Econdiscussion
  • Home
  • Articles
  • About Me

Articles

5. "Game Theory: Prison Breakthrough" - The Economist

5/12/2016

0 Comments

 
Synopsis: The fifth in The Economist’s series on seminal economic works discusses the theories behind the Nash equilibrium, how it changed mainstream economics and its relevance outside economics.

Click here to read the original article.
Discussion:

What is game theory?

Game theory describes a situation in which a finite number of players analyze not only his or her own decisions, but also the decisions of the other players, before acting in a way that optimizes his or her utility.

Take the most famous example of game theory, the Prisoner’s Dilemma. In this scenario, two mobsters are held in separate cells, each one being interrogated by a policeman. They are both given the same deal. If both of them confess to the murder individually, they each get ten years in prison. If one of them confesses, then he gets to go free while the other one is imprisoned for life. If neither of them confesses, they both spend a year in jail.
​
The decision diagram is shown below:
Picture
In this diagram, Mobster 1’s payoffs are indicated in blue while Mobster 2’s payoffs are indicated in red. To make a decision about whether or not to confess, they must both consider the other person’s actions. Note that life sentence is indicated by 100 years in prison.

A game theory scenario must satisfy a few conditions:
  1. There are a finite number of players who face a finite number of decisions;
  2. Players are rational, i.e. they will do what is best for them;
  3. Players cannot collaborate to settle on the best outcome.

Nash Equilibrium

The Nash equilibrium describes which scenario the players will settle at. Let us consider the Prisoner’s Dilemma again.

Take Mobster 1. If Mobster 2 confesses, it is better for Mobster 1 to confess rather than not confess, as the former will get him only 10 years in prison while the latter lands him a lifetime. If Mobster 2 does not confess, it is better for Mobster 1 to confess rather than not confess, as the former will let him avoid jail entirely while the latter means he is in jail for a year. Thus, whatever Mobster 2 decides to do, it is in Mobster 1’s best interest to confess.

If we take Mobster 2 now, the reasoning is exactly the same, and it is in Mobster 2’s best interest to confess.

Thus, both mobsters will confess and they will each spend 10 years in prison.

Despite the fact that if they had both not confessed, they would each have received a much lighter sentence, they both did what was best for themselves individually. Herein lies the importance of the Nash equilibrium: the optimal action of an individual does not optimize utility in a society.

Dominant, dominated, and mixed strategies

For only a single Nash equilibrium to exist, there must be either a dominant or a dominated strategy.

A dominant strategy is one where regardless of what the other player decides to do, a certain action is always better for a player. For example, regardless of what Mobster 2 decided to do, it is always in Mobster 1’s best interest to confess.

A dominated strategy is the exact opposite: regardless of what the other player decides to do, a certain action is always worse for a player. In the Prisoner’s Dilemma scenario, not confessing is always worse for Mobster 1.

It is important to note the distinction between a dominant and a dominated strategy. In a two-player game, if there is a clear dominant strategy, then the opposite decision is the clear dominated strategy. However, once a game expands past two players, a dominated strategy might be clear while a dominant strategy is not.

Some games have neither a dominant nor a dominated strategy. These games are called mixed strategy games. Take, for example, a game of football. A footballer who is about to score a goal faces a goalie, who tries to prevent a goal from being scored. This is shown in the decision diagram below.
Picture
Here, there is neither a dominant nor a dominated strategy. Take the footballer trying to score a goal, whose payoffs are indicated in blue. If the goalie dives left, it is in the footballer’s best interest to kick right. If the goalie dives right, the footballer should kick left. There is no dominant strategy.
Similarly, for the goalie, if the footballer kicks left, the goalie should also dive left. If the footballer kicks right, the goalie should dive right. There is no dominant strategy here.

Resultantly, there is no single Nash equilibrium. Instead, we find two Nash equilibria, depending on the decisions each player takes.

Game theory in the real world – Braess’ Paradox

Let us consider a real-world example of game theory: the construction of a highway in a transport system.

Braess’ Paradox holds that, counterintuitively, adding a highway to a transport system will increase travel time for an individual. To read about Braess’ paradox in full, click here. I will be outlining the basics of the paradox.
Picture
Consider a transport system with two highways that take 30 minutes each, regardless of the amount of vehicles on them, and two roads that take x/100 minutes each, where x is the amount of vehicles on the road. To get from A to D, there are two different paths – an individual can either travel through B or C.

Assume 4000 vehicles using the transport system. The Nash equilibrium is that 2000 vehicles would go through route B while 2000 go through route C. Why is this the Nash equilibrium?

Assume 3000 cars go through route B; consequently, there are 1000 cars travelling through route C. It takes people on route B [(3000/100)+30]=60 minutes to get to point D. It takes people on route C [(1000/100)+30]=40 minutes to get to point D. Thus, people on route B will switch over to route C to reduce traveling time.
​
Recall that the Nash equilibrium is when an individual does not want to change their choice of action, as they have nothing to gain from doing so. This will only happen when travel time between the two routes is equal, i.e. when there is an equal amount of traffic on either route. Thus, there will be 2000 vehicles on either route, and it would take them 50 minutes to reach point D.
Now, assume a highway has been added between point C and D, as illustrated below:
Picture
Now, assume that travel time on the highway is 0 minutes (this is, of course, an exaggeration). Let us assume that the highway is being used. Thus, to travel from A to D, an individual has two choices:

A to B, B to C, and then C to D, or A to C, C to B and then B to D.

Let’s consider the former with 1000 vehicles. It would take everyone [(4000/100)+0+(4000/100)]=80 minutes to get to point D.

For the latter case, it takes [30+0+30]=60 minutes to get to point D. Thus, individuals will choose to travel from A to C to B to D.

