Blog: Drift into Failure...or, Mathematics and a Few Thoughts on Risk

Wednesday, December 31, 2014
Drew Hardesty

What is your Level of Acceptable Risk?  How did you determine this?  Some will center-punch Superior on a CONSIDERABLE danger while others feel happy going to Powder Park everyday.  Everyone is different and it's insulting at worst and a waste of time at best to look askance at others who are on either end of the pendulum. It's - how shall we say - inelegant to look upon some as suicidal and others as boring and unfulfilled.  

The key points here are 

  1. To be aware of your level of acceptable risk
  2. Understand factors that may influence your risk taking
  3. Find others who have a similar level of risk acceptance  

Ok, now the math  -

Essentially -      RISK = Avalanche Danger times Mountain Terrain        R=DxT

Here are the variables: D - Avalanche Danger/Danger Scale  (Low, Moderate, Considerable, High, Extreme)

                                     T - Mountain Terrain (Avalanche Terrain Exposure ie: Powder Park or Superior)

Here is the constant:    R - RISK 

KEY POINT:  Now what this pre-supposes is that you want to keep your RISK at a constant level.  If that's the case, then D (Danger) and T (Mountain Terrain) must have an inverse relationship to one another.  When the danger goes up, you scale down your terrain choices.  When the danger goes down, perhaps you 'up' your terrain choices - otherwise, you're taking less risk.  And that's ok too.  

Easy, huh?  Well, not so fast.   It may be news to you that humans are complicated people.  

As it turns out, most of us feel like our decision-making looks like what you see on the left.  What actually happens is on the right.  

The author Sidney Dekker, in his book Drift into Failure, posits that a "series of small relaxations on safety standards can lead to a catastrophic system failure.  These systems drift and that from inside the system, such drift isn’t visible until the failure occurs."  Please read that sentence again - I had to.  

So What?  How does this apply to us in the avalanche world, particularly when it gets back to Math and especially when it applies in our current Low Probability/High Consequence persistent slab problem?  I'm glad you asked.  

Here's what happens - 

  2. We choose appropriate terrain.
  3. We see no signs of avalanches.
  4. We see no shooting cracks.  We hear or feel no collapsing or whumphing. 
  5. Test results are inconclusive.
  6. Cornice drops and slope cuts provide no results. 
  7. We see other tracks on the slope(s).

Now what?

  1. We "drift" into steeper terrain because we perceive the danger is lower than it actually is.

Result?  You guessed it.  

These are when conditions are the most tricky and dangerous.  In reality, the danger remained the same, you've 'upped' your terrain choices...and therefore you've upped your RISK.  

It's common during persistent slab, deep slab, or hard wind slab conditions.  It's just part of the waiting game - and it takes discipline.  One might argue that these are times when we need to assume a "Ulysses contract" (common in medical and other situations).   You'll remember that Ulysses tied himself to the mast and had his men put wax in their ears to avoid the lure of the Sireens (as Delmar/Tim Blake Nelson called them in O Brother Where Art Thou).  In the end, it didn't work and we know what fate befell them (at least in the Coen brothers film).  Be patient!

Happy New Year - I wish everyone a happy, healthy and fulfilling 2015



"If snowpack is the question, than terrain is the answer"--??

I think it is important to note that by following the tracks (skin and ski) of others you may be getting lulled into assuming their risk tolerance which may be higher than yours.

Not surprisingly, this rings quite loud to me today.

I think the math equation aught to address testosterone. R=2TxD.

When you talk about Bruce and the Ulyssees contract, I assume that you mean you decide before you venture out that you will only access certain types of terrain. Am I correct in Thinking about this in terms of red light vs green light terrain? If so, isn't one of the issues with this that you need to trust in the Avalanche center? In Utah I believe in you guys, but here in Oregon, NWAC is so far from the Mt Hood zone that the forecast is chronically exaggerated/conservative. If I didn't use my judgment I'd never ski anything soft here. Still, I cant ski often so I have to look to them for my pre-trip planning. Any ideas on how to make a reasonable Ulyssees pact with less pre-processed information? Also, Happy New Years and thanks for all of your great work! Cheers!

You talking about me Bruce? I am the guy who is always in Powder Park.

