I’ve had some close calls in my life. My last wasn’t close, it was right under my feet and over my head. On February 16th 2014, I and two friends remotely triggered an avalanche on the mountain Lebnes in Kittelfjäll, Sweden.

The avalanche broke about 100 m above us and caught and partly buried us all. The crown was roughly 300 m wide and the path was about 400-500 m. In some miraculous way, we all survived. However, the incident did not leave us unharmed: one tibia, one fibula, a rib, a chin and one femur where broken, and so was our self-confidence.

How could we have let this happen? We did not consider ourselves to be daredevils by any means, and yet there we were like jackstraws in our self-made debris.


The crown of the avalanche on Lebnes. Photo: Martin Stefan.


One broken Fibula, and one broken Tibia.


A very lucky girl. Photo: Martin Stefan.

Ever since the accident, my brain has been preoccupied with that question. What happened that day? Where did we go wrong, and why? Because errors were made. Multiple ones.

I have been both fascinated and frightened by avalanches, and by how we as humans handle the complex risk situation that avalanche hazard expose us to, for a long time. On February 16th, 2014, I had the opportunity to do an admittedly very unscientific case-study on myself. Let me start with a few facts about the accidents, the weather, snowpack and the victims, i.e., us.

Facts about the weather and snowpack: After a period of cold temperatures and relatively little wind, Kittelfjäll was hit by a snow storm that lasted for 10 days. During the storm, the temperature in the valley was around freezing and there was not much wind. On February 15th, the wind picked up and the snow fall intensified. The local weather station measured 27 cm of new snow on the night between the 15th and 16th of February.

Facts about the victims: The group consisted of three close friends, I, my partner Martin and our friend Maria. I and Martin had spent the previous winter in Wyoming, Colorado and Utah touring the backcountry. We started the season in the US with an AIARE avalanche 1 course in Jackson, WY. The three of us had toured together several times before and knew each other from way back. We knew that we were amateurs, but we considered ourselves to suffer from slight snow nerdiness. We talked about avalanche hazard ALL the time. We dug pits, we read avalanche books and bulletins, we talked about communication and responsibilities. At the day of the accident, everyone in the party was wearing a transceiver, and everyone was carrying a shovel and a probe.

Facts leading up to the accident: Kittelfjäll is located in the mountain range that separates Sweden from Norway. It takes about 5 hours to drive there from Umeå, where we all lived at the time. We had no big intentions with our trip to the mountains. I was recovering from a month of heavy work load and from the shingles, and Maria suffered from a bad foot. Our only mission was to get out and play in the snow. During the day before the accident, we followed our plan and did a small tour up Borkan. A snow pit analysis on the way up revealed a weak layer underneath storm snow. It took relatively brute force to get a collapse, but we got a clean sheer when it collapsed and even though we did not make an ECT we concluded that there was enough energy to produce propagation. We consequently decided to stay in mellow terrain and to keep security distances.


Maria and Martin analyzing the collapsing layer.

On the day after, Sunday 16th of February, 15 cm of fresh snow had fallen on the parking lot outside of our rented cabin and temperatures where slightly lower. Our plan of the day was to tour up on the opposite side of the valley from Borkan to investigate a small treed gully that we had spotted the previous day. The terrain was unknown to us, but we knew that we could approach and check out the run via a low inclination ridge that we deemed to be very safe. If we found out that our plan was too risky, we could always just turn back. On our way up the mountain, we felt several collapses in the snowpack. The collapses were large. At times, we could see the trees shaking off their load of snow due to them. To test stability, we used a test slope and tried to get it to slide. We got a collapse but not a release. As we got higher up the mountain side, the collapses stopped but when we reached tree line we still decided to turn around. The run we had planned to ride was too steep and the outrun too uncertain. We didn’t want to end up in a terrain trap. We wanted to play it safe, so we decided to turn back. At the time we reached our decision, we were standing at tree line at the end of the ridge we had come up. Above us, the terrain was substantially steeper and exposed to the wind. We did not want to be there. Below us was a a ridge that required poling to get down. We did not want to go down there. Before we took off our skins and pointed our skis downhill, we therefore decided to do a small uphill traverse, just to get a few turns in. So we stepped away from the trees, and out into a larger snow field. It collapsed. The avalanche came down like a tsunami from above. It hit me in the neck and tossed me like a toy into the trees.


