Route difficulty

Route difficulty score

Not all FKT routes are created equal. A 50-mile route with 3,000 feet of gain on smooth trail is a fundamentally different undertaking from a 50-mile route with 10,000 feet of gain concentrated in steep, relentless climbs followed by punishing descents. Distance alone tells you very little. This tool attempts to quantify that difference in a single, transparent score grounded in peer-reviewed exercise physiology.

What the score measures

The score reflects the objective physical difficulty of the route geometry — distance, elevation gain, elevation loss, and the distribution of grades across the route. It does not account for surface type, trail condition, remoteness, navigation difficulty, weather, or the athlete’s fitness. Two routes with the same score may feel very different on the ground; the score captures what can be computed from a GPX file alone.

How it is calculated

The score has three components that are summed to produce a composite index.

Base effort starts from the km-effort metric used by the International Trail Running Association (ITRA), which defines route difficulty as total distance in kilometres plus one point for every 100 metres of elevation gain [1]. This gives a number that scales naturally with the demands of trail running and is already familiar to anyone who has looked at ITRA race classifications.

Grade severity modifier addresses a significant weakness of the ITRA base: it treats all elevation gain as equivalent regardless of how steep the climbing is. The physiology literature is clear that this is not accurate. Minetti et al. measured the metabolic cost of running across gradients from −45% to +45% and found a strongly nonlinear relationship [2]. At a 45% uphill grade, the energy cost per metre is roughly 5.5 times the cost on flat ground. A route that gains 2,000 metres in two brutal 35% grade climbs demands substantially more from the body than one that gains the same 2,000 meters across gentle 8% slopes — yet ITRA scores them identically. This modifier adds a penalty proportional to the excess metabolic cost incurred on segments steeper than 20%, using the Minetti 2002 fifth-order polynomial cost curve as the underlying model.

Descent penalty corrects a second weakness of the ITRA model: it ignores descents entirely. Descents are not free. The energy exchange between climbing and descending is asymmetric — uphills cost approximately 50% more per unit of grade than the energy recovered on equivalent downhills [2]. More importantly, steep descents impose heavy eccentric loading on the quadriceps, and Jessup et al. demonstrated that variable-grade conditions — the kind found on any real mountain route — produce 8.3% higher total cumulative energy expenditure than fixed-grade conditions with identical total elevation change, because of the nonlinear shape of the metabolic cost curve [3]. Minetti et al. further showed that the running cost curve on descents is U-shaped: energy cost decreases to a minimum at around −20% grade, then rises again on steeper descents as the braking demand increases [2]. The descent penalty therefore has two parts: a base term weighted at 50% of the equivalent uphill gain, and a surcharge for segments below −20% where the metabolic cost curve turns upward again.

Jessup et al. also noted that the cumulative muscle damage from sustained or repeated descents degrades running economy in the hours that follow [3]. This effect is partially captured by the grade distribution analysis but is not fully modelled, since doing so would require fatigue accumulation parameters that cannot be estimated from GPX geometry alone.

Elevation smoothing

Raw GPX elevation data carries vertical errors of roughly 3 to 15 metres depending on the recording device and conditions [4]. Computing grade from noisy point-to-point elevation differences produces spurious spikes that would inflate the grade severity modifier. The scorer applies a moving-average smoothing pass over the elevation profile before computing grades, and aggregates points into 50-metre horizontal segments before calculating segment grade. This does not eliminate error but substantially reduces the influence of GPS noise on the score. If you have a DEM-corrected GPX file — for example one processed through CalTopo or Gaia GPS — the score will be more accurate.

Limitations

The score is a geometric approximation of difficulty, not a physiological measurement. Several factors that are highly relevant to FKT difficulty cannot be extracted from a GPX file: surface type and trail condition, technical exposure, navigation complexity, altitude acclimatisation, and whether the route is point-to-point versus a loop with a resupply. The Minetti 2002 metabolic data came from ten elite fell runners on a treadmill; real rough-terrain running costs more, particularly on technical descents where runners cannot run at metabolically optimal speeds for safety reasons [2]. The score should therefore be treated as a relative index for comparing routes rather than an absolute measure of difficulty.

References

[1] Fogliato, R., Oliveira, N. L., and Yurko, R. (2020). TRAP: a predictive framework for trail running assessment of performance. arXiv:2002.01328. https://arxiv.org/abs/2002.01328

[2] Minetti, A. E., Moia, C., Roi, G. S., Susta, D., & Ferretti, G. (2002). Energy cost of walking and running at extreme uphill and downhill slopes. Journal of Applied Physiology, 93(3), 1039–1046. https://doi.org/10.1152/japplphysiol.01177.2001

[3] Jessup, L. N., Kelly, L. A., Cresswell, A. G., & Lichtwark, G. A. (2023). It is not just the work you do, but how you do it: the metabolic cost of walking uphill and downhill with varying grades. Journal of Applied Physiology, 135(6), 1263–1267. https://doi.org/10.1152/japplphysiol.00349.2023

[4] Wehr, A., & Lohr, U. (1999). Airborne laser scanning — an introduction and overview. ISPRS Journal of Photogrammetry and Remote Sensing, 54(2–3), 68–82. https://doi.org/10.1016/S0924-2716(99)00011-8