Decisions, Decisions

Most project managers know that the best decisions come from logical, deductive thinking. Taking over a family business, however, led the author to muse about bounded rationality and the limits of logic.

My husband bought his father's business last month. It wasn't an easy decision. Like most modern managers, I've been taught that the best decisions are the product of logical, deductive thinking. In an ideal world, we would have reviewed the financial reports of our existing business, examined the finances of the company we were considering buying, built models of future cash flows under varying conditions, and then unemotionally followed whatever course of action would maximize our profits while minimizing our risks. But we didn't have time for that. We had less than a week to make the decision and finish the paperwork, or we'd miss some crucial advertising deadlines. And in our industry, no advertising means no business.

Traditional economic theory teaches what is called the "rational judgment" model. This theory holds that the best possible decisions can be reached only by comparing all of the relevant information and by statistically weighing all of your options. Decisions should be logical, deductive and thorough. Unfortunately for theorists, this entire process seems to have little connection to what actually happens in real life, where information is often missing, incomplete, misleading or just plain wrong. In real life, we rarely have the luxury of all the time we need. In real life, pesky emotions intrude on the most logical of decisions, forcing us to take into account more than dry statistics and profit-and-loss sheets. In real life, even if we can isolate all of the relevant factors, the combinatorial explosion that occurs when we try to account for how everything affects everything else quickly makes the exercise unrealistic.

Rules of Thumb
More than 40 years ago, the Nobel Prize-winning psychologist Herbert Simon (1957, economics) described an alternative to the rational judgment model. According to his theory of "bounded rationality," it's often impractical or impossible in real life to calculate an optimal strategy. In fact, in many situations, there may be no single best strategy. Instead, he believed that people use heuristics—simple rules of thumb based on their experience—to reach a "good enough" decision.

Bounded rationality, which is gaining renewed attention in such divergent fields as psychology, economics, artificial intelligence and animal behavior, is explored at length in cognitive psychologist Gary Klein's book, Sources of Power: How People Make Decisions (MIT Press, 1998). Unlike most of the popular business press, which is often anecdotal with little data or rigor to substantiate the opinions, Sources is backed by Klein's many years of field research. He is interested in what he calls "naturalistic decision-making"—"the study of how people use their experience to make decisions in field settings." He writes "We all have areas in which we can use our experience to make rapid and effective decisions, from the mundane level of shopping to the high-stakes level of fire fighting." In Sources, he explores the decision-making process, examining how we do it and how we can improve.

In particular, he describes what he calls sources of power—four mental strategies employed by effective decision makers when they're under pressure: intuition, metaphor, mental simulation and storytelling. These mechanisms draw on our experience to facilitate instantaneous judgments about a situation. Intuition helps us to size up a situation quickly. Metaphor uses our experience by suggesting parallels between the present and the past. Simulation lets us mentally follow a series of steps to a conclusion before we actually carry them out. Storytelling helps us abstract and consolidate our experiences, so the lessons we've learned will be available to us and to others in the future.

Klein defines intuition as "the use of experience to recognize key patterns that indicate the dynamics of the situation." Once the decision maker has zeroed in on the patterns that identify a familiar prototype, he has available at least one reasonable course of action that has worked in the past and can start acting immediately. Intuition can be spooky; some decision makers almost seem to pick a decision out of thin air. In reality, of course, these patterns are subtle and don't always reach a conscious level of our awareness.

Spooky or not, psychological studies have indicated that intuition is based in biology, as some brain-damaged patients lack the intuition that normal subjects show as a matter of course. Since intuition grows out of experience, it can be cultivated by deliberately seeking a wide range of experience, and can also be formally trained.

Trained intuition is remarkably effective—I relied on mine a lot last month. My husband and I have owned a similar business for 20 years, so I had a gut-level feeling for what kind of cash flow, income and outgo we could expect with the new company. Of course, intuition isn't perfect, and that gave me a few nightmares while we were struggling to decide.

Mental simulation is intuition's natural partner. Intuition helps us see what is there; mental simulation helps us imagine what may be there in the future or what might have been there in the past. Klein defines mental simulation as "the ability to imagine people and objects consciously and to transform those people and objects through several transitions, finally picturing them in a different way than at the start." Mental simulation allows a decision maker to mentally experiment with several different options and to try out various steps before choosing a course of action.

Long before I knew the formal term for the technique, I learned to rely on my "mental movies." Many years ago, I borrowed a friend's horse so I could take my husband trail-riding. Smoky was gentle, but got nervous when cross-tied. I happened to look up while I was saddling another horse, to see my husband standing in front of Smoky. I was uneasy for some reason, and asked him to move. A couple of minutes later, Smoky reared without warning and fell over backward. If my husband had stayed where he was, his head would have been kicked off. As it was, I caught the horse, soothed him, and we had a nice ride. My husband still thinks I'm clairvoyant, and I gained even more respect for my mental picture shows.