Traces of the avalanche, summer 2014.

If we list all the warning signs that we got that day:

  • Heavy snow fall
  • After a period of cold temperature and little wind
  • Increasing wind
  • Clean sheer from the CT
  • Multiple collapses in the snow

The inevitable question that you have to ask is, where we completely out of our minds?

The short answer to that question is yes, we were, and still, it appears as if we were also pretty “normal”.

Reliable research on human factors in avalanche accidents is still all but missing, but there is suggestive evidence that many victims disregard seemingly obvious warning signs. Just like we did. Ian McCammon (2004) and Dale Atkins (2000) have identified a number of situations in which victims miss more warning signs than others. These situations have gotten the name “heuristic traps”. A heuristic is a mental shortcut that is often used when analytic reasoning demands a lot of cognitive effort and there is a need to make speedy decisions. Examples of this method include using a rule of thumb, an educated guess, an intuitive judgment or common sense. In most cases these rules of thumb work relatively well. In avalanche terrain however, the use of them can have catastrophic consequences. McCammon identified 6 heuristic traps that victims of avalanche accidents appear to have fallen into. Although reliable evidence of that these traps caused the accidents is stil scant, the mechanisms behind them has been documented in both psychology and economic research.

My idea with this blog post is to relate the decisions that led up to our accidents, to the heuristic traps as defined my McCammon, and to relate those to previous research in psychology and economics.

McCammon’s 6 heuristic traps have gotten the nickname FACETS:

  • Familiarity – our tendency to perceive the familiar as safe.
  • Acceptance – our tendency to do things to gain respect and acceptance from others.
  • Consistency – our tendency to do what we planned to do, and to confirm our hypotheses.
  • Expert halo – our tendency to follow the leader.
  • Tracks (scarcity) – our fear of losing out, and tendency to compete over scarce resources.
  • Social facilitation – our tendency to feel safe when others appear to feel safe.

All of these tendencies are linked to the cognitive and social functioning of human beings. We have a very advanced brain, and to make the most out of it, we have become very good at avoiding using (cognitive) energy on unessecary things. We are absolutely super at making quick judgements based on our previous experiences and clues in the environment. Vi are also very good (at least at times) at coordinating behavior and cooperate in social groups. Finally, our innate desire to be happy pushes us to constantly try to improve our situation. All of these skills are great, and have taken us to where we are today. However, they can also cause us to make really bad decisions.

Lets take a few examples.

  1. Familiarity

When we walk, or drive to work, we do not waste cognitive resources on evaluating which way to go, or how conditions have changed since yesterday. Instead we think about other, more important stuff while we transport ourselves from our home to our work. Great!

When we walk up a mountain that we have walked up many times before, we may neither “waste” resources on re-evaluating the conditions and the best route to take. Instead we think about other, seemingly more important stuff. As long as our familiar mountain is considered avalanche terrain, this type of energy conservation is not so great. Because the snowpack is never the same, and failing to devote our cognitive resources at evaluating it may cause us to miss important warning signs (see e.g., Kahneman, 2011).

Did we fall into the trap of familiarity on Lebnes?

I, Martin, and Maria had not skied that particular slope before, but we considered Kittefjäl to be our “home mountain”. I and Martin spent the previous skiing season in Wyoming and Colorado, where the snowpack is infamous for being avalanche prone, and we had just returned from a month in Jackson. Kittefjäll felt very safe. I had actually even articulated my feeling of safety two weeks earlier when I expressed how good it felt to not have to worry so much about avalanches (!). When we skied in the Rockies, we always took great precaution to avoid exposure from above whenever there was the slightest risk of remote triggering. In Kittefjäll, we all felt the collapses, we all saw the clean sheer, and we all agreed that the main risk of the day was remote triggering. Still, we went up and exposed us to the steep terrain above.


