You've gathered information on snow stability, but will it guide your decisions? One study found that 89% of recreational avalanche victims had prior indication that the danger level was high.
On March 1, 2003, eight experienced backcountry skiers set out to ride Microdot Peak in Alaska's Chugach range [4]. With four feet of new snow and bluebird skies, conditions were ripe for some awesome turns.
They were also ripe for big slides. Three days before, the avalanche forecast for the area was “Considerable to High”, meaning that avalanches, both natural and human-triggered, were likely. A pit dug by the group near the base of the run revealed an instability, yet the group proceeded.
After 4 skiers successfully descended the slope the fifth skier triggered a large slab avalanche and was carried 700 feet. Two skiers watching from the bottom were knocked down by the powder blast. Fortunately, the skier caught in the slide survived, and the two knocked down were unhurt. It could have turned out different.
Why did this experienced team ignore the avalanche forecast and their own snow stability test results? The answer probably lies in what’s known as “Human Factors” in the avalanche business.
Briefly stated, human factors comprise the decision-making component of our overall risk while traveling in the backcountry. If the only ingredients to our decision-making process were objective facts on snowpack stability, then avoiding avalanche hazard would be a simple matter of applying a formula. But as researchers such as Ian McCammon, Bruce Tremper, Jill Fredston and Doug Fesler have shown, our decisions can be influenced by factors that have nothing to do with the situation, let alone objective facts. What makes those factors insidious is that for the most part, we are unaware of the “illogic” of our own thinking.
Heuristic Traps--A Conceptual Framework
"Most of us sort out our probabilities not with statistics but according to what we have actually experienced. Psychologists call this the application of "heuristics". Heuristics are the generalizations we accept about the working of the world, drawn from the grab bag of what we've seen personally or what we've heard from trusted sources like friends and relatives."
Source: James R. Chiles, Inviting Disaster: Lessons from the edge of technology, HarperCollins 2001
There are many formulations of human factors. Some, such as the current Level I avalanche curriculum published by the American Institute for Avalanche Research and Education (AIARE), list many dozens of examples, while others, such as Fredston, et al [2] attempt to place them into abstract categories. Both approaches have their merits, but for conceptual understanding I find the “heuristic traps” identified by Ian McCammon among the easiest to grasp.
In a 2004 article [5] in Avalanche News, McCammon investigates the role of unconscious heuristics, or rules of thumb, in 715 avalanche accidents in North America occurring between 1972 and 2003. As human beings, we develop these heuristics over the course of a lifetime as a means of simplifying the task of living. While they function well in the social and technical milieu from which they arise, in the context of avalanche safety they can be dangerously wrong. Just how wrong is indicated by another study, which found that 89% of recreational avalanche victims prior indication that the danger level was high [1].
Rules of thumb become “heuristic traps” when they are applied unconsciously to situations for which they are inappropriate. McCammon identifies six heuristic traps to which backcountry travellers are susceptible: Familiarity, Consistency, Acceptance, The Expert Halo, Social Facilitation, and Scarcity. Each has its roots in other aspects of our lives, where they are both relevant and essential.
Indicator
Frequency
Obvious slide path
82%
Recent loading by wind or snowfall
66%
Terrain trap
58%
Posted hazard above "Moderate"
55%</TD >
Recent avalanches
35%</TD >
Temperatures above freezing
20%
Signs of instability
17%
Hazard indicators used to quantify hazard exposure. McCammon [5] assessed an exposure rating from 0-7 by tallying up the number of the above indicators present in each accident. Of the 715 incidents studied, scores of 3 or 4 accounted for half of all incidents, while scores of 2 and 5 occurred in approximately 16% and 18% of incidents, respectively.
Table 1. Hazard Assessment
Thus, it is not hard to understand why they are so influential. But how influential depends in complex ways on other variables such as the size of the party, prior training, gender and personality composition of the group. McCammon evaluated the effect of each trap by analyzing over 700 incidents and characterizing them according to hazard level (see Table 1), group size, and training level. A synopsis of the results is given below.
Familiarity
Familiarity with a place or situation allows us to function efficiently, as we do not have to figure out from scratch what to do each time we encounter it. We simply behave as we have always done. Unfortunately, in avalanche terrain the effect of prior experience can lead us to take chances we might not take in unfamiliar territory.
Curiously, familiarity with an area has the strongest effect on highly trained parties. McCammon found that groups with advanced training in terrain with which they were familiar exposed themselves to nearly twice the hazard level of less well-educated groups, and about the same level of hazard as parties with no avalanche training at all.
