50 Miles Off the Beaten Track: A Navigation-Heavy Race Report
The Roo 50 is a 50-mile loop that starts and finishes in Hahndorf, in the Adelaide Hills, and rolls out over Mount Arapiles before climbing to the top of Mount Lofty. It is a small club race, run by volunteers who set the aid stations, time the checkpoints and hand out watermelon slices that taste like the best thing in the state. I had wanted to run it for three years. This time I trained through an Australian winter, kept my long runs slow and repetitive, and arrived with a plan that assumed the course would be marked properly. That assumption turned out to be the most expensive part of the day.
Races like this live in the details nobody writes on the entry form. The forecast from the Bureau of Meteorology promised a mild start, a southerly change around midday and a top end near 31 degrees. Volunteers had laid drop bags under cover near the pub on the main street, and the local hash house crew had a tent open with Gatorade, bananas and a speaker playing loudly enough to cover the nerves. Forty of us stood on the start line. By the time the sun cleared the ridge, the temperature was climbing and somebody's barbecue smoke was already drifting across the road.
My plan was simple. Run the first 15 miles easy, eat every 25 minutes, refill at every station, and make one pacing decision per hour. Simple plans fail in predictable ways. Mine failed at the creek crossings, where a forestry crew had cleared the signposts in the week before the race and both options on the map looked like trail. I added roughly four kilometres, lost about twenty-five minutes, and spent the middle hours of the race arguing with myself about which direction to point.
I finished in 9 hours 47 minutes, third overall and first in my age group, which is a decent result and not the story. The story is what my watch battery did, what the map did, and how a race that was mostly physical turned into a long administrative problem. Navigation, pacing and decision-making under fatigue behave like the same skills as running a long project: they degrade in the same order, and the failures are rarely dramatic. They are small wrong turns that only become expensive at kilometre 60.
The course, the heat and the map on paper
The first half is fir forest above Hahndorf, where the soft sand and the roots take more out of your quads than the distance suggests. You climb to the summit of Mount Arapiles, descend through gum trees to the old fire tracks near the Adelaide International Raceway, and then face the exposed climb to the top of Mount Lofty. From there the course drops into gullies with a creek crossing at 28 kilometres, climbs again, and wanders through the ferny sections where the light goes green and the temperature drops ten degrees in half a kilometre. Compare that to the flat, fast trails around Melbourne's parklands, where a wrong turn costs you ninety seconds. Here it costs you a quarter of an hour.
The heat was the other variable. An early southerly kept the exposed sections reasonable, but by eleven o'clock the fire tracks were radiating and the creek beds felt like a different climate. A bushfire warning for the district meant haze in the gullies, which made the going even harder to read. There were kangaroos sitting in the grass on the ridges at dawn, unbothered, and a magpie that followed me for close to a kilometre before losing interest.
The course notes came as a single A3 sheet, hand-marked in highlighter and taped into a plastic sleeve. Checkpoints were 11 to 13 kilometres apart, which is fine when the route is obvious and dangerous when it is not. I loaded the track into my watch, noted the two creek crossings, and packed a laminated card with the route on it that I never once looked at. That card was the most useful thing I owned and the thing I was proudest of remembering to bring.
Where the navigation fell apart
Around kilometre 22 the forestry crew's clearing job had pulled the pink tape from the trees. My watch lost satellite lock under the canopy and the two remaining options were a wide 4WD track and a narrow singletrack that looked newer. I took the singletrack. It dead-ended at a collapsed fence line 600 metres later. Turning back on the same track, with the sun already up, cost me twenty-five minutes and a chunk of my patience that I did not get back.
The rest of the middle section was a lesson in borrowed equipment. My phone battery dropped to eight percent; I put it into airplane mode and used the screen to check the time, which is a crude but effective trick. At the third checkpoint a volunteer handed me a spare battery from a pouch of donated ones, and two other runners in my 40-plus group were walking the same wrong creek I had driven, which is both embarrassing and comforting. The race had a six o'clock cut-off at the far checkpoint, and that single number on the wall changed how I thought about every junction afterwards.
By kilometre 60 my decision-making had narrowed to a single question, and I could hold exactly one answer in my head at a time. This is where running stops being about fitness and starts being about administration. The same thing happens on a long project: at the start you can hold the plan, the budget and the deadline in view, and by the end you are holding one number and one next move. Accepting a wrong turn early is far cheaper than debugging it in the dark, because reversing cost is a function of how much you have already committed.
Eating and drinking past the halfway mark
What worked was boring. Watermelon at the first aid station, a gel every 25 minutes, a Bika at halfway because it is cheap, available at every service station in the country and does the job without fuss. Volunteers ran out of iced soft drink at one table and that single fact shaped the last two hours more than any coaching cue. Australian race food is not glamorous, and most of it is better for being unpretentious.
What failed was the same thing that fails in every ultra: I stopped eating for nearly forty minutes after the halfway aid station. My gut shut down, my hands got shaky, and the effort went from hard to grim. The stomach behaves like a fourth limb that you train separately from the other three, and by kilometre 45 mine had stopped taking instructions. A quarter of an energy gel every fifteen minutes, a few sips of water, then another gel, held it together better than anything else I tried.
Recovery was simple too. I sat on the grass near the finish with a Milo, a Tim Tam and a lawn chair, and let somebody else's kitchen do the work. The Boost juice arrived at the same time as the medal, which is the correct order of operations. The whole lesson of the fuelling section is unoriginal but worth repeating: small amounts, often, and never the heroic thirty-minute starvation that feels like discipline while it is actually just an unpaid debt.
The kit that earned its place and the kit that did not
Most of my gear was justified, and the failure was concentrated in two items: the navigation setup and the food plan. Worth naming clearly, because gear reviews that only praise the gear are not reviews at all.
What earned its place:
- The laminated route card, consulted once and immediately useful.
- The spare phone battery that came from the checkpoint volunteer's pouch.
- A soft flask that refilled at every station without me unscrewing anything.
- Trail shoes with a grippy lug pattern, which mattered more than the plate choice on the wet rock.
What I would leave behind:
- The full-size 6-litre pack, replaced by a 4-litre that did everything asked of it.
- The second pair of navigation watch settings, which I set up and never switched to.
- The extra pair of dry socks, since the creek crossing happened at kilometre 55, not kilometre 28.
- The 200 grams of bars that tasted good and did nothing useful.
What I am taking into the next training block
The physical preparation was fine. Over four months I ran about 900 kilometres, kept three long efforts above 30 kilometres and never missed the easy runs, and none of that needed changing. The gap was contingency planning. I had a pacing plan and no navigation plan, which is like budgeting a project carefully and assuming the internet will stay on.
While I was re-reading old notes, I spent a couple of evenings on weight and training load, a Brazilian site whose practical take on fuelling for long efforts lined up almost exactly with what the course taught me: small inputs, repeated often, with the boring consistency doing most of the work. I have written up the physical side of these races in my previous race reports, but this one was not really about running speed. It was about what happens when the plan meets an unmarked track.
The first thing in my calendar is a 6:15am meet at the Old Magill Railway Trail, where I am re-running the climb toward the top of Mount Lofty with a paper map, a compass and a spare battery in my pack. It is forty-five minutes of running I have already done a hundred times, and it is the closest thing I have to a rehearsal for the part of the race I got wrong.