IMG_5691 (1)

Maria traversing at tree line.


2. Acceptance

AcceptanceHumans’ ability to read social signals makes it possible for us to adjust our behavior so that we are not excluded from valuable social groups. Most of us base our self-esteem on how we think that we do in the eyes of others. We tend to evaluate how good we do by comparing our behaviors to that of others. This has the benefit of making us behave in accordance with social norms, which in turn facilitates cooperation (see e.g., Baumeister and Leary, 1995). However, a negative side effect is that we also feel bad when we don’t meet the expectations that we think that others have of us, and that we tend to compete for status. The former of these two can make us refrain from speaking out when our opinion or evaluation goes against that of referent others. The latter means that we get a raise to the top (or bottom) but without any gain. Economist call this phenomenon “keeping up with the Joneses”, and the cause of it “positional preferences” (see e.g., Alpizar et al, 2005, and Carlsson et al., 2007). To gain social status in the backcountry community, we may feel that we need to ski lines that are considered as “rad enough”. How steep or exposed “rad enough” is depends on how steep and exposed other members in the group skies. If people start skiing more rad lines, the other have to follow to keep up. Now, skiing rad lines may give us pleasure in itself. Afterall, it feels good to master that skill. The point is not that we should not ski rad lines, but when we just raise the stakes to gain status, nobody wins.

Did we fall into the trap of Acceptance on Lebnes?

Well, I can’t speak for the others, but  I certainly did. I still do, over and over again. I constantly fear that I will be considered to be that “boring, coward, chick”. I want to be that “cool, funny, and skilled” touring partner. I want people to want me to be their friend and ski buddy. I want people to see my pictures on social media and think, wow she is bad ass ! I constantly evaluate what I ski to what others ski, and so I know that I will never be bad ass, and I feel bad about it. Even though I am terrified, I force myself to go out into avalanche terrain and push my fear limit. I am working on this. Still working.

3. Consistency

Humans’ genious ability to use energy efficiently not only “helps” us to make automatic decisions in familiar terrain. This characteristic of ours also makes us prone stop searching when we’ve found what we are looking for, and it diverts our attention so that we find what we are looking for. Psychologists call this phenomenon “confirmation bias”, because it makes us confirm our hypotheses (see e.g., Klayman, 1995, and Nickerson, 1998). This would all be fine and dandy if a hypothesis is proved once it is confirmed. The problem is however, that many false hypotheses can be confirmed if we do not look for the right information. Consistency also implies that we tend to act in accordance with our plans. If we don’t, we experience what psychologists call “cognitive dissonance” (see e.g., Aronson, 1968). Cognitive dissonance is not a nice experience. It makes us question ourselves, and we therefore have a tendency to rationalize our choices and stick to our plans.

Did we fall into the consistency trap on Lebnes?

I would say, yes. We heard several massive whumpfs, but we were committed to go up and check our line. We followed our plan. To test stability we used a slope and tried to get it to slide. We got a collapse but not a release, and we drew the conclusion that there was enough friction in the snow to prevent it from sliding, or perhaps to little energy to generate propagation. We failed to search for other factors that could have prevented the slide. The slope was short, and supported from below… When we came up to the treeline, I made a hand sheer test. I had to pull quite hard to get a release. When the block came loose, the sheer was clear. I focused on the first factor.

IMG_5689 (1)

Martin walking up the whumpfing ridge.

4. Expert halo

Groups can make us smarter, and dumber. I think that we’ve all found ourselves in a situation were our cooperation with others helped us solve a problem that we would have been unable to solve alone. In other settings, however, results are not so benign. One example of this is the Challenger disaster of 1986 (read more about it here), where NASA chose to launch the rocket in spite of obvious and documented problems with an O-ring months before the launch. The shuttle exploded 73 seconds after launching because of the problem with the O-ring. Groupthink (coined by Irving L. Janis, professor in psychology at Yale) is mainly found in cohesive groups, when members’ strive for unamity overrides their motivation to realistically ascess the alternatives.

Did we fall victim of the Expert halo on Lebnes?

