Why Your Backcountry Observations Matter to Your Local Avalanche Forecast Center

[Utah Avalanche Center]

The Wisdom of Crowds
Saturday, November 10, 2012
Hardesty

Last February, at a TED (Technology, Education, Design) conference in California, a PhD student named Lior Zoref brought an ox onstage to perform a social experiment. He had each person in the crowd write down what (s)he thought the animal weighed and turn it in. While his assistants were tabulating the answers, he then went on to talk about the wisdom of crowds. James Surowiecki posed the same argument in his 2004 book The Wisdom of Crowds: Why the Many Are Smarter Than the Few and How Collective Wisdom Shapes Business, Economies, Societies and Nations. Its central theme is that, quite simply, the collective brain-power of the many is often more accurate than the analysis of the "expert". This is not a new concept. In 1906 the British researcher Francis Galton -who was later knighted - observed the same "phenomenon" at a country fair. The researcher ​Harri Oinas-Kukkonen goes on to submit that

• In some cases, groups are remarkably intelligent and are often smarter than the smartest people in them.
• The three conditions for a group to be intelligent are diversity, independence, and decentralization.
• The best decisions are a product of disagreement and contest. - Productive brain-storming
• Too much communication can make the group as a whole less intelligent. - Sound familiar?
• The right information needs to be delivered to the right people in the right place, at the right time, and in the right way.
• There is no need to chase the expert. - Aha - the avalanche 'expert'

But there's the fine print. It implies that each individual must come up with his or her own answer or opinion independent of what others offer or opine. For example, through social experiments, the researchers Lorenz, Rauhut, Schweitzer, and Helbing^1 found that as testing progressed, the average answers of independent test subjects became more accurate, in keeping with the wisdom-of-crowds phenomenon. Socially influenced test subjects, however, actually became less accurate. In other words, Group Think is not only less accurate, but perhaps dangerous.



And this is where Observations from the backcountry come in. We, the "experts" go in the mountains and come away with our subjective opinion on the state of the snowpack. But it's just one man's (or in Evelyn's case - woman's) opinion. How could we actually be as accurate as the crowd in the backcountry? This is why - at the end of each class, lecture, or presentation, with some tongue-in-cheek - I offer that at the Utah Avalanche Center, our forecasts are always right. But you...you can help us be more right. Submit your obs to the UAC. Or to your local avalanche office. Stick to your convictions - don't be swayed by everyone else's opinions - you might be right after all. And what's more, your info helps save lives.



And the ox? All the independently arrived-at answers averaged out to be 1792 pounds. The actual weight of the farm animal? You guessed it - 1795 lbs.



We look forward to your reports from the backcountry -

Drew Hardesty

"Avalanche Expert", Utah Avalanche Center



^1 - Jan Lorenz, Heiko Rauhut, Frank Schweitzer, and Dirk Helbing. Proceedings of the National Academy of Sciences, Vol. 108 No. 20, May 17, 2011

[GoldenBear]

Bumping because this needs to be read by YOU!

[Shredgar]

Great post & great book reference Drew, thanks.

[Slaag Master]

deep bows to UAC, mr Hardesty, and the rest of the fine crew!!!! for obs not making it to your web site please check this local tr

http://www.tetongravity.com/forums/s...nditions-12-13

again, thank you for taking such deep concern for the safety of so many!!!!!!

[icelanticskier]

WONG FOWUM JONG!!!!

try here, 1st poster: http://www.tetongravity.com/forums/f...The-Slide-Zone

hopefully you get many obz sent in from da sheep pen.

hello to gordon for me, eh?

rog

[MakersTeleMark]

Just put your self help books down, re-fund WOW, and fire half your staff. His approach, analysis, drive, and presentation is bar none. So put up with a little BS and get a lot in return. Including, I would bet, a few saved lives.

[~mikey b]

plus 3 for WOW

Bob is just about the best avalanche un-professional I know

I can only hope to be so un-professional

[jfost]

fine for guessing the weight of Oxen, but when the average of safe and unsafe is taken, someone still dies.

better than 2 I suppose, but you need both extremes as well as other middlemen to get the best answer in a group!

[gpcharles]

The question of whether the collective is wiser than the "distinguished expert" is not new, and despite what this talk implies, is far from fact. There's lots of research that's "proven" both sides of the argument. The biggest trap is (as noted) groupthink, and the importance of independence in the individuals' thought processes and observations etcetera. Problem is, groupthink is almost impossible to avoid. Even when you tell people to avoid it.

My thinking is simple: BE CAREFUL on this one UAC et. al. It scares me when the belief that brainstorming > individual-expertise is held for something so technical and mission critical as avalanche safety.

For a nice summary of the argument for why it may be that individual-expertise > brainstorming see here:

http://www.newyorker.com/reporting/2...fa_fact_lehrer

[Slaag Master]

the folks that help save lives in UTAH - the uac - simply asked LOCALS to provide them with obs from the backcountry. let's not over think this or anal lies it to death!!!!!!!!

[icelanticskier]

the folks that help save lives in UTAH - the uac - simply asked LOCALS to provide them with obs from the backcountry. let's not over think this or anal lies it to death!!!!!!!!

this place is slipping............big time.

rog

[bestskieratsnowbird]

this place is slipping............big time.

rog

that's fucking hilarious coming from you.

[gpcharles]

the folks that help save lives in UTAH - the uac - simply asked LOCALS to provide them with obs from the backcountry. let's not over think this or anal lies it to death!!!!!!!!

fair enough. i'm just genuinely interested in the debate about experts vs. crowd-sourcing in this and other realms.

it's really quite fascinating to think about here. used wisely, i bet it can be pretty good for the UAC. just have to be careful.