Consistency
Politicians and corporate managers who, having made one bad decision, follow it up with a whole string of bad decisions merely because they don’t want to change their original position provide an endless supply of fodder for the likes of Bill Maher and Scott Adams. But let’s face it, we all do it from time to time. When consistency with previous decisions helps us cut through distractions and stay focused on the task at hand it is a good thing. But when it blinds us to new information that suggests that “staying the course” is a bad idea, it can lead us into trouble. McCammon and others found that there was a significant increase in the level of risk taken by parties who, for one reason or another, were committed to a particular course of action. Groups larger than two, and parties with at least some formal avalanche training seem to be most susceptible.
Heuristic Trap
Accident party size
1
2
3
4
4+
Familiarity
+
Acceptance
Consistency
+
+
Expert Halo
+
+
+
Social Facilitation
+
+
+
Scarcity
+
+
The influence of some heuristics may vary with party size. A blank cell indicates weak or no correlation. A plus sign indicates positive correlation, a minus sign, negative correlation. Orange cells indicate a stronger correlation (90% confidence level), and red cells indicate strongest correlation (95% confidence). See McCammon [5], figure 8.
Figure 1. Impact of group size
Acceptance
The desire to be noticed and accepted has a powerful influence on human behaviour. There are many possible variations on this theme, including peer group acceptance and gender acceptance. McCammon focused on gender acceptance; specifically, how the presence of women in a group affects the behaviour of men. He found that mixed gender groups exposed themselves to higher risk than all-male groups. The effect does not vary with the number of people in the party. However, groups with only avalanche awareness, but no formal training, exposed themselves to the highest risk.
Expert Halo
Following a leader is a very human trait. It helps simplify the task of deciding how to respond to our complex world. But when our follower instinct is triggered by a person due to their personality or perceived level of experience rather than their actual qualifications, it can lead us astray. Indeed, in examining the risks taken by groups with and without a recognized leader, McCammon found that groups without a leader exposed themselves to less risk than those with clear leadership.
Heuristic Trap
Accident party training
None
Aware
Basic
Advanced
Familiarity
+
Acceptance
+
Consistency
+
+
Expert Halo
+
+
Social Facilitation
-
+
+
Scarcity
+
Training can affect how a party is influenced by heuristic reasoning. A blank cell indicates weak or no correlation. A plus sign indicates positive correlation, a minus sign, negative correlation. Orange cells indicate a stronger correlation (90% confidence level), and red cells indicate strongest correlation (95% confidence). See McCammon [5], figure 9.
Figure 2. Impact of training
This was most pronounced in groups with minimal avalanche training: unskilled teams, with or without leadership, exposed themselves to less hazard than avalanche-aware groups with a poorly trained leader. Group size also is a factor; groups of three to ten people take bigger risks when there is a perceived leader.
Social Facilitation
Have you ever felt emboldened by the fact that you were not alone, and that you might have an audience for your exploits? Or conversely, have you ever felt like you didn’t want your relative lack of ability to show when others might be watching? If so, you have experienced what McCammon calls the Social Facilitation trap. He found that groups with at least some formal avalanche training took significantly higher levels of risk when they encountered other parties, whereas those with no training actually exposed themselves to less risk.
The only thing between you and an awesome ridge climb is this steep slope. Has your desire to do the route on this splitter day influenced your judgement of its stability?
Scarcity
The human tendency to value a resource more highly when it is perceived to be scarce has had a profound influence on history. Whether it be oil, gold, or that last jacket on the 80% off clearance rack, people will compete for their share before supply runs out. Not a bad strategy when you’re talking about things like food and water. But when McCammon looked at powder hounds headed for an untracked slope, he found that parties who saw another group headed for the same place took significantly higher risks than if the slope had already been skied. Untracked snow is but one form of scarcity that may affect our decisions in the back country. Maybe our time is scarce; if we don’t ski this weekend, we won’t have another chance until next season. Or maybe we want to be the first on a route to avoid being showered by rock or ice knocked down by another party. Whatever the motivation, a strong sense that we must seize the moment or lose a valuable opportunity can be a warning sign that it’s time to take a deep breath and ask if we’d do the same thing if that opportunity was secure.
Strategies
Heuristic reasoning may be a built-in feature of human thinking, but that doesn't mean there isn't anything you can do about it. One of the more powerful strategies is good communication and teamwork. Each team member must be empowered to speak up with an observation or concern. Good leaders will actively elicit input from everyone, and all team members must treat each other with respect to avoid shutting down the inexperienced or timid. An explicit process is more likely to expose poor reasoning, and hopefully excise it from the process. When planning a trip, the composition of the group should be considered. Are there any participants with a penchant for overruling others? If so, the group's safety could be compromised.
