This is the third installment of a three-part series on game theory.
July 21, 2021
By Rachel Smith
A few weeks ago, we explained game theory, and last week we dug into how it applies to sales. Many theories and theoretical frameworks get disproved or lose popularity over time because they aren’t the best model for describing a phenomenon. For example, we no longer believe the sun goes around the earth or that toothaches are caused by tooth worms—yes, that was a thing! Game theory, however, is still being refined and successfully applied in more and more situations.
From disease treatment to artificial intelligence, game theory continues to help us model possible outcomes and make good decisions. This week we’re taking a look at how game theory is being applied right now to today’s latest treatments and technologies.
The conventional treatment protocols for metastatic cancer are based on the maximum-tolerated dose (MTD) that a patient can endure. The MTD is administered continuously or repeatedly until the patient experiences negative side effects that are too much to handle or the tumor grows instead of shrinks.
The idea is that it’s best to kill the cancer as quickly as possible. This can be a successful approach if there are no drug-resistant cancer cells at the beginning of treatment and none emerge during treatment. The problem is that for metastatic cancers, this is almost never true. In fact, tumor cells have an amazing ability to evolve drug resistance. An MTD approach often results in a patient being left with a smaller tumor, but one completely made up of drug-resistant cells.
This is where evolutionary game theory comes in. When cancer treatment is looked at as a competition between a physician and cancer cells, the physician actually has an advantage in a few ways. First, the physician is acting rationally, meaning she can make deliberate decisions based on her information and knowledge of cancer. Second, she gets to be the leader, administering a treatment and waiting for the cancer cells to respond.
Game theory exploits the physician’s advantages. Instead of killing cancer cells as quickly as possible, the physician is anticipating cancer’s next move and aiming for a stable state. This might not mean a complete cure—it might mean living with a chronic disease instead of a terminal one. Evolutionary game theory has previously been successfully applied to antibiotic resistance and controlling agricultural pests. Hopefully its application to cancer treatment will have similarly successful results.
Most of the research in cybersecurity is focused on a specific vulnerability or a specific defense for a vulnerability. The problem is that cyberwarfare does not exist in those separate bubbles. Cybercriminals attack, defenders respond, the criminals change their strategy based on the response…sounds like a good scenario to apply game theory, doesn’t it?
A cyber-defense plan based on game theory will consider the changing interactions between attackers and defenders. It will also consider how attack and defense strategies evolve in response to each other. Game theory allows those defending their cyber network to anticipate the attacker’s next move. This allows defenders to create their own deceptive traps for potential attackers.
Applying game theory to cybersecurity doesn’t just take into account attackers and defenders. There is a third player in the game—users. Users want to be able to perform whatever they’re doing like they always have. They see most cybersecurity measures as a burden. They are a key player to consider since their actions (think: using “Password” as a password) can severely hinder defenders.
Using game theory as a framework lets cybersecurity system providers and developers have more of an upper hand than they would otherwise. By anticipating their opponents next move, they can actually take the initiative for a change.
When determining how to deal with a disease under normal circumstances, you look at what has worked in the past in similar situations. That has not been possible with COVID-19, however, for two reasons. The first is that serious pandemics are a relatively rare occurrence. Second, in past situations, the world was not so interconnected. Luckily, game theory has provided a useful tool during the pandemic in figuring out everything from social distancing to vaccine distribution.
Game theory provided the perfect framework for vaccination decisions since it is, in essence, “the prisoner’s dilemma” playing out over and over again. Players are deciding whether they should cooperate (get the vaccine) or just consider what is easiest individually (not get the vaccine). The more people who get vaccinated, the less the virus will spread and the safer we’ll all be. But if I think everyone else is getting vaccinated, maybe I can just rely on herd immunity and avoid getting vaccinated myself.
The problem with many infectious disease models is that they don’t consider human behavior. Game theory, on the other hand, is all about human behavior (or at least rational human behavior). It provides a more realistic projection of what will happen.
Researchers employed game theory to figure out who should get the vaccine first. There were two models being considered. One was direct protection, which would mean the most at-risk populations would get the vaccine first. The other option was indirect protection—vaccinate young people first because they spread the disease.
It turns out that the answer to which model was best depended on when the vaccine became available. If it was between January and March (which it was), it made the most sense to vaccinate the older, at-risk population first. If a vaccine had not been available until July or later, it would have made the most sense to vaccinate young people first.
Malcolm Gladwell has just begun a sixth season of his wonderful podcast, Revisionist History. It is, in his words, “a podcast about things that are overlooked and misunderstood.” I highly recommend it. The first episode of season six was about self-driving cars, and whether they will or won’t solve all of our traffic problems. Mr. Gladwell and game theory both say no.
The idea of self-driving cars is exciting and promising. A computer can react faster and see better than any person. Car accidents would be a thing of the past. Traffic jams would disappear. This model, however, is only considering the self-driving car and the people inside that car as the players. But what about the people outside of the car? I’m talking about pedestrians.
To understand how pedestrians ruin our idyllic, no-more-traffic-jams world, we have to understand the Nash Equilibrium. John Nash (the mathematician portrayed in A Beautiful Mind) was a great contributor to game theory. He proved that there is a point in any finite game at which no player can do better than they are already doing by changing their own strategy—the Nash Equilibrium. In the prisoner’s dilemma, the Nash equilibrium is achieved when both prisoners confess.
Think about a pedestrian who wants to cross a busy city street. They use the crosswalk and follow the rules because they don’t want to get hit by a car. If all of the cars are self-driving, however, they are all programmed to stop when something (or someone) is in the way. The Nash Equilibrium for a world of self-driving cars is for pedestrians to keep walking regardless of what’s barreling toward them and for the self-driving car to stop. Imagine New York City with invincible pedestrians. And you thought traffic jams were bad now!
While game theory arose in the 1920s, the framework is still predicting outcomes and guiding behavior in such advanced fields as cybersecurity and driverless vehicles. It’s rare that a theoretical framework proves to be useful in such a wide range of areas. From the simplest two-player, zero-sum games to national military strategy, game theory provides an incredibly useful way to look at things.
Are you ready to learn how game theory can improve your sales strategy? Schedule a Question Trees workshop with Maestro and be ready for any scenario. Contact us at mastery@maestrogroup.co.
Get the Maestro Mastery Blog, straight to your inbox.
