Training Log Archive: TheInvisibleLog

A short break this morning before starting on logistics work for Long. I hope some find the following explanation of course setting for the Middle illuminating. I am still debating with myself whether this event or the 2002 Vic Champs the day after WMOC was the most complex course planning course I have undertaken.

There were major constraints. First, we had only one quarter of a square kilometre of middle suitable terrain to work with for the hard courses. The second constraint was the existence of only one feasible route between the terrain and the arena - a narrow bridge. The route along the river was choked with fallen timber and scrub.

We dealt with the second constraint by the judicious use of electric chainsaw and pruning loppers. Before you feel shock about this, the day we were out Parks was also out with much noisier chainsaws. A year prior to the event a massive wind storm went through, clear felling the south east slopes of the hills, damaging infrastructure and causing general chaos. The bridge was damaged, the road to the start was unpassable and many of the tracks used for novice courses were buried under trees.
https://photos.app.goo.gl/nwB5M7vTuwZSe7hq6

Now the course design process. For the first time in my life I started by creating a grid of controls before even starting to think about course shape. The grid covered the available terrain, except areas near the Yankee Creek in the north that we felt were too rough for enjoyable orienteering. Each control was designed to have no adjacent controls on a similar feature (whether appearance on ground or representation on map) and minimum control separations were adhered to. This meant that sometimes we were very careful about the side of the feature that hosted the control. GPS measurement helped us here. There was a very good reason why the control might have been on the north rather than south side of the knoll. Sites were chosen to provide reasonable flow through the controls. Here is the all controls map.

Next step was to design the M21E course. This was not because this was a priority over other courses, but to allow control of course interactions as further courses were created. Course 1 needed to do two loops of the available terrain to get the required winning time, and the two loops needed to cover different parts of the terrain. The first loop was an outside loop, the second was an interior loop. A map flip was unavoidable. Towards the end of the first loop was one of only few route choice legs for the event. This was the one chance to give orienteers the opportunity to choose a fast running option. This leg was common to quite a few courses.

The next off the rank was course 2. I copied course 1 and then adjusted it to the required length. Half the controls were in common with course 1. This approach ensured there were no legs where runners coming out of a control would lead those on the other course into the control.

So now we have two courses and it looks a little spaghetti like.,

But we have 15 more hard courses to create, so the spaghetti has only started. Each course was created by copying the previous one and further adjusting to the required length. When it was all done it looked like this.

Observations.
* Note the fan of lines at the start. Every course had a different first control, with the shorter courses going gradually more to the left from the start. The first leg was a primer on the nature of the terrain so that when the route choice leg came up, runners knew what they were choosing between - the rough terrain or the wide track.

* There were few controls where runners were expected to be entering and leaving on the same line. To get to this point explains many of the course versions. Resolving one control issue often led to another issue elsewhere. Eventually we realised we couldn't resolve every issue. In mathematically terms, it was an indeterminant matrix. The remaining issues were mostly caused by the design of the shortest hard courses where the decision was made to prioritise keeping the oldest age classes out of the rougher terrain. The numbers on these courses were small so the impact was marginal. The Event Adviser decided to address this through the start windows for the relevant age groups. The way we could keep track of this complexity was by using the condes tools for identifying approach and exit angles and between-course acute angles. For those that use other course setting software, are your tools up to a task like this?


* Separate courses were needed for M16 and W16. The complexity of the terrain meant a fully hard course was probably a step to far for those running moderate courses the year before. You might say these runners had a moderate hard course.


* As if this was not enough, we took into account the variable roughness of the terrain. We used a mud map of roughness zones to try and keep the courses with older competitors in the less rocky terrain for a greater proportion of their course.

* After loading in the classes, course groupings and competitor number estimates, we had to tweak the latter part of quite a few courses to reduce control load on any individual control to 300 or below.

* And finally, we made a decision to give age classes W65 up and M75 up choice of crossing the river or the bridge. I believe the river was faster, but it appeared many women in particular chose instead to keep their feet dry.

Course design like this is about balancing competing objectives. Some feedback was that a few of the legs on the elite course were not elite standard. For most courses, improvement was obviously possible if one ignored the impact on other courses. We couldn't ignore them though. In the end, the legs with a disappointing outcome were not the ones I expected. On M21E leg 12-13 was not optimal. On the other hand, 17-18 proved much more "elite" than I expected.

Optimising all the competing course setting objectives was a true team effort between this course planner and the IOF adviser (Chris Norwood) with input from Warren Key (Technical Director), mapper Fredrik and test runner Kerrin. No one person could keep on top of it all. The final event file was labelled version 18_dry (The mapper Fredrik may well have been through a similar number of versions). I think I re-started the course design process from scratch at least twice, but I suspect that is an underestimate.

Did you catch the word "dry"after the course version number? There was a river crossing. For some months prior to the event we had been maintaining separate wet and dry course versions as a hedge against a high river flow. And as a postscript, a week before the event the river was in flood and the BOM forecast was for more rain. We had been given a print deadline of a week before the event. Based on a nasty river level and the rainy forecast, we printed the shortened wet set of courses with a remote finish.This decision caused some angst.

The forecast rain during the week leading up to the event didn't eventuate and by Thursday we could see the river falling so there was some pressure for a map reprint. On the Friday, two days before the event, a frazzled Jim Russell printed the original courses (dry version). Part of the pressure he was under was that other carnival maps remained to be printed (mainly terrain exclusions due to the wet weather). I acted as printer's assistant and came to understand why Jim was frazzled. The task of printing took four hours. Each sheet of paper needed to be hand fed into the printer (Pretex tends to stick). Each map flip had to be matched to its mate. Colour checks need to be made every now and then. Every course must have the correct numbers and be free of interlopers. If you think its just a case of pressing a button and going and having a coffee, think again. So a massive thanks to Jymbo for stepping up on that one.

The final decision about which courses to use was made at 3pm on Saturday when the arena was being created.

C • Thank you.. 9

C • Course version number 7

C • I would add 7

Snuck in a run...

No log entries for a while. Put that down to carnival. So many people are over it. Tempers are fraying. Deadlines are being missed. Trying hard not to point fingers.

C • Ditto 21

One of those rare efforts where I gained no pink cells on the winsplits display. Every control just fell into place. I think all the recent time on Nuggety helped.

C • No pink cells 3

Gee this one was hard work. Need3d a run to keep back issue from getting worse.

Fiddlers Green. Massive versioning fail for club. I hadn't added the new map to the database so the old one was used. Other than that personal disappointment, a good day. First 9 minute per k race in a long long time. No errors (for once having field worked an area might have helped, though it was four years ago). Julie had a good run as well, and was ahead of me until the technical controls in the last third of the course.

C • map version 2