"The Edge... there is no honest way to explain it because the only people who really know where it is are the ones who have gone over." Hunter S Thompson

The "Sireens" are certainly calling these days. The high pay-off of super nice powder conditions is also certainly a factor. I'm much more likely to commit to a potentially dangerous slope in good pow conditions, because I can anticipate to pay-off (face shots). Even understanding the risk, the luring call of cold smoke is so loud, it's perhaps impossible for us mortals to avoid unless we're tied to the mast or deafened by wax in the ears.

On the other hand, it's easy for me to opt out of even a nice run if conditions aren't ideal. It seems most insane to take serious risks for a lower pay-off, like skiing a steep slope in dangerous and crusty conditions for example....

This math doesn't help me except to describe the retreat to lower angle terrain when the conditions are unstable. There IS however, other math that DOES change how I think about avalanche risk. I used to try to evaluate if I thought a slope would slide or not before deciding to ski it - yes or no? Today I think instead of the PROBABILITY that a slope could slide. What if you were 95% certain a slope would not slide, and so decided to ski it? That sounds very conservative. If you skied it, there is a 95% chance you'd have a great run and be fine. But what if you skied it 20 times? 40? 100? A low-probability 5% chance happens all the time to people that roll the dice a lot. Skiing 100 days per year in the BC means that even if I was 99% sure every slope I skied was safe I'd probably be caught in a slide every single year. This should make you think. Or at least it made ME think. And it made me choose to be more conservative. Now I can't count the number of times I've walked away from a good looking slope even though I felt it was PROBABLY safe to ski. The funny thing is when someone else then goes and skis it and it doesn't slide - that person thinks they were right in judging the slope, and I was wrong or too conservative. What's funny is that I don't care if they skied it 9 times in a row with no releases - I KNEW it probably wouldn't slide. But I wanted to know what would happen if you rolled those dice 20 times in a row, because I hope to do it a hundred times this year. Of course, over a span of 23 years in the Utah and Colorado backcountry, you see a lot of slides. Strangers die. Friends die (RIP Jim Jack, Basti Haag). I've barely skied out of a couple that almost certainly would have done me in. So maybe it's not some mathematical analysis that made me more conservative, just the experience of a guy who likes to do stupid stuff! Or maybe I'm just getting old and chicken-shit!! I bought an airbag. And my new Vapor Nanos mean I can crank fast turns and catch air on 28 degree slopes. When I was in Utah it was clear the mountains didn't care about you; you had to watch out for yourself. But here in Colorado they DO care about you - and they want to kill you right now all year long!! (Or at least they've demonstrated they want to kill ME). So I keep a mental list of clues as to how badly the mountain wants to kill me every time I'm out there (recent slide reports, collapsing, cracking, loading, hasty pit results, etc). But mostly I keep in mind that I'm a big fat idiot and I'll NEVER know for sure, so I try to strategize for rolling the dice over and over and over again. That's the math that's helping me make better decisions out there.

It's an illogical argument. When you throw powder lust into the equation, all logic fails. Or so it seems these days. Just get a go bro camera and charge it, brah. **sarcasm thick** Young, dumb and full of..... luck.

I like what you have to say about the perception of risk and how people can be lulled into complacency, but I have to point out that your graph of what people think their decision making is like makes no sense at all. You say that if risk is to remain constant, "...then D (Danger) and T (Mountain Terrain) must have an inverse relationship to one another." This makes some sense, however the graph on the left shows exactly the opposite relationship. In your graph the value of X increases as the value of Y increases. This is a direct relationship, not and inverse relationship. If you were to graph an inverse relationship the line would start high on the Y-axis and slope down towards the the X-axis as you move further to the right. In other words, as the value of Y increases the value of X decreases and vise-versa.

I like Chris Cawleys take, doesn't mean you can't ski steeper terrain, maybe try a slope that has already avalanched or at least a slope with a clean run out, void of rocks and trees in it's path, trauma kills most avalanche victims, especially in a thin snow pack.

Agreed with Chris and Mark, terrain is the answer. But sometimes safe terrain means the Nordic track. Clean run out? That's not an acceptable risk for me. I'm pretty sure I could get trundled hard enough to (at least) end my season even in a "clean" run out. Hard slabs are hard.

This video will haunt me forever. How easy is it to ski onto a slope that has already been skied. The answer: really easy.