SEE MORE LEGAL SOLUTIONS »
Your premier source for legal solutions, including Thomson Reuters Westlaw and law books.
Practice Innovations  —  Managing in a changing legal environment
Gray Rule
March 2018 | VOLUME 19, NUMBER 2
Gray Rule
Using Expected Value Calculations and Big Data to Guide Decision-Making
»For easy printing – view as PDF

IN THIS ISSUE:
spacer

»Skill Fade: The Ethics of Lawyer Dependence on Algorithms and Technology
»Using Expected Value Calculations and Big Data to Guide Decision-Making
»Fending Off Incursions by the Big Four into the Legal Industry
»The Opportunity for "Back Office AI" in Law Firms
»Suffolk Law School: Leading Transformation of Legal Education
»The Future of Change is Client/Law Firm Collaboration
»Back to Contents

LINKS:
spacer
» Subscribe
» About Practice Innovations
» Editorial Board
» Past Issues
» Reader Feedback

rss feed View Feed XML

Using Expected Value Calculations and Big Data to Guide Decision-MakingBy Mark Thorogood, Director of Application Services, Perkins Coie LLP, Seattle, WA
In today's hyper-competitive environment, the best way to win the race is to ride the right horse. We may pride ourselves on being superb jockeys, but how do we identify the best option? In this article, I summarize the technique of calculating the expected monetary loss or gain from a given alternative. I also review some psychological hardwiring that may thwart our decision-making process.

As professionals, we seek to maximize the wealth of our clients, our companies, and our communities. We succeed by preventing and resolving problems, and by enabling people to achieve their dreams. In a basic sense, our driving purpose is to produce the most value for our clients at the lowest cost. There are patterns and tools that can propel us towards realizing this ambition; however, there are also patterns that can derail us. The ability to make effective decisions is an essential competency. There are techniques that aid us in selecting the best course of action out of a crowd of alternatives. There are also insights into human behavior that can help us steer clear of decision-making pitfalls.

We all want to devote our energies and talents to the most valuable endeavors. The challenge becomes: how can we rank alternatives so that we can identify options with the greatest expected potential? We could assess options by looking solely at their absolute values. However, that strategy overlooks the critical dimension of probability, which is intertwined with the concept of relevancy. The process of finding an alternative can be improved by looking at probability in combination with absolute impact. The technique of expected monetary value (EMV) takes into account both of these dimensions. To calculate EMV, multiply the probability (percentage chance of occurrence) by the impact associated with an event. The outcome is a ballpark estimate of the expected value in terms of dollars. Ascribing probabilities to a given alternative may seem tricky, but fortunately the probabilities can often be sourced from big data analysis, the internet, or expert opinion.

This concept can be readily applied to real-life scenarios. For instance, say we need to decide whether a client should settle a patent litigation suit. In this example, the cost to settle would be $900K. On the other hand, the exposure from an unfavorable judgment is estimated to be $1.2M. Based on an analysis of these data points, litigating the case would be out of the question—$1.2M is clearly greater than $900K. However, that reasoning does not factor in the likelihood that an unfavorable judgment would result. Using decision science and big data, we can gain a sharper economic understanding of whether to settle or not.

There are many data services that provide statistics on judicial outcomes. Big data enables us to harness the experience from many outcomes rather than relying on the memory of a few individuals. This strategy inherently leads to greater objectivity. But the data by itself is not sufficient to produce insights. We need a framework, such as EMV, to analyze and reason about the data. In the scenario above, the data may ascribe a 50% chance of unfavorable outcome with a potential judgment of $1M. Applying the EMV formula, we would find an expected monetary loss of $500K (i.e. multiply $1M by 0.50). In this case, the path of least expected cost would be to litigate. Alternatively, the analysis (with backing data) could be used to negotiate a less costly settlement.

Applying the concept of EMV may seem simple, but there is evolutionary hardwiring that may get in the way. In his book The Art of Thinking Clearly, Rolf Dobelli describes how human beings are naturally biased to focus on overall impact while neglecting probability. Take air travel, for instance. It is common for people to fear flying, likely because they focus on the potential of a fiery death while overlooking the probability associated with such an event. Everyone knows that flying is much safer than driving. Nevertheless, I know many people who willingly get in a car and drive around town on a Saturday night but refuse to board an aircraft.