Procrustean Problems
The very fluidity of a mental simulation is both its greatest asset and its greatest bane. Working with thought-stuff, we can construct simulations that would be impossibly expensive or time consuming in the real world. This is an incredible advantage when time and resources are short, but if we're not careful, we can use the power of our simulations to explain away any inconvenient facts. Nobel Prize–winning polymath Murray Gell-Mann's book, The Quark and the Jaguar: Adventures in the Simple and the Complex (W.H. Freeman & Co., 1994), tells a story about two physicists who were climbing near Aspen, Colorado. When they started their descent, they got turned around and came down on the wrong side of the mountain. As they descended, they looked down and saw a lake, which they identified as the familiar Crater Lake—which they were expecting to see about then. One of the men casually mentioned that there was a dock on the lake. Crater Lake has no dock. Unfazed, his friend replied "They must have built it since we left this morning." They did make it home finally, several days later.

We can smile at the impossibility of building a dock in a morning, but these were not stupid people. Just the reverse, they were smart enough to fit any inconvenient facts into their mental simulation of where they were. Luckily for us, we can improve the way we use mental simulations. I know several software managers and consultants who like to use what they call a "premortem strategy." Once a planner has created a plausible mental simulation, it's more difficult for him to imagine events unfolding in any other way. In Klein's version of the technique, the planner is asked to imagine that it's now the future, his plan has been tried and it has failed. That's all the information he has—now he must determine why it failed. This perspective helps to jar the planner out of his overconfidence. Of course, once he starts thinking seriously about the flaws in his plan, he's in better shape to guard against them or to create a better alternative.

Mental simulations are a conscious strategy. We usually know when we're using a simulation, and we can watch ourselves think as it unfolds in our imagination. A metaphor or simile is equally powerful, but much more subtle. When Robert Burns wrote "My love is like a red rose," he framed our understanding and expectations of love. These types of speech draw on one domain to explain another, and a good metaphor or simile affects both what we're able to see in the world around us and how we interpret what we see. As an example, Klein points out how the metaphor "Arguments are war" makes certain actions seem natural and other actions less likely. When we're under the influence of this metaphor, it's obvious that we need to attack our "opponents" and to defend our own weak points. But if we learn "Arguments are like music practice," we may be alert for ways to work together more harmoniously.

Management Metaphors
We use metaphors constantly in user interface and hardware design. But it's harder for us to understand that metaphors also affect the ways we talk to each other and the management decisions we make every day. In the new edition of Peopleware (originally published by DLC: Management House in 1987; 2nd edition, with eight new chapters, Dorset House, 1999), management gurus Tom DeMarco and Timothy Lister explore how our view of "teamwork" changes when we shift from thinking about teamwork in sports to the harmony produced by a choir or glee club. In describing what happened to their partnership when they took choir lessons together, they're following a teaching tradition that is tens of thousands of years old: storytelling. A good story is an abstraction that filters out the unimportant factors in a situation so that we can see the important points clearly. A good story is dramatic, featuring characters with whom we empathize. A good story has lessons to teach us, with a conclusion that resolves the conflict.

As a profession, we're fortunate to have a literature filled with good storytellers. People like Gerald Weinberg, Tom DeMarco and Tim Lister, P.J. Plauger, Henry Petroski, Robert Glass, Don Norman and many others have been telling us their stories for decades, if we will but listen. And closer to home, well-run peer reviews and postmortems are excellent ways for a software organization to create and to pass on its own stories.

Metaphors and stories work because they "chunk" experience into smaller units. Herbert Simon believed that experts develop their skill by building a large store of remembered patterns, each individual pattern compressing and summarizing an experience. He called this the experts' "vocabulary."

Experience is the underlying commonality in all of Klein's sources of power. Accurate intuition and pattern matching depend on our experience in the domain. Our ability to create useful simulations is based upon our knowledge and experience. The best stories are always a distillation of personal experience, while metaphors encode group experience. Of course, these aren't the only sources of power available to us. Bounded rationality should never be used as an excuse for ill-informed and impulsive decisions.

Rational analysis is an important source of power that has fueled the explosive growth of the computing field and the modern world in general. Klein compares rational, analytical thinking to our foveal vision. It allows us to make very fine distinctions that are simply impossible any other way. But if we don't use our peripheral vision at the same time, we quickly lose track of where we are and what's going on around us.

The lesson I learned last month is: "Trust your gut, but verify your assumptions." We decided to go ahead with the purchase, based on our gut feeling that the cash flow would be sufficient to break even the first year, and recognizing the many powerful emotions involved when a family business passes from father to son. That emotional stake was at least as important as the financial analysis in my husband's decision. But we made sure to check out as many of our assumptions as we could in the week we had, and I spent that week running mental simulations of possible futures, trying to spot any rearing horses so we could avoid them or at least prepare for them.

Was our decision the absolutely best one we could have made? I don't know. I'll never know. But it was the best one we could have made at that time, under those conditions. And that's good enough for me.