It’s tough being a manager; so many decisions to make. Here’s a blog that will improve management decision making.
Most decisions that managers make are complex multi-round affairs. They often involve making one decision, and then waiting to see what someone else will do before making another.
Examples abound: in manager-employee relationships; in employee – customer scenarios; or in buyer – supplier issues.
Decisions can be aided by understanding game-theory. The important measure is the payoff for each party. Games are often zero-sum – someone wins and someone looses. But if compromise can be reached, both parties win.
Here’s how it works.
Keep going or swerve?
Two drivers are racing towards a single-track bridge. Neither wants to give way – both are in a hurry and want to be first over. They must both make a decision at the same time. Do they keep going and try to win, or do they give in and lose?
How do they decide what to do? By understanding the basic principles of game theory, conceptual models can be used to understand complex problems in order to predict what actions others will take.
This blog considers two models and then shows how they can be used in real-world examples.
This game of chicken (above) has serious consequences if both drivers choose to keep going. The drivers must decide what action to take in a split second, without knowledge of what the other driver (or ‘player’) will do.
There are actually four choices in the above scenario. Let’s consider them.
The worst-case scenario is that both drivers continue straight on. This has severe consequences (or ‘payoffs’), certainly injury and possible death. It does, however, prove they are strong-willed. To swerve would show weakness.
Both drivers prefer a zero-sum game where the other swerves and they win – particularly if they’re both men! There are however two options where they tie, one of which unfortunately risks their lives.
The decision making process is complex and split second decisions must be made by each driver without knowledge of what the other will do.
A second decision-making model is knows as The Prisoners’ Dilemma. As with the above example it’s a one shot game. The person has to make a decision without the knowledge of what the other person will do.
Bill and Ernie have been arrested for stealing a car. They are told that they will each serve two years in prison for car theft. The police believe that they have both also been involved in a bank robbery, but they have no evidence of this.
Bill and Ernie are held in separate rooms and told that if they confess to the robbery they will serve a total of three years for both crimes. They are also told that if one confesses and the other doesn’t, that person will serve one year in prison for assisting the police whilst the other will serve ten years. So, how does this look on the decision making model.
In this game the best outcome for Bill and Ernie is to deny that they were involved in the robbery. They will then service two years in prison.
But what if Bill confesses and Ernie doesn’t? Bill will serve one year in prison and Ernie will service ten years. Likewise, if Ernie confesses and Bill remains silent then Bill will serve ten years and Ernie only one.
In order to remove the possibility of serving ten years in prison the best option for each is to confess since neither can be certain that the other hasn’t reached the same conclusion. By confessing they will serve one year if the other person remains silent. If they both confess they will each serve three years.
These examples are interesting, but the real power of game theory as a decision-making tool comes into its own when considering more complex issues.
Decision Making in Business
Southern Rail is currently in dispute with the RMT Union over who should close the train doors. This has resulted in months of misery for commuters across the south east of England.
By modelling the interaction using game theory it is possible to see why the two sides are taking the stance that they are.
Southern wants to make change; the RMT Union is resisting. As with the game theory discussed above there is an option where both parties can compromise and reach a solution that both parties can live with. At the moment both parties are seeking maximum payoff.
The outcomes that can be reached in any decision making process have a value. It is this value that drives the choices made. By understanding how this works it is possible to anticipate what action the other party is likely to take, based on the payoff they will get. In the rail dispute the best option for Southern (S) is to make the change (Hold). The RMT union believes the changes are wrong, so their best option is to oppose the changes (Hold).
Adding numbers to the example helps to understand why certain decisions are taken.
Considering the numbers in the payoff column, it is clear that the best payoff for Southern is to hold but to have the union concede. Southern have their maximum payoff of 10 units. The best payoff for the RMT union (10 units) is for them to hold and for Southern to concede. There’s currently an impasse because both parties are driving to maximise their payoff.
The greatest payoff for each player is to hold. This is the zero-sum option. Where one player holds and the other concedes there will be a huge loss of credibility.
The solution where both parties concede will result in compromise on both sides. Both players will win something – just not what they expected when they set out to win. This is easy to see on paper, but not easy to achieve in negotiations. If both parties are sitting in separate rooms at ACAS they are making decision in a similar manner to that in the prisoners dilemma example!
By considering all the options available at the payoff stage it’s possible to work backwards to determine the approach to the contest.
Decision making in business is not usually a single stage game. There can be many rounds before a conclusion is reached.
The example above considers a multi-stage decision making process with six stages concerning the conflict between an employee (EE) and his employer (ER). The employer knew the outcome they wanted and worked at each stage toward achieving that. In each round the employer considered options for the following rounds. This informed their choices at each stage of the process. In this scenario the employer achieved the best result possible.
When engaged in decision-making, parties will act independently and rationally based on their own interests.
As demonstrated, this does not always deliver a result that provides the optimum for both parties. To minimise impact, each player considers their actions based on their understanding of the other’s intent. There are many factors that come into play including mutual co-operation based on trust. Trust is built over the months and years preceding the game. Often trust is weak. Often the solution that requires the trust of the other party is ignored, since it leads to a solution that is high risk if one party defaults.
The game theory lens provides a model allowing players to consider options before acting. No decision is a two-choice decision. There are four options based on the two decisions that a player can make.
The game theory models illustrate the outcomes possible based on assumptions. Whether, of course, these initial assumptions are correct is a different story.