The “best” route choice has to balance speed vs risk, right?

With 100% GPS tracking we can determine, for each runner, how fast they were on average, by taking their relative speed (vs the fastest runner) on their medium position leg.

I.e. if you have a few very fast legs, and a few relatively slow legs, and the majority around 10-20’th position, then we can take that 15’th position speed (maybe 5% down from the leg winner?) as the base speed of that runner, right?

At this point we can then determine how much each runner have lost due to mistakes on each leg, and allocate each mistake to a given route choice, and get a measure of the average loss due to misses for each alternative.

Adding this to the calculated optimum time for each route choice gives the _expected_ time, which is quite likely to be higher for a risky, direct route choice than a safer & easier alternative.

From the spread of the route-choices, we see that there are a lot of alternatives, both on micro and macro level, and the split times show that some are significantly better than others.
You comment that the path going up the hill on the eastern route choice is slow, and as this is not a path, but a ‘narrow ride’ (symbol 509), the runability is not better than the area the ride runs through.
One of the key ideas of course-setting is that legs should be “visually” balanced, but still require different time and skills on the different route-choises. By moving control 7 east to(wards) the open area, most of the runners would probably prefer the eastern choice (with a simple entrance to the control), even if the leg then is balanced in time. (It was many enough of simple “TV-controls” already decided…)
This shows that this tool is more (and excellent!) for analysis than course-setting.
Keep up the good work, Jan!
..and Merry Christmas to you all!

So was the map bad?