Posts Tagged ‘risk’
“As far as the laws of mathematics refer to reality, they are not certain; and as far as they are certain, they do not refer to reality.”
The distinction between risk and uncertainty arose from the field of economics and is based on the work of Frank Knight. According to Knight, “risk” refers to a situation in which the probability of an outcome is known or can be roughly determined, while uncertainty refers to an event or an outcome whose probability is not or cannot be known. A common example used to illustrate the point is that games of chance are risky (because the odds of winning vs. losing can be calculated) and the outcome of a war with its multitude of changing environmental situations on the ground is uncertain. But it is not as simple as all that, it certainly can be very confusing, and a deeper understanding of risk vs. uncertainty can help people make better decisions.
People board airplanes routinely, strapping themselves into what is essentially a lounge chair (even if it is an uncomfortable one), inside of what is little more than a controlled missile, whose paper thin walls of plastic and metal guard you against external conditions that could not possibly sustain life, while jet engines furiously burn enormous quantities of highly explosive fuel within a few feet of your location, hurtling you and your lounge chair through the sky at hundreds of miles per hour, with thousands of pounds of airplane coming back to the earth with a typically ungraceful thump, on small rubber wheels (which always look flat to me), and then hoping that the plane will somehow slow down and avoid plunging into the water (I land at LaGuardia), or worse. Why in the world would they do that? Because it is not all that risky. Airplanes have very good track records and there are very few accidents. We can manage risks, uncertainty is more difficult.
But what about Wilber and Orville when they first attempted flight? They were not dealing with risk they were dealing with uncertainty, for there was little real understanding of whether man could build airplanes that could stay aloft for a period and then safety return the occupant to the ground. There was no track record to calculate risk upon, there were no computers to run simulations, there were no wind tunnels which could test airplane models. Yet facing this uncertainty they persevered. Events can begin with uncertainty and then as track records about them build they can become simply risky. Though, some people who treat a situation as risky, when it is actually uncertain, can accumulate really awful track records of performance.
For instance, Gerd Gigerenzer of the Max Planck Institute analyzed the performance of 22 major international banks on predicting currency fluctuations. These annual forecasts of currency values which occurred from December of 2001 to December of 2010 were used by the banks to guide their investment decisions. These annual forecasts were wrong, very wrong, for nine out of those ten years. Gred’s conclusion about the track record of the people who produced these annual guides to currency values was that “highly paid people produced worthless predictions.” He went on to explain that based on his analysis the risk modelers at the banks didn’t distinguish appropriately between risk and uncertainty. They treated the currency fluctuations as risky but in fact “it is uncertainty that rules in the real world, where risks can’t be known in advance because of a complex tangle of factors triggers new, extremely unlikely hazards.” What he meant was that many factors that could affect currency values, (e.g. oil shortages, war, weather, natural disasters, deepening recession) were not adequately accounted for in the prediction model and in fact could not be known as they were uncertain. The analysts at the banks though treated them like risks, however, with underlying probability distributions and they got it wrong.
When Steven Jobs, who famously stated that “people don’t know what they want until you show it to them”, produced his first computer he had no idea if people would be able to see its promise, what they could accomplish with it and whether they would buy it. Over the years with each new technology his team at Apple developed there was uncertainty, sometimes uncertainty with great consequence, as some of the products rolled out were “bet the company” kinds of decisions. Steven Jobs reveled in the world of uncertainty and showed that mastering the world of uncertainty can lead to enormous financial reward. But Apple’s ultimate financial success came as those uncertain new technologies became simply risky products. Was there any doubt in anyone’s mind if the iPhone 5 would be a success? The question was not “if”, the question was “how big”. “If” is uncertain, “how big” is risky.
There are hoards of managers out there who want to emulate Steve Jobs, or at least his success or even more precisely his financial success. They look at his management style, which at times bordered on abusive and wonder if that is the path towards their own success. Perhaps if you beat up your employees, driving them really hard, you and your company can also succeed and become like Apple. The scientific literature casts doubt on that approach (big time) as working for the majority of managers (from a risk management perspective). I do know of a few very successful CEO’s whose success could only be described as coming off the backs of their employees rather than through or with their employees. Yet Steven Jobs, with his style, and his ability to deal successfully with uncertainty, like the Wright Brothers, was able to build the most financially valued organization in the world.
In the retail world (and real estate in general) there is an expression, “location, location, location”, meaning that without this fundamental element in place, it simply does not matter if you have great merchandise. You will not be successful. And with Apple it is technology, technology, technology or perhaps product, product, product. The success of Apple should not be attributed to an abusive management style, but rather to Steve Job’s genius in developing technologies and products with an uncertain outcome and turning them into mere risks. Arguments about his management style could be viewed as a red herring – he was successful in spite of it, because of his overwhelming other abilities, not because of it.
If I need to hire 100 people for my sales force and I have developed a predictive analytics approach to helping me select the best 100 out of my applicant pool of 1000, I can determine the likelihood of expected performance outcomes across those 100 new hires by creating a probability distribution. The distribution of job performance across 100 sales people is something that can be known and so my hiring decision can be described as risky, not uncertain. But if I want to know and predict how a specific sales person will perform, that is more uncertain. I can create a probability score for that individual, but I cannot say with certainty what the performance outcome will be as external factors (e.g. a death in the family, a pregnancy, a spouse relocating, a divorce or marriage, a new educational degree, or a new opportunity) cannot all adequately be accounted for. One thing to keep in mind, as the quote by Einstein above describes, is that our models are representations of reality, accounting for only a portion of the variance, they are not reality.
Yesterday, a horrendous crime was committed in Newtown, CT where a 20 year old gunman shot and killed 28 people, 20 of them school children between the ages of six and seven. This was terribly disturbing and I had great trouble concentrating on anything else after I heard the news. As I saw the images of the parents finding out about their children I could feel my heart ache for them. Our own local school, about an hour away from Newton went to a heighten security status. I am concerned that there will be the usual hand-wringing about firearms and second amendment rights and then nothing will be done. This time is has to be different. Our children are dying.
From an uncertainty standpoint it would be very difficult to determine if any one individual is a risk of committing a gun related crime. As with the sales force example above a probability score can be created, but at the individual level what you are dealing with is closer to uncertainty than risk. Common-sense gun laws would suggest background screening and eliminating those with various mental illnesses and track records of violence or abuse from gun ownership.
But beyond that if you look at the methods that can be used to manage risk, and at the likelihood of gun violence, more guns simply provide more opportunity for guns to be used in gun violence. It is a simple relationship. Red herrings are constantly thrown up about arming people to stop the perpetrators of gun violence, as though if we simply have more bullets flying around that fewer people will get injured or killed. Very unlikely. The arithmetic simply does not add up. From a risk management standpoint fewer guns mean fewer gun crimes. Period. End of sentence. Trying to create the odd one-off scenarios whereby having the right person in the right place with the right weapon and the sensibilities to stop a crime in progress without creating further injury to other by-standers is just not logical.
A first step would be to ban the type of firearms that allow for mass-murder to happen within a few seconds without reloading. Our ultimate goal from a risk management standpoint should be to reduce the number of guns available. Period. Given the difficulty of dealing with uncertainty, you cannot accurately reduce the number of guns available to only those who will commit gun violence, you will get it wrong. So the solution must be one that works with the probability distribution. Ultimately, we must reduce the overall number of guns that are floating around in our society. To paraphrase a quote about eating an elephant – how do you remove 300,000,000+ guns from our society? One at a time.
© 2012 by Jeffrey M. Saltzman. All rights reserved.
Visit OV: www.orgvitality.com