Decision engineering rests on the following three fundamental premises:
- A good outcome is not the same as a good decision.
- A decision is an irrevocable allocation of resources. Decision making is thought about action or actionable thought. Once an decision is made and action has been taken it cannot be reversed.
- We can increase the probability of good outcomes by improving decision making.
We all know that a good outcome is not the same as a good decision. Suppose a drunk insists on getting into his car to drive himself home after a party. Suppose further that he arrives home safely. We can agree that he had a good outcome. But did he make a good decision? Most of us would say no. He made a bad decision and, fortunately, he had a good outcome. Decisions and outcomes are different. We can make good decisions and have bad outcomes, and vis versa.
It is easy to recognize a good outcome: we win the lottery, our cancer is cured, or we have a successful career. But how do we know that a decision is a good one at the tiime it is made?
I studied engineering at Stanford in the 1970′s. I learned from Professor Howard that a good decision is one that is logically consistent with what we know, what we want, and what we can do. The three legged stool is the icon for this way of looking at a decision. The legs are the alternatives, the information, and the values. We really don’t have a decision unless we have all three legs of the stool. The triad, together with the frame it sits on, has a technical name, the decision basis.
That is how it was taught to me 40 years ago by Professor Howard. That’s the way he teaches it today. The definition certainly has passed the test of time. I have used this framework successfully throughout my 40 year research and consulting career. It is theoretically sound and practical.
The implication of Professor Howard’s definition is that we can make better decisions if we improve our logic. For example, in the movie Moneyball, the character played by Jonah Hill showed Billy Bean (Brad Pitt) how he could improve his decisions by taking available player data and combining it logically. He improved the baseball team’s success dramatically using pure logic.
Another definition of a good decision that I find compelling is offered by the late Chris Argyris of Harvard. This definition has its roots in organizational development. The definition is particularly useful when there are many people involved in a decision.
Argyris defined a good decision as an “informed choice based on valid information.” Further, that a good decision is one where the parties involved are “internally committed to the choice.” People universally agree that we make better group decisions if we think and act according to the Mutual Learning Model shown above. You can learn more about the Mutual Learning Model in the landmark book, Overcoming Organizational Defenses.
We can combine the Howard and Argyris definitions and say that it takes a combination of logic and mutual learning to make good business decisions. In other words, there is a hard side and a soft side to a good deciison. As we see in the film Moneyball, Billy Bean used logic to formulate his recruiting strategy, however, his staff deliberately did their best to subvert his strategy. They were not “internally committed.” That’s how life in organizations works. Logic alone does not win the day.
As decision engineers we need processes and tools for dealing with a broad range of dysfunctional behavior, intentional and unintentional. Chris Argyris and his colleagues have developed several tools to help people move toward Mutual Learning. The tools include the ladder of inference, balancing advocacy and inquiry, and the “left hand, right hand columns.” I have found all of them extremely useful.
Do you have a another definition of a good decision? Is it operational? To see how I think this all fits together see my recent INFORMS presentation titled, The Collaborative Design Process and Decision Engineering.