Register | username: pw: 
Attackpoint - performance and training tools for orienteering athletes

Discussion: Developing a Map analyisis tool

in: Orienteering; General;

Sep 2, 2016 7:16 PM # 
Hey All,
I am trying to develop a process of map analysis to gauge the technical and physical challenge of a map and put a number to it between 0-5. I am looking for feedback in two specific areas.

1. What makes a map technical or physically challenging to you ie soft ground lots of point features close together, many detailed contour features, lots of climb etc.

2. Does anyone know a way to quickly measure if a map has say 75% green on it or 25% smooth contour lines vs squiggly lines or little hills, and percentage of contours lines close together vs spread etc? Does anyone want to make this your project to develop an analytical tool to measure these things accurately?

My ultimate goal is to be able to gauge an athletes performance based on the challenge of the map and their min per K time to figure out how to help them get better, learn to apply the right technique to the terrain situation and to set a standard for the level of ability it takes to compete at a high level in competitions like JWOC. Your specific congruent feedback is greatly appreciated: )!!

Sep 2, 2016 8:10 PM # 
An interesting initiative. Maybe Europeans or Oceanans have tried this. The description in the Swedish course design guidelines of various levels of difficulty are useful, though a bit tied to Swedish terrain. Note that detailed terrain can be technical (think Pawruckaway in NH or Saskatchewan sandhills or Flock Hill west of Christchurch), but so can low detail (think Round Mountain in CO).
Sep 2, 2016 9:04 PM # 
I think the Swedish scale is about the technical difficulty of the leg, rather than the map. It is possible, obviously, to make easy legs on a difficult map (and in fact as Controller this is probably one of the most common bits of feedback I give to course planners).

My gut feeling is that the objective of scaling each map is maybe not the right way to go?(Of course, as usual, nobody asked me)
Sep 2, 2016 10:02 PM # 
Yes, measuring the technical difficulty of a course, rather than a map, is more ideal. I mentioned the Swedish course standards because I think that the description of several levels of difficulty is one of the best I've seen.
Sep 2, 2016 11:55 PM # 
Does anyone know a way to quickly measure if a map has say 75% green on it

Assuming you've got OCAD:

* Draw an area symbol around the boundary of the map area, select it, use the measure tool to get the area.
* Select all the 'green' symbols (hold down Ctrl to select multiple).
* Right-click the symbol list, Select by symbol..., All objects with selected symbol.
* Measure tool to get the cumulative area of everything selected.

Assumes the mapper hasn't overlapped green areas, but should give you something to work with.

Point-feature density is easy - count point features, divide by area. Ignore rocky/broken ground.

or 25% smooth contour lines vs squiggly lines or little hills

Define 'squiggly' mathematically. :)

If you had lidar for the area you could probably work out steepness/average gradient...? I want to suggest average length of contour line symbols, but I'm sure there's a mapper out there who draws all their contour lines in 20m sections.
Sep 3, 2016 11:43 AM # 
I endorse the negative, but well intended comments above.
Interesting, maybe for some, but I think ultimately unproductive.

I certainly agree course evaluation trumps map evaluation, but even quantifying courses is very imprecise.

In either case, one of the most important determinants of technical difficulty is the visibility of the white forest. How do you quantify that(?), especially when it varies within a terrain, and also seasonally in many parts of the world.

Related to visibility, is the skill level of the orienteer. For a developing orienteer, complexity is highly correlated to tech difficulty, but for someone with mature map reading skills, visibility makes a huge difference, regardless of complexity.

For physical evaluation, sure some tools will count climb, but what about the footing which is similarly relevant. I think rocky footing has a very low correlation to mapped rocks, and how does one quantify the softness of the footing or height of the bracken and heather which play a major role in many terrains, especially for non-locals. The recent WOC terrain is a prime example of a Swedish/Norwegian region with an especially "heavy" reputation, that is difficult for firm-footers to adjust to.

For tech evaluation, I think subjectively assigning a 0-5 number based on experienced eyes in the terrain, or even worst case, off site, will be at least as useful as any sophisticated number crunching.
Sep 3, 2016 4:58 PM # 
Are the ranking system gnarliness factors useful for your purpose? Of course, they only exist for US ranked races after the fact. Not for local or international events.
Sep 4, 2016 1:47 AM # 
Related to Eric W's comment about visibility in white (or the various green shades), these can vary tremendously by season as well. Small features visible from 100m in winter or early spring might be visible from only 10 or 20 in summer, and the competitor's speed of movement through said area may also be substantially reduced.

