March 11, 2004 Avalanche in Elderberry Cyn on Mt. Tom
Steve Burnham Photo

In the wake of two avalanche-caused fatalities occurring on Mount Tom on March 26, 2005, as well as a serious avalanche-caused injury to a third victim earlier this winter, the destructive power of even the most familiar slopes and ski runs of the Eastern Sierra has become heart-breakingly obvious.

Most research and training regarding avalanche safety speaks to the technical aspects of avalanches - the how's, why's and what's of snow-loading, crown depth, slab formation, slope, base layer, beacons, shovels and the like. Less commonly studied but just as critical, are the human factors in avalanche accidents - the how's, why's and what's that go into every decision that every person must make before he or she enters any slope that could possibly avalanche.

In recent years, studies of human decision-making patterns regarding avalanche accidents have gained credibility. Researchers have found that as much as 93 percent of every avalanche accident studied in the United States has, at its roots, socially-motivated decision-making errors made by the people involved, rather than occurring due to unpredictable subtleties of terrain or snow-pack.

The most prominent and accepted study on human decision-making trends regarding avalanche safety, done by Ian McCammon with the National Outdoor Leadership School (KNOLS), looked at over 600 different avalanche incidents that included human involvement and human victims (although not all victims were fatalities). The accidents occurred in the United States between 1972 and 2001, and included victims of every degree of avalanche safety expertise.

What McCammon found may surprise those schooled to think of avalanche safety as a function mostly of increased technical knowledge. He found that six common decision-making assumptions were consistently used by individuals and groups when assessing avalanche risk. According to McCammon, using one or a combination of these six assumptions was often the main reason that recreational winter trips turned tragic.

It is also useful to know that in the vast majority of the incidents McCammon looked at, there were several very clear hazards present before the party started out, such as a clear avalanche path (in 82 percent of the cases), recent snow loading (66 percent of the cases), extreme or considerable hazard postings prior to the trip (55 percent), among others.

Adhering to the six assumptions listed below caused people to make riskier decisions, so McCammon called them decision-making "traps." These decision-making traps were consistent, he found, across a wide variety of factors, such as gender, age, experience levels, and terrain type. When McCammon adjusted for all possible mitigating factors, the common trends that he found that contributed to poor decision-making in winter recreational outings are as follows:

1) The familiarity trap

McCammon found that familiarity with the terrain increased the probability that a recreational trip would involve an avalanche accident. In addition, the more highly trained the individuals in the party were in these cases, the more apt they were to be in an avalanche accident.

"The tendency of highly trained accident victims to make risky decisions in familiar terrain is disturbing," McCammon wrote. "While this group seemed capable of recognizing and avoiding obvious avalanche hazards, it appeared to do so only when traveling in unfamiliar terrain. In familiar terrain, this group seemed to suspend its ability to heed obvious warning signs, and this exposed party members to significantly more risk." McCammon said that the experienced group was up to four times more likely to be involved in an accident in familiar terrain as in unfamiliar terrain.

McCammon said that, in the end, despite their advanced training, the accident rates for trained groups were about equal to the accident rates for untrained groups, mostly due to this "familiarity trap."

2) The commitment trap

Another decision-making trap that McCammon noticed was that groups that were highly committed to a certain objective, say due to approaching dark, due to previous bad weather that had made a trip impossible up to that point, or other constraints, were more likely to become part of an avalanche accident than groups that were more flexible about where and when they were going. This trend held true, even in the presence of high avalanche danger warnings or other high risk factors.

3) Gender acceptance trap

McCammon also discovered that avalanche victims consistently played into what he calls the acceptance trap - a tendency to do things that others in the groups might find courageous or daring, in order to increase their social acceptance. He said that this tendency swayed accident victims most distinctly when two genders were present in the party. He said that mixed-gender groups consistently were involved in more accidents than single-sex groups. McCammon noted that the increase in risky behavior was not due to the trained women in the groups exposing themselves to known hazards, but to what he states, in much more technical terms, as male group members showing off, thereby exposing the entire group to more risky conditions.

4) The "expert halo" trap

The expert halo trap occurs when a group picks a leader, who may not be more highly technically trained for avalanche safety than the rest of the group. The bottom line, according to McCammon, is that most groups eventually settle on some type of leader, if only because that person might have just a bit more technical skill than the rest of the group or, say, because of other leadership qualities that may or may not be related to technical avalanche-safety training.

The expert halo trap resulted in significantly more accidents per group than in the groups that did not have any identified leader but, instead, achieved their goals via consensus. Again, this trend occurred regardless of the technical abilities of the leader. The larger the group, the more risky the conditions the leader exposed the group to, which, according to McCammon, lends credence to previous studies done on group dynamics that show people tend to increasingly conform to one position as group size increases.

5) The social facilitation trap

Meeting other people near or at the avalanche site increased the tendency of highly trained groups to expose themselves to risky conditions, McCammon found. On the other hand, lesser-trained groups seemed to decrease their risky behavior when they encountered other groups. McCammon said that this might be attributed to the tendency of people who are already good at something to do it even better when others are watching, and the corresponding tendency of people who are not "good" at an activity to try to minimize others from noticing the their lack of knowledge.

6) The scarcity trap

People took significantly more risks in unmarked slopes and terrain than in terrain already visited by other people, McCammon said. The scarcer such terrain might be, quite often the more apt it is to avalanche. In other words, right after a storm is when a slope is most untrammeled, and thus most perceived as being scarce, or rare, which then drives people to take more risks as they rush to make first tracks. At the same time, due to the natural laws governing snow behavior, this is exactly when the terrain is the most prone to avalanching.

To summarize the above six decision-making traps, McCammon states, there is "a disquieting learning curve amongst avalanche victims. It seems that social cues play an important role in novice decision making, while those with more expertise plan far too much on familiar terrain and the presence of other people" and their own ability to mitigate dangerous situations.

In addition to the above traps, McCammon also found that group size was a significant factor in determining a group's exposure to risk factors. The larger the group, he states, especially if the group is larger than three to four people, the more of the above traps the group is likely to engage in.

McCammon also acknowledged the limits of his study, saying that the study identifies correlation, not causation, and that because the study is only done on accidents, it is impossible to study other situations where there were no accidents.

"Nonetheless," he states, "one trend is certain: the majority of avalanche accidents happened when the hazard would have been obvious even to someone with minimal avalanche training." Meaning something other than logical risk analysis is going on every time people make decisions regarding avalanche safety, Indeed, if McCammon's analysis of human behavior is accurate, in every potentially risky situation that people take part in.