As I See It: Sweating the Little Stuff

Meteorologist Edward Lorenz was in a hurry. The year was 1961 and he was running weather simulations. Lorenz plugged equations into his computer and waited. And waited. He was working on a Royal (as in typewriter) McBee LGP30, a 740-pound “portable” desk system, and the going was slow. Today, of course, computers crunch numbers with the efficiency of an elephant stomping on a peanut, but in the 1960s, computers crunched numbers with all the vigor of a squirrel gnawing on a coconut. So when he decided to run one last simulation, he took the numeric results from the middle of a prior run, rounded off the values in order to hasten the calculation process, and restarted the simulation assuming there would be a slight but negligible impact on the results.

Boy was he wrong.

Instead of the trifling changes he expected, Lorenz saw wild fluctuations. Seemingly inconsequential changes had produced vastly dissimilar outcomes. In retrospect, you could say the results were positively chaotic, because that was exactly what Lorenz had blundered upon: one of the governing principles of what later became chaos theory.

In succeeding years, as chaos theory was being developed, science-speak was assigned to Lorenz’ discovery, the essence of which was that systems have an acute sensitivity to initial conditions. Chaos afficionados call these conditions “attractors,” presumably because they attract future outcomes. Lorenz himself chose a more memorable and poetic name, calling his discovery the “butterfly effect.” In the context of weather system formation, Lorenz discovered that a butterfly flapping its wings in Brazil can cause (or prevent) a tornado in Texas.

It’s curious that W. Edwards Demming, the statistician and father of the quality movement, doesn’t get credit for contributing to chaos theory, because he instinctively understood how initial conditions profoundly effect systemic outcomes. Demming began the quality revolution in Japan shortly after the Second World War. More than 10 years before Lorenz made his discovery, and 25 years before mathematician James Yorke actually coined the term “chaos,” Demming was applying chaos theory principles to manufacturing. One of his fundamental beliefs was that it made little sense to have quality inspections after an item comes off the assembly line. Build quality into the front-end of the manufacturing process, and there is no need for inspection and rework when it’s too late and too expensive to correct defects. In effect, Demming changed the attractors (the initial conditions) in the manufacturing system and, by doing so, helped Japanese auto makers produce generations of vastly more reliable cars.

If natural systems such as the weather and manmade systems such as manufacturing lines react radically to changes in initial conditions, the same is likely true for workplace systems and individuals.

In IT, the attractor principle can most clearly be seen during the requirements phase of a software development project. If specifications are incomplete or rushed and inaccurate, a predictable set of one or more outcomes is almost certain to follow. Users will have a dysfunctional application, and they will be dissatisfied. As a result, there may be inter-departmental shakeups, and the project leader or responsible manager may be fired. From a chaos theory perspective, specifications are one of the attractors that influence future outcomes in the IT department; and these outcomes are acutely sensitive to initial conditions. As Lorenz discovered, even minor variances can have huge impacts. A slight coding error in an accounting program, for example, can produce criminally inaccurate financial statements; an error in a voting machine’s operating system can elect the wrong candidate. And, about a decade ago, a small flaw in the auto-pilot software of a commercial airliner cut power to the engines when the plane dropped below a certain altitude, preventing the pilot from aborting the landing, and the plane crashed.

Likewise, we see chaos at work in system security. The tiniest security flaws allow hackers access to private systems. When those flaws exist in software that is widely distributed, the full impacts are so broad as to be unknowable.

One of the most powerful attractors for individuals is the ability to tell the truth. The most blatant and tragic example is the distortion of pre-war intelligence which resulted in regional and possibly global consequences that will span generations. But initial conditions can also be influenced by omission. In the wake of the 1986 Challenger disaster that killed seven NASA crew members, it was discovered that engineers had concerns about the defective o-ring that caused the explosion, but no one spoke up. No one wanted to be responsible for postponing the flight. Thus, the seemingly inconsequential choice to keep silent produced a horrific departure from anticipated results.

It would be fascinating to know how many ill-fated IT projects could have been avoided, how many billions of dollars could have been saved, if someone who knew better spoke up to say: “This will never work.”

For that matter, what are the unanticipated results of providing inadequate training? What outcomes do incompetent employees attract? What will the outcome be of having 45 million people without health insurance? Or workers without pensions? The butterfly is flapping its wings, but by the time the tornado forms it’s too late to change the attractors.

I recently wondered if having a bad day was the result of initial conditions. The likelihood of having “one of those days” seems to increase in proportion to how early the mishaps begin. If I spill my morning coffee, I seem more likely to cut myself shaving, get stuck in traffic, and have a fight with my wife. Bad days have a way of snowballing. Conversely, if I take delight in the sunrise, and feel joy in being alive, it is less likely that anything will throw me off stride, and much more likely that a parking space will appear precisely when I need one. So, although I can’t prove it, perhaps my mood attracts more of what created my mood to begin with.

What seems to be true is that there are conscious and unconscious attractors. While we can’t control all of the variables that contribute to future outcomes, we can apparently influence them by making seemingly inconsequential changes in initial conditions. This has broad implications for change and problem resolution–both personal and professional. If small course corrections lead to very different destinations, making the right choices initially may do more to attract the future we want than hard work and persistence.

However, there are mysteries we call coincidence or luck that have yet to be adequately explained by any theory.

Eight months ago, my wife was driving on a two-lane rural highway when she hit something. Unsure of exactly what she hit, she turned around to see if perhaps she had injured someone’s pet. She was on the shoulder of the road looking at a very dead skunk when a woman going about 55 mph slammed into the back of her vehicle.

By the time I got to the accident scene, my wife had already been taken to the hospital for precautionary care and the other driver had been arrested. It would be some time before we would know the full extent of my wife’s injuries. As I stood looking at the wreckage, I had the same thought countless others have probably had in the aftermath of an accident. If she had left the house one minute earlier or one minute later, this never would have happened.

From the perspective of chaos theory, the precise time my wife left the house was the attractor that produced these near-deadly results. Such events appear random, but chaos theory suggests that perhaps they are not.

For the moment, the best we can say is that even minor changes to initial conditions can radically influence future events, and although most of the time the butterfly batting its wings produces benign results, sometimes, as Lorenz discovered, they produce a tornado.