[wstdeep]

the folks that help save lives in UTAH - the uac - simply asked LOCALS to provide them with obs from the backcountry. let's not over think this or anal lies it to death!!!!!!!!

^^^ +1

adding: collective wisdom does have value...say this was an avalanche forecasting scenario AND the guess results of the collective were known beforehand (the forcast)...the people who were way off the mark have a chance to re-evaluate before they get onto the snow.

Don't know WOW personally but I like reading his stuff, that said it just doesn't seem like the place here to be firing shots at the UAC re: him and their relationship.

[gpcharles]

if everyone doing the observing is scientific in their observations, and working independently, crowd >>>>> expert (read: crowd WAY BETTER THAN expert).

the problem is if a group of people don't behave as above, and instead develop a groupthink that the stability is OK. perhaps also this groupthink is confirmed by stretch of a zero of accidents (for instance, maybe all the stability concerns are deep hard-to-trigger instabilities, and which are hard for non-experts to fully evaluate). In that case, we end up with a large body of bad data, and hence bad decision making.

[icelanticskier]

the problem is a group

solo, the way to go to stay alive in avy terrain. my ride? not solo. WOW's ride? not solo. how many folks do you know that are dead or still with us that were going solo and caught vs in a group? something to think about..........

rog

[Swine]

But there's the fine print. It implies that each individual must come up with his or her own answer or opinion independent of what others offer or opine. For example, through social experiments, the researchers Lorenz, Rauhut, Schweitzer, and Helbing^1 found that as testing progressed, the average answers of independent test subjects became more accurate, in keeping with the wisdom-of-crowds phenomenon. Socially influenced test subjects, however, actually became less accurate. In other words, Group Think is not only less accurate, but perhaps dangerous.

Groupthink = Bad. Aggregating data from informed individuals = good. This isn't that hard.

[The Gnarwhale]

Here's the page of observations from this year: http://utahavalanchecenter.org/new-avy-content

In an organization that attempts to cover 100's of miles of mountains with a team of 7 or 8 forecasters, what's the alternative to not crowd sourcing? Either find $$$ to hire a few dozen more forecasters, restrict forecasts to the small areas where one has high confidence in one's assessments, or do like Canada and publish vague conjectures that cover areas bigger than some NE states based on sparse reports from heli operators?

In the avy forecasting world, the UAC has become one of the most envied because they have so much current information to work with. They read each obs submissions before it is posted to the public. Presumably they filter out a few that are just wack. On top of that, many of the observerions come from guides, ski patrol, etc who're trained in the snowgeek ways.

I submit an obs when a) I see something notable, like a fresh avalanche or b) go skiing someplace outside of PC/Mill/Big/Little even if the snow conditions there are unremarkable.

[PNWbrit]

something to think about..........

rog

Rog relies on the windbag.

[icelanticskier]

Rog relies on the windbag.

yer wrong brit, as is per usual. rog relies on witchcraft.........voodoo........

rog

[icelanticskier]

This isn't that hard.

neither is reading these daily whether you plan on touring that day or not.

http://utahavalanchecenter.org/advis...lt-lake#simple

talk about taking a lot of the guesswork out. they even tell you where the good snow is, how generous is that!?

rog

[wra]

Here to post in this bit of nonsense.

solo, the way to go to stay alive in avy terrain. my ride? not solo. WOW's ride? not solo. how many folks do you know that are dead or still with us that were going solo and caught vs in a group? something to think about..........rog

Ooops, trolled by an expert.

[Slaag Master]

i'm confused WRA. so you broke the rules got caught got injured and your advise is : "Seldom view the danger report or any of the crowd info, no interest. My limited info is free and available if you were to look. I'd also call Alta central if there was a need. A person should learn how to think for themselves and forget the crowd, they have no wisdom."

maybe i misunderstood something. otherwise i am not buying!!!

[Swine]

An example which I still dream about occasionally. Almost killed me and I was alone....
Seldom view the danger report or any of the crowd info, no interest.
My limited info is free and available if you were to look. I'd also call Alta central if there was a need.
A person should learn how to think for themselves and forget the crowd, they have no wisdom.
Hope to be back ski bumming shortly.

So you go alone, and you don't pay attention to expert reports and data? This reads like fucking gibberish. And your "limited" info is free... amazing groundwork. Or did I just get trolled... jesus?

And rog, are you arguing that going solo is statistically advantageous to being in a group? Back that shit up if so.

[icelanticskier]

And rog, are you arguing that going solo is statistically advantageous to being in a group? Back that shit up if so.

going solo, imo, leaves out that extra bit of comfort and tendency for complacency that can often be more prevalent when with others in the bc. when you go solo in avalanche terrain, you are much more likely to err on the conservative side vs being in a group. now i don't have the stats, but i'm pretty sure that if you read most reports of any accidents/fatalities, they more often than not involve groups.

for someone that tours solo more often than not, and having been caught while not solo, would i have been caught that particular day if i was solo? i think not when i think of where my head was at that day.

made me think of wow cuz that's someone i'd often see solo and sometimes with others. when he got caught he was not solo and would probably not have been caught if he was as his partner dropped in behind him while wow was skiing partway down the slope when slope let go. obviously poor/no communication, but that's group dynamics at work.

two good friends and touring partners of mine, an avy forecaster here and another out west, who also both tour solo often have both taken big rides, but only when with others.

yes, i think going solo is advantageous as you are more likely to keep things scaled back and oftentimes one focussed clearly thinking brain is better than multiple distracting brains getting wires crossed.

rog