I, Martin and Maria are really good friends. We go way back, and we talk a lot about safety whenever we are on a mountain. As we discussed our options on Lebnes, everyone participated and talked about their concerns. So did we escape the expert halo trap? I don’t know. What I do know is that I always feel uncertain about wether my concerns are valid or based on ghosts. I therefore usually express them, but then let someone else decide. In other words, I choose a leader and leave the responsibility to that person, regardless of wether s/he wants it or not. The expert halo trap is also related to that of social acceptance. Constantly crying about being afraid doesn’t make you a fun ski buddy.

5. Tracks (scarcity)

Most people have experienced the fear of loosing out, and this fear may make us accept greater risks than we would otherwise do. In a number of experiments, both psychologists and economists have found that people, who are risk-averse in the gain domain, display risk-loving preferences in the loss domain (see e.g., Thaler et al., 1997). In more normal language this means that a lot of people prefer to take the risk of loosing a lot if there is a small chance that they will not loose anything, to a situation with a small but certain loss.

Did we fall on the Tracks on Lebnes?

For skiers, powder is a scarce resource. If we do not ski it, others will, or the wind (or sun) will destroy it. Not skiing a run filled with untouched powder is a loss. We were the only ones on Lebnes that day, so we did not compete for powder with other people. However, as we were standing there at tree line, it definitely felt like a loss going down the same flat ridge as we came up. In addition, we were driving home to Umeå the same day. There are no mountains in Umeå and powder is definitely a scarce resource in the Swedish mountains.

6. Social facilitation

Our ability to judge a situation based on others’ reactions on the one hand helps ut at responding to dangers. If we hear someone scream “get down!!”, we react in a split second without spending too much time at thinking about why we should get down or if this is the adequate thing to do. We just do it, and that may help us avoid getting hit by something. However, in some cases the failure to react by others, causes us to fail to react as well, even though there are obvious signs of danger. One example of this is an experiment conducted by Bibb Latane and John M. Darley in 1968 (read more about it here). In this experiment, participants were asked to fill out a questionnaire in a closed room. While the partcipants were filling out the questionnaire, smoke started comming in underneath the door to the room. If the participant was alone in the room, it took on average 2 minutes for him/her to report the smoke to the experimental leader. However, if there were others present, who did not react (these people were experiment confederates), only 10% of the subjects actually got out of the room or reported what was a serious problem. In other words, a majority of the participants sat passive (and coughing) and did nothing in spite of the room being filled by smoke!

Did we socially facilitate our decisions that day on Lebnes?

I always look to others for comfort. If no one else is scared, why should I be? I guess we may all have been in that smoked filled room on Lebnes.

So I’ve had a close encounter with my own cognitive errors. I can say for sure that I at times am completely stupid, and human. I am very happy to be alive.


Alpizar, F., Carlsson, F. and Johansson-Stenman, O. 2005. How much do we care about absolute versus relative income and consumption? Journal of Economic Behavior and Organization 56: 405-421

Atkins, D., 2000. Human factors in avalanche accidents, Proceedings of the 2000 International Snow Science Workshop, October 1-6, Big Sky, Montana

Baumeister, R. F. and Leary, M. R., 1995. The need to belong: Desire for interpersonal attachment as a fundamental human motivation, Psychological Bulletin 117: 487-529

Carlsson, F., Johansson-Stenman, O. and Martinsson, P., 2007. Do you enjoy having more than others? Survey evidence of positional goods, Economica 74 (2007): 586–598

McCammon, I., 2004. Heuristic traps in recreational avalanche accidents: Evidence and Implications, Avalanche News 68

Kahneman, D., 2011. Thinking fast and slow, Macmillan.

Klayman, J., 1995. Varieties of Confirmation Bias, Psychology of Learning and  Motivaiton, 32.

Nickerson, 1998. Confirmation bias: A ubiquitous phenomenon in many guise, Review of General Psychology 2(2)

Thaler, R.H., Tversky, A., Kahneman, D. and Schwartz, A., 1997. The effect of myopia and loss aversion on risk-taking: an experimental test, Quarterly Journal of Economics






Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s