Another reason to pay attention to your prospective team mates is "risk homeostasis" [6]. The theory is that each individual maintains a more or less constant level of risk in their activities; too little and things get boring, too much and things get scary. Dramatic differences in attitudes towards risk can frustrate attempts to implement group decision-making. If you are more risk-averse than others in your party, it might be time to seek new partners. Risk homeostasis is also a possible explanation for why avalanche training seems to have little effect on incident rates. Armed with information on hazards, groups may actually choose to take more risk, believing that the mitigations (terrain selection, travel techniques, etc.) they put in place will compensate.
Advance trip planning can make a big difference. Colin Zacharias [7] describes the AIARE-recommended approach of formulating an Ideal, Safer, and Safest trip before arriving at the trailhead. The ideal trip is what you would do if all objective indicators pointed to a safe trip. The "safer" option is one that avoids hazards that might be present in your ideal trip. It may be a variation on the ideal trip, or an alternate trip through safer terrain. The "safest" trip might stick to low-angle terrain or simply be to stay home that day. It's important to take pre-trip planning seriously, making sure that your alternate plans are attractive enough to make it easy to accept information that says your ideal trip would be unsafe.
Gaining experience can mitigate the effects of heuristic reasoning [6]. Though research shows that experts still occasionally get into trouble, they tend to approach problems based on an extensive background of similar problems, and thus are able to classify situations as belonging to a known type. They are able to test their hypothesis, which if confirmed, leads to a response that is appropriate to conditions. The challenge to recreational backcountry climbers and skiers is how to acquire this experience without being caught in a slide first. One approach is to take follow-on avalanche training periodically to reinforce good habits and uncover bad ones.
Whether you are a backcountry novice or long-time practitioner, seeking additional training is never a bad idea. Research has shown [3] that people under stress practice what they have physically done rather do as they have been told. Thus, when choosing a course it is important to select one that emphasizes field work and hands-on application of classroom lessons. Instructors who teach concepts versus rules of thumb ("a snowpack can only adjust to additional loading given adequate time" vs. "snow fall greater than an inch an hour leads to instability") reinforce the student's ability to think about the information available rather than apply overly simplistic rules. A good instructor will also model good group decision-making and present participants with scenarios designed to force them to apply what they have learned. This may seem artificial, but it is an important aspect of training. Don't just "take" a course, participate actively [2].
Conclusion
Human factors play a significant, if not dominant, role in avalanche accidents. Indeed, McCammon's results [5] suggest that only 4% of avalanche incidents might truly be called "accidents" insofar as there were no hazard indicators present. That leaves a lot of room for improvement, both for the recreational back country traveller and the avalanche education professional. Some of the tools needed to correct the situation are available in the references provided, and through professional instruction.
Great article, Steve. I emailed the link to my son, who is itching to ski the backcountry powder with his friends, but does not yet have the years of avalanche study experience to draw upon.
You know, it's not fair that the slope, if it's going to slide, doesn't slide under the first rider. It would be nice that if one rider makes it down, then the slope is proven safe.
In regard to the article, it is so true that the group mix and dynamics really alters the decision making process.
Excellent work Steve, Very thought provoking. Looking back on some of my bumbling experiences, my "thought" could well have done with being provoked a long time ago.
Steve Larson - Apr 19, 2006 1:46 am - Hasn't voted
Re: Great Article
Yeah, I know the feeling. It's all stuff I'll never tell my kids about!
Looking back at it now, we probably did take some risk when we crossed that slope. I won't call it excessive and unacceptable though. What do you think?
I compared the situation that day with the seven hazards used in my article, and the only one I would say was clearly present was terrain traps (trees, rocks, small cliffs and a bottleneck in the runout). I recall checking rather carefully as I went across, looking for any sign that the snow was weak, but I didn't find any. If I had it to do all over again I'd probably make the same decision, at least, that would be my input to the group. I do think we did the right thing spending the extra hour to get to the top of the slope before we crossed it.
However, thinking back on the day, I'd say we took a slightly higher risk getting to that point. The west-facing slopes we ascended (above where we dropped the snowshoes and skis) had the additional hazard factor of temps above freezing, and the terrain traps were worse. I'd also probably ascend that slope again as well, as the snow was pretty cold and dry despite the sunshine and warm air temps.
What an excellent article. It's easy to overlook human factors in favor of sexier things like snowpack analysis and weather but at the end of the day, it's the human element that we have the most control over. By the sounds of things, it also has the greatest chance of making a difference too. Thank you for this insightful and thought provoking read.
GREAT article, Steve. Thanks for taking the time to put this one together.
After my, shall we say, heuristically induced avalanche incident last year, I came across this in a bookstore: Deep Survival - Who Lives, Who Dies, and Why by Laurence Gonzales. Amazon Link
People might want to read this in addition to your references. More of a layman's book, but touching on similar subject matter and analyzing several incidents, and the choices and behavior of people involved.
Page Type:
Activities:
Comments
Post a Comment