based on the minimum width spec (if impassible), or leave a 0.4mm gap (if passible). There doesn't seem to be any dire unsolvable dilemma here. To a few, touching corners suggests some third state of a gap too small to get through, but that doesn't seem like an essential element (nor is it in the spec). No problem creating a map here, or numerous other interesting urban terrains, with clear indication of passibility. And that should take precedence as a top mapping priority. No apologies for calling for that.

There should be no assuming and guessing. As I often reply when someone says "good luck", there should be no luck in orienteering, at least in regards to the things we have control over. The competitor should be able to tell from the map if something like a space between buildings is passable, and it shouldn't be ambiguous.

assume and guess - or only map and run in boring terrains.

I didn't think that interesting terrain and assuming and guessing went hand in hand.

The map extract illustrates another (related) issue - ISSOM specifies minimum dimensions for areas of colour:
"Blue, green, grey or yellow full colour: 0.5mm2
Black dot screen: 0.5mm2
Blue, brown, green or yellow dot screen: 1.0mm2
All features smaller than the dimensions above must be either exaggerated or omitted..."
Symbol 419 Prominent bush or small tree is a green dot 0.75mm diameter, i.e. ~0.45mm2 - so no features should be mapped smaller than this - but clearly many were.
It's not the first time it's been an issue at UK major events, I've raised it with planners/controllers for 2 British Sprint Champs but generally they don't give a shit - not their job to ensure the map meets the spec, apparently

The diameter for a calculated 0.5mm^2 area is 0.8mm ("0.798mm").

Sticking to the specs really would make maps more readable. It sounds like the tendency to map everything, or map as much as possible. It's actually harder to ruthlessly generalize, but yields a more useful map for navigation (and a more pleasant one in my experience). It's a paradox, but difficult to read is not needed or helpful for navigational challenge. Packing the map with non essential details does not improve it for the purpose of an orienteering race. As simple as possible, but no simpler, is a closer metric.

@cedarcreek that's true, but there aren't any 0.8mm diameter ISSOM symbols on the map to use as a gauge....