Another example is the penchant for some to play lotteries. People who play lotteries focus on the riches they may win, overlooking the scant chance that a payday may be realized. The reality is that the expected monetary gain from a $1 lottery ticket is roughly 60 cents. Logically, that proposition does not make sense, but many people still buy lottery tickets because they maintain a one-dimensional view of the game. We should remain mindful that we may have a predisposition towards emotional and illogical reasoning, which may obscure our objectivity. To counter these predispositions, we can use tools such as EMV to reason about choices or risks objectively.

The concept of focusing on impact while neglecting probability is particularly relevant to the legal industry. Lawyers are trained to spot issues. Research also indicates that they tend to be pessimistic by nature. Because of this, lawyers tend to focus on the worst case. In doing so, they focus on the magnitude of negative outcomes. This behavioral bias can sometimes be referred to as the "dead tree syndrome," where legal professionals will proclaim an entire tree to be dead after spotting just one dead branch. I have seen deals sour because people cannot negotiate past a sticking point even when the sticking point is unlikely to materialize. When I come across these situations, I instinctively ask how much it would cost to insure against the risk rather than continuing to argue about a carve-out in a contract or settlement. I take this path, because insurance industries are expert in making expected monetary loss calculations. In a sense, we can leverage insurance quotes to objectively reason about a negotiating point.

In addition to behavioral biases, there are social biases that impair data-driven decisions. For example, a common trap is defaulting to the highest paid person's opinion to drive decisions. This decision-making pathology occurs so frequently that it is referred to as the HiPPO (or Highest Paid Person's Opinion) Syndrome. I am not saying to dismiss the highest paid person's opinion. However, it is always prudent to triangulate opinions with data and analyses where possible and warranted.

Another barrier to using EMV is the perception that probabilities are difficult to source. That may have been the case a decade ago, but in the age of big data and the Internet of Things, data is plentiful. There are many free and paid services that provide statistics to inform the business, practice, and administration of law. Moreover, new services are brought to market each day. These data services cover a spectrum of use cases. Some services provide comparative analysis between firms. Others provide insights into how judges are likely to rule on a specific motion. Still others provide estimated judgment amounts and win/loss statistics. There are also many services that provide benchmarks on law firm operations.

In addition to data services, there is also software that can help firms mine their internal data and convert their proprietary information into actionable insights. In a recent example, DLA Piper announced a program at ILTACON 2017 that will identify the key variables that affect client retention. In their study, they found the key variables to be team size, team composition, and the existence of a client-focused marketing program.

I use the technique of EMV to weigh alternatives almost on a daily basis. It has enabled me to get the most from the resources allocated to me. For example, my team was responsible for upgrading a major system at my firm. During a meeting, a senior engineer implored me to move the system that was being replaced to a new set of servers as soon as possible. He believed a hardware failure was "very likely" to occur within the next two months. He based his belief on the need to replace a failed hard drive in the server and assumed that other hard drive failures were soon to follow. The project to replace the system was on a tight timeline. Moving the legacy system to new hardware would have delayed the upgrade, but moving the system would avoid the prospect of a system failure. I decided to test the assumption that a failure was "very likely." I calculated the probability of server failure using the annualized drive failure rates that I sourced from the manufacturer. I found the risk of failure to be less than 1 chance in 10,000, which is hardly in the realm of "very likely." After discussing the analysis and data with the engineer, he concluded the same. Fast forward, the server did not fail, and the project stayed on track.

The concept of EMV is one of several techniques used in data-driven decision-making, but I chose to spotlight it because of its simplicity and practicality. Within the domain of decision science, there are other techniques that are worth exploring, including Maximax, Maximin, Minimax Regret, and decision trees. When selecting a particular model to use, consider their affinities, strengths, and weaknesses. Some are better fits for a specific circumstance than others. For instance, if the client is a high-roller, Maximax may be the best choice. But, if the client is risk averse, Minimax Regret may be the better choice. The models can be used in combination. Personally, I start with EMV and may add other models to tease out nuances and angles to a given decision. Details on these tools and examples of how to use them are readily available online.

If our goal is to maximize wealth and produce value, we should seek to identify the best course of action from a crowd of alternatives. We may pride ourselves on being superb jockeys, but in order to win the race, we need to ride the right horse. There is little hope for someone to win the Kentucky Derby while riding a plow horse, even if that person is the best jockey of all time. The concept of expected monetary value in combination with big data is one method to objectively identify the right horse and to avoid the pitfalls concomitant with emotional reasoning.

Back to Contents