Cutting-Edge Project Management: What Game Theory and Poker Can Teach Us

By Gil Junqueira | minute read

As project management continues to grow around the world and become increasingly commoditised, more specialisation will be necessary in order to secure the best opportunities. One of these areas of specialisation that will have increased demand in the future, I believe, is risk management.

World-class competition, lower costs and increased demands for scarce resources are pushing companies to operate in that frontier - the fine line between chaos and efficient use of tools, people and materials. The future operating style of in demand Project Managers will fall somewhere between a conservative money manager and a poker player.

Poker players have to constantly make good decisions with incomplete information, the goal being to maximise returns while minimising risks. They use the power of observation and deduction to make reasonable assumptions about their competitors, use calibrated estimates to calculate odds of winning and then choose the best course of action - all while keeping their own emotions in check. They are always prepared to change strategies, using a combination of both social skills and technical fundamentals. In this dynamic, ever changing environment, a poker player knows it is futile to think in terms of exact numbers, in a domain that belongs to probabilities.

Project Managers who desire to be seen as the best in their field will be required to employ these same tactics.

The link between strategy games and the business world is not new. Innumerous articles and books are available detailing how one can learn concepts of game theory and how they apply to business situations and even life in general. As the Project Manager moves from the deterministic domain to the probabilistic one, he will have to consider the risks associated with every given decision.

So what can world-class poker players teach managers?

• How to measure and acquire information: Risk (uncertainty) exists because we have incomplete information. If we could somehow acquire more information or make the existing information more reliable, we would be able to reduce uncertainty. In the real world, we acquire more information through measurement. Thus measurement is and will continue to be a key tool that can be used to mitigate risks. Managers who can make use of well-placed measurements will have the leading-edge against the competition.
• Measuring risk: When risks are involved, there has to be a way to measure them. Most companies only go to the extent of saying a given risk is "low", "high" or "medium". However, this connotation is disconnected from any meaning. Wherever real business is done, measuring risk has to go deeper and forgo unrealistic precise values and use ranges. With ranges, managers do not need to assume what they don't know - they might not know exactly when a given task will be complete, but they might have a high-degree of confidence that it will take more than 4 days, but less than 15 days. Ranges are better representatives of uncertainty. Monte-Carlo simulation is one of the tools that uses ranges to analyse and measure risk.
• Calibrated estimates: A good project starts with a good estimate. No matter how well you execute your projects - if your estimate is wrong you are in for a rough ride. With calibrated estimates, you don't really have to assume anything you don't know. What does a calibrated estimate look like? You estimate a Confidence Interval (CI) and a range of values. You may not know exactly how much a product costs, but you might be 90% sure (CI) that the product in question costs between $2000 and $7000. In this case, your CI is 90$ and the range is 2000-7000. With a little practice, you might be able to estimate anything from the percentage of sun radiation that strikes the earth to the number of piano tuners in Chicago.
• Calculating the value of information: When we are faced with a binary decision that has a risk involved, we know we might choose the wrong alternative. And if the wrong alternative has a price tag associated with it, we might want to conduct measurements (obtain more information) before making a decision. The information itself, therefore, has monetary value. If we knew what this monetary value was, we would probably choose to spend more time to reduce uncertainty in the decision, or at least think twice before stating that a given measurement would be too expensive to conduct.
• Choosing alternatives with positive expected returns: Every poker player knows that he should enter hands where pot odds (rewards) are larger than odds of making the hand (risk). If he does this consistently, he will have positive returns in the long-run. We can all benefit from this little advice. By stacking the odds in our favour, we increase our chance for success. On the other hand, if we choose alternatives where the risks are greater than the rewards, we have negative expected returns, and we will not stay in business for long.
• The 80/20 rule: Most of us are familiar with the 80/20 rule. However, we do not totally appreciate what far reaching applications it has. 80% of revenue comes from 20% of customers, 20% of problems causing 80% of lost productivity, 80% of losses coming from 20% of risks, 80% of consequences coming from 20% of the decisions, 80% of value coming from 20% of tasks, 20% of the customers generating 80% of the complaints, 20% of employees performing 80% of the work, and the list goes on and on. Learn how to understand and appreciate this simple rule and you will start paying attention to what is really important.
• The rule of 5: The rule of 5 is about sampling. As explained above, one way to reduce uncertainty is to collect more information. We can learn a good deal of the entire subject of study (population) by analysing only a fraction of it. You don't have to be a Statistics wizard to know how to take a sample to learn more about a population. In fact, if you have to remember a statistic law that you can use in your everyday work, learn the Rule of 5 - Randomly sample 5 subjects from a population you want to learn something from. There is a 93.8% chance that the median of the population is between the smallest and the largest values in the sample. Want to have an estimate of employee commute time to work? Randomly select 5 employees to ask this question and you will learn a whole lot more than you did, for almost no effort at all.
• Building a toolbox: The problem today is not so much a lack of information, but how to organise and find relevant information in an easy and timely manner. A serious Poker player will tell you that he has spent a considerable time creating, growing and fine-tuning his toolbox. A toolbox is an easily accessible knowledge system containing information that can be used at the drop of a hat. Concepts, rules of thumb, formulas and contacts could all be included in a toolbox system. In a dynamic, competitive environment, you will have to be able to make complex decisions using back-of-the-envelope type analysis. This can only be accomplished if your knowledge system is up-to-date and at your finger-tips.
• Mitigate risks with buffers: The player with the most chips on the table is called "The Table Captain". This is because the amount of chips a player owns can be a distinct advantage (or disadvantage). Other players know that the Table Captain can wipe out half of the players on the table if he so chooses. Poker is a game of high-variance, which means that there is a lot of uncertainty involved. One way to deal with such uncertainty is to use buffers - a reservoir of whatever we wish to maintain at a certain level. Do you have risks associated with the delivery of a certain raw material? Find more suppliers able to deliver and you are increasing your buffers and mitigating your risks. Toyota is a classic example of implementing an inventory system free of buffers, by using what is called "Just-in-Time" manufacturing processes. Unfortunately, such a system leaves little room to deal with variances and uncertainty. And we all know where Toyota's quality ratings stand now.
• Social skills: Last but not least, an important, but often underrated skill is the ability to read people's emotions, intentions and desires. No amount of technical skills can replace these abilities. A Poker player that cannot read other player's "tells" will not rise to the top. This is also true with a manager who does not correctly evaluate his team's emotional state and reads past superficial intentions. Best of all, social skills, including Emotional Intelligence, is a learned ability that can be honed and expanded.