While some maps are inherently more challenging than others, seasonal variation and the course setter's skill greatly influences the "gnarliness" of a particular course. I understand what Erin would like to do, but the number of variables beyond the map itself seems to make it pretty much impossible to quantify the map itself in isolation from other factors. Granted, you could say "with the assumption of optimal season for visibility, and expert courses set by a highly competent course setter". But that kind of defeats the purpose of all-season evaluation of athletes on courses from various sources.
Sep 4, 2016 3:08 AM # 
Can any climbers jump in here with background on how their scale and how it is applied?
Sep 4, 2016 7:22 AM # 
Climbing grading is entirely subjective. Climbers wouldn't have it any other way. Where would the sport be without endless camp fire disputes about how hard or easy a climb is/was. Where would the sport be without sand bagging- purposefully under grading a climb as (to quote a celebrated Australian climber) a hedge against inflation. The first ascensionist has the right to name the climb grade, and subsequent repeat ascenders either accept or dispute. Generally a consensus is reached. But not always. What climbing grading has in common with orienteering debates about difficulty is that the difficulty depends upon the body build. How much reach do you have? How small are your hands? The latter was the subject of a celebrated climb called Passport to Insanity. The crux pitch is a thin crack through a down sloping 20 foot roof. Only four people have ascended it. The top male climbers of the day failed. Their hands were too wide for the crack. Whilst sitting around a campfire discussing the climb, one of the failed aspirants looked at the diminutive Nyrie Dodd and quipped that her hands were small enough to fit in the crack. So off she went and climbed it. She graded it 28. An debate erupted as to whether it really was grade 28. One argument was that Nyrie had never climbed 28 before. The counter argument was that none of the males capable of climbing 28 had succeeded on it. There you have the subjectivity of climb grading. Since then at least one male has climbed the route, using holds outside the crack.
Sep 4, 2016 5:43 PM # 
I will second Eric's endorsement (above) with an example: compare the US Individual Championships in Wisconsin, October 2009, with the US Classic Championships at Prince William Forest, November 2007; I was a contestant on Brown in Wisconsin and course setter for day 1 at Prince William; By inspection you would be hard put to say that the maps were similar; yet the results were similarly scrambled; both were on very fast terrain and the difficulties contestants had, including me, had to do with visibility; technical errors among very good orienteers were rife and similar; in Wisconsin it was the moraine bumps and in Prince William it was the green vegetation and both were easily subject to parallel error.

A numeric scheme, unfortunately, is likely to be subject to more handicapping than are subjective judgments of competent orienteers familiar with specific terrains.
Sep 5, 2016 1:42 AM # 
Hey All,
Thanks for the feedback and thanks Juffy for the OCAD tips! I'm pretty convinced at this point that it would be to complicated to do what I would like to do due to many of the factors mentioned. Seems like figuring this out from another direction would be more optimal. Thanks
Sep 6, 2016 1:52 PM # 
It could be sorted by doing weighted overlay analysis tool in ArcGIS, I am a GIS analyst, however, the main issue to my view is to identify the weights that accumulate the physical challenge levels etc.. This does not include the conversation from OCAD environment to GIS environment and rasters and re-classifications may have to be used, I am happy to keep discuss this as I would like to develop something similar to MTBO. Having said that, even if we sort categorizing this somehow, that will not take the subjectivity out...

Juffy, you are in?
Sep 10, 2016 5:18 PM # 
Gudeso I'd be happy to talk more offline.
Sep 11, 2016 2:20 PM # 
One aspect not mentioned here is generalization level of the map. See, sometimes the size of features if just right for mapping them, so all of them can be mapped and no much generalization is needed in mapping phase and also no much interpretation needed when using it. And there isn't much features not mapped. But sometimes there is way too much going on and everything cant be mapped, so stuff must be generalized or left out. And runner needs to be able to figure how it is generalized and understand the map instead of just reading it. So maps of these two areas essentially look the same, same detail density and sane contour squigglyness, but the other one quite easy and straight forward to use, and the other is a lot more difficut, far more challenging and technical.
Sep 12, 2016 4:02 AM # 
This nice observation would be useful in other threads about mapping, scale, ISOM etc. Thanks Jagge.

Please login to add a message.