What Founders Need to Know About Game Theory
Game Theory is the study of decision making in situations of conflict, where the success of one’s decisions depends on the decisions of others. Game theory is a powerful tool. It can be used to predict, to optimize, and to find a path through the fog and chaos that we often believe to be our day-to-day business environment.
However, game theory is not a panacea. It is not a magic bullet. It cannot solve all problems or answer all questions. Instead, it should be used as another tool in your toolkit for decision making and problem solving, and in this article I will explain why you should keep it there.
Why limit ourselves?
Game Theory has made enormous contributions to economics and international relations. Yet these fields cannot solve all of our problems. We still have many unanswered questions about human nature that do not necessarily fit into neat models based on rational actors maximizing utility.
How much does this job pay compared to my other opportunities? Should I hire this person based on their résumé or their interview? How much time should I spend reading science fiction vs actually doing science? These are all questions that are not easily answered by game theory, and I believe that they are also questions that Founders should be asking themselves.
Game Theory is a useful tool for predicting human behaviour, but it does not account for everything we want to predict.
It is an excellent tool for optimising outcomes, but it will not show us the best outcomes.
What Game Theory can do for you
Game Theory can help you understand your competitors better. It can help you predict how they will react to your moves, and thus allow you to anticipate their actions and navigate through the fog of war.
In games such as chess, where players know what strategies their opponent is planning on executing in advance, game theory provides a framework by which players can plan out exactly what steps they will take in order to win the game.
In games such as poker or bridge (or even an investment pitch), where players do not know what strategies their opponent will execute and must instead make decisions based on incomplete information, game theory provides a framework by which gamers can make calculated decisions under uncertainty.
Game Theory is not without its flaws though. To demonstrate, let us consider a simplified example of Game Theory known as “The Prisoner’s Dilemma”:
The Prisoner’s Dilemma
Two prisoners are separately being interrogated in separate rooms by different interrogators who have no communication between them whatsoever (so there is no way for prisoners to communicate with each other).
The interrogator offers each prisoner the following deal: If neither prisoner confesses and both prisoners stay silent (which is called “cooperate”), then both of them will receive no jail time whatsoever and will be released immediately after the interrogation session ends (this option is called “cooperation”).
However, if either of the two prisoners confesses while the other stays silent (which is called “defect”), then both prisoners will receive 1 year of jail time (this option is called “defection”).
Now suppose that one prisoner has been told that he will definitely receive 7 years of jail time if he confesses while his partner stays silent; however, he does not know if his partner has or hasn’t been told about this fact.
The second prisoner also doesn’t know what has been said about him or his partner, but he knows that whatever information has been given to him will also be given to his partner.
Both prisoners must decide whether or not to confess before the interrogator can choose what to do with them.
If both prisoners confess, then they both get 1 year of jail time. If both prisoners cooperate, then they both get 0 years of jail time, which is obviously a better outcome than either of the other two options.
If one prisoner defects (confesses) and the other cooperates, then the defector gets 1 year of jail time and the cooperator gets 7 years of jail time (which is a worse outcome than if they had both cooperated). If one prisoner defects and the other defects, then they both get 7 years of jail time (which is a worse outcome than either of them would have gotten if they had stayed silent).
What should these prisoners do?
This is an important question because it has implications for things that we care about in real life, such as business and science. If both prisoners cooperate, they both get the best outcome.
But if one prisoner defects, then he gets a better outcome and his partner (who did not defect) ends up in a worse position. Thus, each prisoner has an incentive to defect.
However, if both prisoners defect, then they end up with the worst outcome. If one of them stays loyal to their partner and chooses to cooperate rather than defecting, then he can get a better outcome for himself and his partner.
Thus, each prisoner has an incentive to stay loyal to his partner (to cooperate) rather than defecting. But if both prisoners choose to cooperate, then they end up with the worst outcome.
Thus, there is no rational choice that these two prisoners can make at this point in time; they are caught between a rock and a hard place!
What do you do?
Here is where game theory comes into play, there are three possible outcomes here:
- Both prisoners defect;
- One prisoner defects while the other cooperates;
- Both prisoners cooperate.
Game theory provides rules for what players should do when faced with this choice (as well as many other choices).
The two most important rules are:
- Always choose whatever choice leads to the best outcome for you regardless of what your opponent chooses (this is called “maximizing your own payoffs”), and
- Do not reveal information about your strategy until it is too late for your opponent to react effectively (this is called “muddying the waters”).
These two rules can help us craft our optimal strategy for this game:
We should both cooperate on our first move because doing so will lead to the best possible outcome for us regardless of what our opponent chooses — we will only get 0 years of jail time if we both cooperate.
