5 December 2003

Hold sandsynlighed adskilt fra konsekvens!

Nedenstående historie er taget fra Fast Company og er, synes jeg, et godt eksempel på, hvorfor det kan være livsvigtigt at opveje sandsynligheder med konsekvenser når man skal tage en beslutning.

Separate Probability From Consequence: How to Succeed--and Stay Alive--by Understanding the True Risks
In the summer of 1975, a young climber named David Breashears set his sights on a beautiful, unclimbed sheet of rock rising from the ground on a cliff south of Boulder. For years, no rock climber had given serious thought to climbing this section. The challenge lay not in the apparent difficulty of the climb, but in the absence of natural protection. Breashears saw no cracks where he might slot wired climbing nuts, and he was climbing in an era before it became acceptable to drill expansion bolts directly into the rock for protection. The wall rose vertically for about five stories, with little pebbles and sharp edges, then the angle kicked back to 85 degrees with holds that looked larger.

Breashears headed up the route, trailing a rope and carrying a small selection of wired nuts that he hoped to slot into one of the upper pockets after the hard climbing. At the 50-foot mark, he had a horrifying realization: the climb ahead would be more difficult than the opening moves, and there were still no places to slot nuts. The rock became water-polished from thousands of years of runoff from above, and the sloping handholds had no sharp edges to grip. If he fell, he would plummet 60 feet straight down onto the jumble of boulders strewn at the base. At 32 feet per second squared, he would slam into the boulder field at nearly 50 miles per hour at a force of 20 Gs. Ka-smack! One dead climber.

Was this a risky situation?

Well, it depends on what you mean by "risky."

For David Breashears, it was not a risky situation. Sure, the consequences of a fall were severe, but the probabilities of a fall were close to zero. David was such a gifted climber in his prime, that--to him--the route formed a puzzle to solve, but not a particularly difficult one. It would be like handing a world-class crossword-puzzle expert the Wednesday New York Times crossword (a challenging puzzle, but well within her capabilities) with the following instructions: If you don't get the puzzle right, we're going to drop you off a 60-foot cliff to your death.

If Breashears had allowed the 60-foot ground-fall potential to infect his brain, he might have died. But he didn't. He was able to separate the probabilities of falling from the consequences of falling, and he climbed with focused precision to the top, establishing a new route aptly named Perilous Journey.

Separating probability from consequence applies not just to climbing but also to work, life, and business. In 1994, when Intel Corp. first discovered the floating decimal-point flaw in its Pentium microprocessor, engineers estimated it would cause a rounding error in division only once every 27,000 years for the average spreadsheet user. This infinitesimally small probability blinded Intel's leaders from worrying about the astronomically high consequences on the other side of the coin. When that one-in-a-billion event happened to a math professor, it ignited an explosion of Internet chat, which, in turn, caught the attention of the media. As then-Intel CEO Andy Grove described in his book, Only the Paranoid Survive (Random House, 1999)--a good title for climbers, by the way--the company found itself hounded by CNN, pilloried in the press, and jolted by unhappy customers. On December 12, 1994, Grove awoke to read the horrific headline, "IBM Stops All Shipments of Pentium-Based Computers." Ultimately, Intel took a $475 million write-off--an amount equal to half a year's R&D budget, or five years of Pentium advertising spending.

While Intel didn't die from the fall, it certainly crashed onto a ledge and shattered its leg. To Intel's credit, it changed its way of doing business to account for the consequences, not just the probabilities. To date, we have not seen another problematic Pentium event from Intel.

Jim Collins er amerikansk professor og bl.a. forfatter til bøgerne "Built to Last" og "Good to great" hvor han beskriver resultaterne af en stor undersøgelse af, hvad der adskiller virksomheder med fra de rigtigt succesfulde. Han beskæftiger sig i det hele taget med "greatness-faktoren" - hvad gør at nogle virksomheder klarer sig bedre (økonomisk) end andre? Collins er aktiv klatrer i sin fritid.