The End-game - Decision Analysis

Doctor Karl - Doctor Karl is Karl Edward Misulis, a physician, scientist, educator, and university professor. He has a BSc from Queen's University Canada, MD from Vanderbilt University, and Ph.D. from SUNY Syracuse. He has authored more than 20 books, some in multiple languages, and has lectured around the world to the public as well as to fellow academics.

Introduction

This is part of a series of lectures and briefs on decision-making. We have already discussed algorithmic decisions, cognitive bias, and logic, among other topics. In this brief, we consider the end-game - decision analysis.

We made our decisions, and we have our outcomes. Now it is time to analyze our decisions. Why do we do this? We do it because we need to determine whether our decision-making is on-target and valid or whether it is flawed and needs to be redesigned for future decisions. Perhaps we might be able to take a step back and fix the outcome of the decision under question.

To illustrate this process, we will analyze my decision-making process as I apply it to my life and career. I have come a long way and made many good decisions and a few for which I would like a do-over.

My own experience

Young Karl was a bright, imperfect child, not far different from most on our planet, with many possibilities for the future, some more realistic than others. I decided I wanted a career in medicine. There are different ways to do it in medicine, but a significant branching point is to be a scientist or a practicing physician. These are related but very different in workflow and lifestyle. And there are a few people who do both, spend part of their effort doing clinical medicine and part doing research. That blend seemed exciting to me, so that was my decision. I got an undergraduate degree from Queen’s University in Canada, a medical degree, MD, from Vanderbilt University in Tennessee, and a scientist degree, Ph.D., from one State University of New York campus. Armed with those credentials, I was ready to take on the world - time to get a job. But which position do I want? Would I be the full-time scientist, the full-time physician, or the rare but exciting physician-scientist, doing both? At that time, I chose the latter. I didn’t think about it much, I was prepared for that job, so I was going to give it a go.

Three possibilities

What if I had done a formal pre-decision analysis? Which would have been the logical choice? It depends on what parameters we want to use to measure success. So let’s step back in time and do that.

All three of these possibilities are complex, and the sustainability is very different. It is elementary to get and see enough patients make a good living as a physician. Even the last person in the medical school class did well. However, science is complicated and depends on getting problematic grants. So here are the numbers.

A pure physician has a 95% chance of making a lucrative career. A small number, less than 5%, is not successful, either through personality quirks, being too picky about their situation, or just deciding to quit medicine.

A pure medical scientist has a 34% chance of getting grant after grant and getting academic advancement to stay in the game. The rest leave to form or join a business or leave science altogether. That is not a happy statistic.

The physician-scientist, blending clinical medicine and research into a long-term career, is even worse, 20%. The other 80% who try this path cannot do both well, so they flip to one or the other, science or medicine, and in that case, medicine.

I took that physician-scientist path, thinking I might be one of that 20%. I knew the odds were against me, but I wanted to take my shot. I missed. So, now I am a physician and not a research scientist. I do other things like teaching and writing, but being a physician pays the bills.

Did I make the wrong decision? No.

I know the odds were against me. But I wanted that physician-scientist role. I did my best, but my best was not good enough, at least my best without giving up on family and fun. It didn’t work out, so I went to plan B and am happy. That was my decision-making and decision analysis.

Now, let us consider a person considering their first or subsequent job opportunity and making a decision with an elementary branch point. One branch is starting a business. The other branch is becoming an employee of an existing business. Before making our decision, we must decide the measures of our success. There are two key performance indicators or KPIs for someone looking for or starting a job. One is monetary compensation, we need to be revenue positive at some point, and different jobs offer different positivity; we cannot go our entire lives revenue negative. Another KPI is personal satisfaction; we are happy and fulfilled with our actions.

So, our personal KPIs are income and satisfaction, and we will make our choice depending on our estimate of what those KPI outcomes will be if we either start a business or become one employee. Can we predict how our KPIs will turn out? To a certain extent, we can have data for other people since millions have started businesses and billions have become employees.

As we look at data, let us first promise to minimize our biases. We are not necessarily smarter or harder working than many others, so while those attributes will affect our success, let us not assume we are gifted or entitled. Let us be objective.

What is the data? We have numbers on those exact KPIs. The numbers are a little dirty because the financial success data is from the US market and the personal satisfaction data is from Europe, but still, that is the best we have. And I have combined data from more than one study and rounded the numbers, but here we go.

How I Use Multiple KPIs to Make Decisions

First, for monetary compensation, if someone becomes an employee, they have a 90% chance of getting the expected income. About 10% did not, and we don’t know whether that is because they dreamed they would get more or felt they were trapped into taking a low-paying job or were being cheated; the data can’t tell us that. For the business starter, the chance of the business being open after one year is 60%, the opportunity of being revenue vivacious is 30% in the first year, and 40% in the business’s lifetime. Ouch!

But before we all decide to be employees, let’s look at the satisfaction data. Of the employee, about 50% have positive satisfaction, more satisfied than not. The other 50% are okay with their job but not personally as confident as they might have wished. For a startup business, the satisfaction is 80%. Why would that be? Because there are elements to satisfaction other than money, between the excitement of building and designing a business and taking their shot, the joy far exceeds the financial success. Of course, we cannot pay our electric bills with satisfaction, so either the business-starter is one of the 40% who becomes sustainable, or at some point, they step back and consider taking another shot at another business or decide to be an employee.

This is parallel to my career. I took my shot at a low-probability-of-success opportunity. I missed it, switched to Plan B, and I am happy. I do not (often) think of what might have been, and I have no regrets about the time and effort I put into taking that shot. The point here is that I am glad that I tried to shoot for the moon. It did not work out, but I took my fallback plan and am satisfied.

Where does this take us on decision analysis?

First, we do our homework. For our key decisions, make sure we know the possible outcomes. We place value on each of those outcomes as possible. Determine what our KPIs are going to be. Make a decision and go with it. When we have interim or outcomes, see how they stack up on the KPIs. When these data are actionable, apply the same rigorous decision to whether we should take some new or alternative action. Perhaps we can change the outcome. Perhaps not.

Second, we see how we did about decision-making. We will make countless essential decisions in our lives, so ensure we make those decisions wisely. 

Were we objective and informed about making our KPIs? Did we make the best effort? A standard failure method is getting off to a good start and then coasting to the finish line, a losing strategy.

Third, apply what we learned through this experience to our next major decision.