Now, some vehicles will choose to take the first path, while others take the second. The Nash equilibrium will be when the time travelled in both paths is equal, i.e. it takes 60 minutes to travel by either path.

This will happen when there are 300 vehicles on the first path. Note, however, that the use of the highway has increased travel times for all individuals by 10 minutes.

Why the Nash equilibrium matters

The Nash equilibrium assumes many things, including rational behavior and an ability to successfully predict a player’s payoffs. The biggest thing it assumes, however, is perfect information. In other words, a player knows exactly how another player will act given any scenario in the game. This, of course, is not the case in reality.

The Nash equilibrium has been adjusted by many others since John Nash himself. One revision includes uncertainty about people’s payoffs. Another includes the knowledge of non-credible threats, i.e. knowing whether another player will actually act in a certain way or whether he is just claiming that he will.

Despite all its flaws, the Nash equilibrium is one of the more revolutionary findings in economics for a few reasons. Firstly, the Nash equilibrium shows that individuals do not just weigh their own costs to maximize utility – they look at other players in the market to do so as well. Secondly, it shows that a utility maximizing individual does not necessarily maximize utility for a society. In perfectly competitive markets, for example, an individual’s choice of selling their product for the price at which it costs to produce it maximizes utility not only for the seller, but for the collective market as well.
​
The third way in which the Nash equilibrium is important is through the lens of government policy. Take a game theory scenario in which the dominant strategy for an oil company is to pollute a river. Taxation, or other methods of decreasing the payoff, shifts the company’s dominant strategy to not polluting at all. In this way, the Nash equilibrium is another way to view equilibrium analysis.
0 Comments

Your comment will be posted after it is approved.


Leave a Reply.

    Categories

    All
    Abenomics
    Adam Smith
    Adverse Selection
    Ageing
    Amazon
    Arrows
    Arthur Okun
    Asset
    Asymmetric Information
    Ben Bernanke
    Booms And Busts
    Braess' Paradox
    Brazil
    Bretton Woods
    Brexit
    Bubble
    Bull Market
    Business Cycle
    Capital Control
    Capital Flows
    Capital In The Twenty First Century
    Capital In The Twenty-First Century
    Carry Trade
    Causation
    Central Bank
    China
    Christopher Sims
    Classical Economics
    Consumption
    Counter-cyclical
    CPI
    Creative Destruction
    Crisis
    Daron Acemoglu
    David Cameron
    David Ricardo
    Debt
    Debt-to-GDP
    Deflation
    Deleverage Cycle
    Demography
    Devaluation
    Developing Economies
    Development
    Diffusion
    Diminishing Returns
    Dominant Strategy
    Dominated Strategy
    ECB
    Economic History
    Economic Theory
    Equality
    Equitity
    Equity
    Equity Investments
    EU
    Exchange Rate
    FDI
    Fed
    Federal Reserve
    Finance
    Financial Crisis
    Financial Instability Hypothesis
    Financial-instability Hypothesis
    Financial Times
    Fiscal Multiplier
    Fiscal Policy
    Fixed Exchange Rate
    Fleming
    Floating Exchange Rate
    Free Market
    Free Trade
    Freshwater
    Game Theory
    GDP
    George Akerlof
    Germany
    GFC
    Gini Coefficient
    Global Financial Crisis
    Globalization
    Government Intervention
    Government Spending
    Great Depression
    Growth
    Heckscher-Ohlin
    Helene Rey
    Hyman Minsky
    ICOR
    Illiquid
    Immigration
    Income
    Inequality
    Inflation
    Infrastructure
    Innovation
    Interest Rates
    Investment
    Italy
    James Robinson
    Janet Yellen
    Japan
    Jean Tirole
    J.M. Keynes
    John Nash
    Keynesian Economics
    Labor
    Lawrence Summers
    Leverage
    Lindau
    Liquid
    Malaysia
    Managed Exchange Rate
    Mario Draghi
    Matteo Renzi
    Michael Spense
    Minsky Moment
    Mixed Strategy
    Monetary Policy
    Monopoly
    Monopsony
    Moral Hazard
    Mundell
    Mundell-Fleming Trilemma
    Nash
    Nash Equilibrium
    Nigeria
    Nobel
    Nobel Laureates
    Nobel Prize
    Paul Krugman
    Paul Samuelson
    Perfect Information
    Phillips Curve
    Politics
    Poverty Traps
    Principal Agent Problem
    Prisoner's Dilemma
    Productivity
    Protectionism
    QE
    Qualitative Easing
    Quantitative Easing
    Redistribution
    Regulation
    Retrenching
    Rich
    Risk
    Robert Shiller
    Saltwater
    Saving
    Secular Stagnation
    Shiller
    Shinzo Abe
    Signaling
    Stakeholders
    Stolper-Samuelson Theorem
    Strategy
    Subsidies
    Tariff
    Taxation
    Taxes
    The Economist
    The Market For Lemons
    Theory Of Comparative Advantage
    Thomas Piketty
    Total Factor Productivity
    Trade
    Trilemma
    UK
    Unemployment
    U.S.
    USD
    Wage Benefits
    Wages
    Wealth
    Wealth Effect
    Wolfgang Stolper
    WTO
    Yuan

    Author

    JANANI DHILEEPAN
    A gap year student trying to explore real-world economics

    Archives

    May 2018
    February 2018
    January 2018
    December 2017
    November 2017
    April 2017
    February 2017
    December 2016
    October 2016
    September 2016
    August 2016
    July 2016
    May 2016
    March 2016
    January 2016
    November 2015
    September 2015
    August 2015
    July 2015
    June 2015
    April 2015
    March 2015
    February 2015
    January 2015
    December 2014
    November 2014
    October 2014
    September 2014

    RSS Feed

Proudly powered by Weebly