The Guard’s Dilemma, a (sometimes) Useful Framework
By Guy Incognito
The Prisoner’s Dilemma is a framework for understanding decision making that can be tremendously descriptive of our reality. When reality gets more complicated, the model can typically be refined through any number of adjustments to accommodate the additional complexity. A hard working army of scholars is fighting to learn the nuances of how to play this game.
Why? They want to be able to make better decisions in an uncertain and highly complex reality. The game describes that reality, so if they can understand how to make better decisions in the game so that they can make better decisions in life. That sounds reasonable enough.
But the question remains: Why?
Here’s the problem: that army of scholars is losing the war. They’re really struggling to learn the game better. For the amount of effort they’re expending, the Prisoner’s Dilemma game analogy isn’t improving their life decisions very much.[1]
That army is fighting a difficult battle. And that’s what I don’t like. I like easy problems. I know how to solve them.
How can we more easily use a game analogy to improve real-life decision making? What type of game will make it easy to get to good outcomes? I can think of three key characteristics it should have.
1) The game’s design should effectively describe parts of reality where we want to improve decisions
2) If you only want to get to good outcomes, the game should always quickly generate good outcomes for all involved
3) The game should lead to good outcomes even if poor strategy is used. Just in case, the best strategy should become obvious to everyone as soon as possible.
Many variations of the Prisoner’s Dilemma game meet the first criteria, which is a necessary condition for improving real-life decisions using the game analogy.
But many of the most interesting variations fail the second two criteria, often spectacularly. So why not try to find a game that meets these three characteristics as much as possible, and try to apply that analogy as broadly as possible? Isn’t that a better way to use a game to improve real life? A game’s descriptive power isn’t enough. It should be prescriptive too. If you want to get good outcomes, find frameworks that lead to good outcomes and apply them where they are descriptive and aren’t being used.
How? Let’s just observe the Prisoner’s Dilemma objectively. Prison Guards do that, and their perspective might help. Presumably they have some Dilemmas. When you stop to think about it, they probably have a lot of dilemmas. Here’s a caricature of the dynamics of a prison:
By Guy Incognito
The Prisoner’s Dilemma is a framework for understanding decision making that can be tremendously descriptive of our reality. When reality gets more complicated, the model can typically be refined through any number of adjustments to accommodate the additional complexity. A hard working army of scholars is fighting to learn the nuances of how to play this game.
Why? They want to be able to make better decisions in an uncertain and highly complex reality. The game describes that reality, so if they can understand how to make better decisions in the game so that they can make better decisions in life. That sounds reasonable enough.
But the question remains: Why?
Here’s the problem: that army of scholars is losing the war. They’re really struggling to learn the game better. For the amount of effort they’re expending, the Prisoner’s Dilemma game analogy isn’t improving their life decisions very much.[1]
That army is fighting a difficult battle. And that’s what I don’t like. I like easy problems. I know how to solve them.
How can we more easily use a game analogy to improve real-life decision making? What type of game will make it easy to get to good outcomes? I can think of three key characteristics it should have.
1) The game’s design should effectively describe parts of reality where we want to improve decisions
2) If you only want to get to good outcomes, the game should always quickly generate good outcomes for all involved
3) The game should lead to good outcomes even if poor strategy is used. Just in case, the best strategy should become obvious to everyone as soon as possible.
Many variations of the Prisoner’s Dilemma game meet the first criteria, which is a necessary condition for improving real-life decisions using the game analogy.
But many of the most interesting variations fail the second two criteria, often spectacularly. So why not try to find a game that meets these three characteristics as much as possible, and try to apply that analogy as broadly as possible? Isn’t that a better way to use a game to improve real life? A game’s descriptive power isn’t enough. It should be prescriptive too. If you want to get good outcomes, find frameworks that lead to good outcomes and apply them where they are descriptive and aren’t being used.
How? Let’s just observe the Prisoner’s Dilemma objectively. Prison Guards do that, and their perspective might help. Presumably they have some Dilemmas. When you stop to think about it, they probably have a lot of dilemmas. Here’s a caricature of the dynamics of a prison:
The average guard’s most obvious dilemma is that he’s unhappy. Let’s assume it’s because he has to spend time in an awful environment. That would make me unhappy. To make things easier we’ll only look at strategies that end well, since those are the ones we’ll want to apply to real life. How could the guard succeed in changing prison conditions? Looked at this way, the guard’s dilemma isn’t much of a dilemma. The easiest way to get them changed would be to make the civilians aware of how awful prison conditions are. More specifically, convince the people who can change the conditions to change them.
I have found this game analogy to be both descriptive and prescriptive. And easy to use.
I have found this game analogy to be both descriptive and prescriptive. And easy to use.