The heuristic traps that get experienced people killed

Read enough accident reports and something uncomfortable emerges. The people involved are usually not beginners. They frequently have the right gear, current training and years of experience. They walked into hazard that, from the outside, looks obvious.

The explanation is not stupidity. It is that human decision-making runs on shortcuts which are efficient nearly everywhere in ordinary life and catastrophically wrong in a small number of outdoor situations. These are usually called heuristic traps, and knowing their names is genuinely protective, because they are recognisable in the moment once you have a label for them.

Familiarity

You have skied this slope thirty times. You have paddled this section every spring. Nothing has happened, so the terrain feels known and safe.

But you did not survive those thirty days because the slope was safe. You survived because on those days the conditions were fine. The terrain is a constant; the conditions are the variable, and familiarity quietly swaps one for the other in your head.

The counter is to ask, out loud, what is different today. Not whether the place is familiar — whether today is.

Commitment

The drive was five hours. The permit took four months. Flights are booked, the weather window is closing, the group has been talking about this for a year.

None of that information is about whether the route is safe. All of it feels like it is. Commitment converts sunk cost into apparent justification, and it is the single most reliable predictor of a group continuing past the point where they had already noticed the problem.

The counter is to name the investment explicitly before you decide. "We have spent a lot to be here, and that is not evidence about the snowpack." Said aloud, it loses most of its power.

Social facilitation

Other people are there, so it must be fine. Tracks on the slope mean somebody already went. A larger group feels safer than a small one.

Tracks prove one thing only: that a slope did not release for one person at one moment. Avalanche paths regularly hold many tracks before the one that goes. And a bigger group is not safer — bigger groups make worse decisions, because responsibility diffuses and nobody wants to be the one who stops fourteen people.

The expert halo

Someone in the group is more experienced, or is perceived to be, and everyone else quietly outsources judgement to them. Often that person has not agreed to lead and does not know they have been appointed. Sometimes they are simply louder.

Meanwhile the least experienced member — who is often the most nervous and therefore the most attentive — says nothing, because who are they to question anyone.

Fix it structurally. Ask the quietest person directly what they think, and ask before you state your own view. The order matters: once the leader has spoken, you are no longer getting independent information.

Scarcity

Powder gets tracked. Swell drops. The ice will not be in condition next week. Anything perceived as limited pushes people to act faster and accept more risk to secure it.

Competition makes this sharper. If another party is heading for the same line, the pressure to move stops being about the mountain at all.

Acceptance

Wanting to be seen as competent, bold, or worth including. It affects everyone and it is particularly strong in mixed-ability groups and among people newer to a scene, who are reading a social situation as much as a physical one.

This is the trap that produces silence — the person who was worried and said nothing. Nearly every accident debrief contains at least one of those.

What actually helps

Decide the limits before emotion is involved. In the car park, cold and unexcited, agree the turnaround time and the terrain limit and write them down. A number agreed in advance is much harder to argue with than a feeling generated at altitude by a person who is tired.

Assign someone the explicit job of arguing against the plan. Give them permission and expect them to use it.

And separate the two questions that always get merged: can we do this and should we do this today. Skill answers the first. Only conditions answer the second, and no amount of competence changes what the snow is doing.