Then we should muddle the waters by choosing different strategies on subsequent moves; once again, doing so will lead to the best possible outcome for us regardless of what our opponent chooses — we will only get 7 years of jail time if one of us defects while the other cooperates.
By following these simple rules we have crafted an optimal strategy that defeats our opponent under all circumstances.
If we were in a real-world situation where we could use Game Theory like this example did, then it would be very useful indeed. However, there are some major limitations that prevent games like this from being useful in real-world situations.
Limitations of Game Theory
Games like the Prisoner’s Dilemma are known as “zero-sum” games. In a zero-sum game, the sum of the payoffs that all players receive is zero. In this game, if one prisoner gets 1 year of jail time then the other prisoner gets 7 years of jail time, and vice versa.
There are two major limitations to zero-sum games:
- They do not represent real life very well, and
- hey do not allow for repeated interactions.
When you look at the world around you, it becomes immediately apparent that there are many situations where these two constraints do not hold.
When two business competitors are bidding for the same contract or two scientists are publishing articles in the same journal or two investors are bidding on the same company, then neither competitor will necessarily get a better or worse outcome than their competitor just because they have chosen to cooperate or defect differently.
When you look at human behavior in general, it becomes immediately obvious that we often like doing things for other people without expecting something in return (for example we may donate to charity), and we often help others because they helped us even though they did not expect anything in return (for example we may help a friend with his homework).
If you think about it carefully enough, you will realize that most relationships between people don’t fit into a zero-sum framework; rather they represent positive-sum games where people can end up better off as a result of interacting with each other.
A perfect example of this is science: The more scientists who study a topic, then the more likely it is that interesting scientific discoveries will be made — not only for researchers who have already made significant progress on their research project but also for researchers who have yet to make any progress at all!
Thus rather than being trapped between a rock and hard place like prisoners in our previous example would be, Scientists have an incentive to cooperate with each other so that everyone can collectively move their research projects forward faster.
This is why science is driven forward by collaborations between multiple scientists from various backgrounds working together towards a common goal rather than being driven forward by individuals competing against each other for publication credit and grants!
So while game theory can provide some useful guidelines about how best to act in certain situations like our previous example demonstrated (even if this advice only works when certain conditions hold), there is no general rule about what kinds of strategies work best in which kinds of situations.
There is no one right strategy that will always work out best for you and your company. Instead, you need to use game theory as a tool to help you think about the problem at hand more clearly so that you can make better decisions for yourself and your company.
For example, if you are a founder who is thinking about suing another founder for theft of intellectual property, then using game theory can help you think about the potential consequences of such an action.
For example, if another founder stole your idea in order to start his own company, then he may be the first to market with your idea and gain an unfair advantage over you.
Or if another founder stole your idea in order to build out the same product as yours but with better features than yours (i.e. he is trying to beat you with a better product rather than market share), then he may win over all of your potential customers because his product has better features than yours does.
Or maybe another founder stole your idea in order to build out a completely different product which only superficially resembles yours (i.e. he is trying to beat you by fundamentally changing the way that people approach this problem rather than just building a slightly different version of what already exists), then it might be hard to prove that he stole your idea (even if he did) and difficult for him to defend himself against any accusations that he stole your idea (even if he didn’t).
You could even try working out some sort of settlement agreement where both founders agree not to develop competing products but where neither one of them has any control over what their co-founder decides to do in the future; however, this would still leave open several other possible attack vectors which have been left unsaid or unaddressed in this agreement!
The point is that by thinking through all these various scenarios ahead of time, a founder can make much more informed decisions about how best to proceed since they know what it is they are really trying to protect themselves from!
In summary, while game theory can provide some useful guidelines for thinking about how best to achieve your goals, there is no one right strategy that will work in every situation, especially when you are dealing with other human beings.
This is why science is driven forward by collaborations between multiple scientists from various backgrounds working together towards a common goal rather than being driven forward by individuals competing against each other for publication credit and grants.
This is the same reason why you should not try to make all of your company’s decisions in isolation; instead, you should actively seek input from others who do not share your personal biases (and may even challenge your assumptions) so that you can think about these problems more clearly.
About the Author
I am the Founder of Cudy Technologies (www.cudy.co), a full-stack EdTech startup helping teachers and students teach and learn better. I am also a mentor and angel investor in other Startups of my other interests (Proptech, Fintech, HRtech, Ride-hailing, C2C marketplaces and SaaS). You can also find me on Cudy for early-stage Startup Founder mentorship and advice.
You can connect with me on Linkedin (https://www.linkedin.com/in/alexanderlhk) and let me know that you are a reader of my Medium posts in your invitation message.