The last hour of a walk is where navigation quietly goes wrong

There is a settled way of talking about fatigue and route decisions. It is worth asking how much of it survives contact with the detail.

Tiredness changes what you notice

Fatigue narrows attention onto the immediate task, which on a hill means the next few metres of ground rather than the shape of the valley you are entering. That narrowing is useful for placing feet and useless for noticing that the stream on your left should have been on your right for the last twenty minutes.

It also reduces how often people check anything, because each check costs effort and the tired brain economises without announcing that it is doing so. The result is longer gaps between position fixes, and a longer gap means a larger error before anybody notices something is wrong. None of this feels like impairment from the inside, which is the property that makes it dangerous rather than merely inconvenient.

Commitment hardens as the day gets longer

The more effort already invested in a route, the more reluctant people are to abandon it, which is a well-described bias and not a personal failing. At the end of a long day the alternative to continuing is usually walking back uphill, so the cost of admitting a mistake rises just as judgement falls.

Worth knowing before you go — this is why parties press on into deteriorating conditions far more often than they turn back early, and why the accounts afterwards all sound similar. It also explains the specific pattern of descending the wrong valley: the ground keeps going down, going down feels like progress, and reversing feels like defeat. The countermeasure is to decide the turnaround conditions in the morning, when the decision costs nothing, and then treat it as already made.

Why descent routes generate the errors

Ascending a hill, all routes converge on a summit, so an approximate line still arrives; descending, every small error diverges and grows with distance. A bearing taken a few degrees off at a summit plateau can put a walker into an entirely different catchment by the time the ground steepens. Plateaux are the worst case because they offer no obvious fall line, and in cloud there is nothing to correct against until the ground is committing.

Locally, the practical habit is to take and follow a bearing off any summit before starting down, rather than after the shape of the ground has already suggested a direction. A short leg walked carefully at the top saves an hour of contouring on a hillside where the alternatives are all steep.

Light disappears faster than people expect

Useful light fades gradually and then quickly, and in a valley or on an east-facing slope the ground is dark well before the sky is. Colour perception drops before overall brightness does, which makes path surfaces, wet rock and vegetation boundaries harder to distinguish while it still feels light. A headtorch fixes visibility and narrows it, since a beam shows a small circle and removes the wider context navigation depends on.

The standard advice everywhere is to be off the difficult ground before that transition, which means treating the last usable light as the deadline rather than the sunset time.

Local sunset times, terrain shading and the season all affect this, so it is worth checking rather than estimating from memory.

The failure modes have names

Confirmation bias is the commonest: the tired walker looks for a feature that fits the assumed position and finds something close enough to accept. Bending the map is its companion, where a stream, wall or track is matched to the wrong one because the expected one has not appeared.

Over a season, following others is a third, and it is especially likely at the end of the day when other people appear to be heading somewhere purposefully. Each of these is easy to recognise in someone else's account afterwards and very difficult to spot in yourself at the time. The reliable protection is not vigilance but process: fixed checkpoints, agreed timings and a habit of confirming position before it is in doubt.

None of this is uniform across a country the size of most of these.

Building the end of the day into the plan

Plan the last section to be the simplest, using an obvious path, a track or a clear valley rather than the most interesting ground available. Set a time by which you expect to be at a known point, and treat failing to reach it as information rather than as something to make up.

Eat and drink before the final stretch rather than after, since low fuel affects judgement more quickly than it affects legs. Agree in the group who is navigating for that section, because shared responsibility at the end of a long day tends to mean nobody is doing it. Where conditions or terrain make this hard, local guidance and a proper skills course are worth more than any list of habits written in advance.

The takeaway

Plan the end of the day first. That is the hour when everything you are relying on works least well.

The route rewards patience more than equipment.