The Core Definition of Nash Equilibrium (NE)
In the world of Game Theory, Nash Equilibrium (NE)
is the "point of no return." Named after the mathematician
John Nash (popularized by the film A Beautiful Mind), it describes a
state where no player can improve their outcome by changing their own strategy
while everyone else stays the same. Think of it as a stalemate of mutual best interests.
1. The Core Definition
A set of strategies is a Nash Equilibrium if, for every player,
the strategy they are choosing is the Best Response to the
strategies chosen by all other players.
·
The
"No Regrets" Test: If you were told what your opponents were going to do before
the game started, and you still chose the same move you originally planned, you
are in a Nash Equilibrium.
·
Individual
vs. Group: Crucially, a Nash
Equilibrium does not mean the group is getting the
best possible result (as seen in the Prisoner's Dilemma). It only means that no individual has an incentive to
deviate.
2. Types of Nash Equilibrium
A. A. Pure Strategy Equilibrium
This is when players choose one specific action and stick to it.
·
Example: In a "Coordination Game" (like
driving on the left vs. right side of the road), everyone driving on the right
is a Pure Strategy NE. No single driver
wants to switch to the left alone—that’s a head-on collision.
B. Mixed Strategy Equilibrium
In some games, there is no single "best" move. If you
always do the same thing, your opponent will predict you.
·
The
Logic: Players randomize their moves
based on probabilities.
·
Example: In Rock, Paper, Scissors, the Nash
Equilibrium is to play each hand exactly $1/3$ of the time. If you played Rock $50\%$ of the time, your
opponent would just play Paper every time, and you would lose.
Famous Examples
The Battle of the Sexes (Coordination)
A couple wants to go out. Player A prefers the Opera; Player B
prefers a Boxing Match. However, they both prefer being together over being alone.
·
NE
1: Both go to the Opera.
·
NE
2: Both go to the Boxing
Match.
·
Interpretation: There are two equilibria. The
"game" is about communicating to settle on one, even if one person is
slightly less happy than the other.
The Cournot Model (Economics)
Two firms decide how much of a product to produce. If Firm A produces a lot, the market price drops.
·
The
NE: They reach a point
where Firm A’s production level is the best response to Firm B’s production
level, and vice versa. Neither firm will change their output because doing so
would lower their specific profit.
4. Why Nash Equilibrium is "Large"
(The Impact)
The discovery of NE fundamentally changed how we understand
human systems:
1. Economics: It broke the old Adam Smith idea that "individual greed
always leads to the best social outcome." NE proved that individual greed
can lead to stuck, inefficient outcomes (like pollution or price wars).
2. Evolutionary Biology: Biologists use NE to explain why animals
develop certain traits (like the length of a peacock's tail) that seem
"irrational" but are actually stable strategies against competitors.
3. Artificial Intelligence: Modern AI (like AlphaGo or poker-playing
bots) is trained to find the Nash Equilibrium of a game to ensure they cannot
be "exploited" by human players.
5. Limitations of the Concept
·
Multiple
Equilibria: Some games have
dozens of NEs. The theory doesn't always tell us which one people
will pick.
·
Rationality
Assumption: It assumes players
are perfectly rational and have infinite "brainpower" to calculate
the moves of others.
·
Predictive
Failure: In real life, humans
often cooperate more than Nash Equilibrium predicts (due to altruism or social
norms).
Summary Checklist
· No Incentive to Change: If I change and you don't, I
get worse off.
· Mutual Best Response: I am doing the best I can, given
what you are doing.
· Stability: Once a group reaches a Nash Equilibrium,
they tend to stay there.
To find a Nash Equilibrium in a 2x2 matrix, we use the Underlining Method (also called the "Best
Response" method). This is the most reliable way to ensure you don't miss
an equilibrium, especially when there are multiple.
Let’s use a classic Coordination Game
(like two firms choosing between two different technology standards: 5G or Fiber).
The Payoff Matrix
The first number in each cell is Firm A's payoff, and
the second is Firm B's.
|
Firm B: 5G
|
Firm B: Fiber
|
|
Firm
A: 5G
|
(10, 10)
|
(0, 0)
|
|
Firm
A: Fiber
|
(0, 0)
|
(5, 5)
|
Step 1: Find Firm A’s Best Responses
Assume Firm B makes a choice, and Firm A reacts. We underline
Firm A's highest payoff in each column.
1. If Firm B chooses 5G: Firm A compares 10 (from choosing
5G) vs. 0 (from choosing Fiber). Since 10 is higher, we underline 10.
2. If Firm B chooses Fiber: Firm A compares 0 (from choosing 5G)
vs. 5 (from choosing Fiber). Since 5 is higher, we underline 5.
Step 2: Find Firm B’s Best Responses
Now, assume Firm A makes a choice, and Firm B reacts. We
underline Firm B's highest payoff in each row.
1. If Firm A chooses 5G: Firm B compares 10 (from choosing
5G) vs. 0 (from choosing Fiber). Since 10 is higher, we underline 10.
2. If Firm A chooses Fiber: Firm B compares 0 (from choosing 5G)
vs. 5 (from choosing Fiber). Since 5 is higher, we underline 5.
Step 3: Identify the Nash Equilibrium
A cell is a Nash Equilibrium if both numbers are underlined.
This represents a mutual best response.
In this game, we have two Pure Strategy
Nash Equilibria:
·
NE
1: (5G, 5G) — Both firms get 10.
·
NE
2: (Fiber, Fiber) — Both firms get 5.
Interpretation of this Result:
·
Stability: If both firms are currently using 5G, neither
wants to switch to Fiber alone (they'd go from 10 to 0).
·
Inefficiency: Even though (5G, 5G) is better for everyone,
if the firms are "stuck" in the (Fiber, Fiber) equilibrium, they
might stay there because no single firm can move to 5G without the other. This
is called a Coordination Failure.
Advanced Tip: "Mixed" Strategies
If you have a game like Matching Pennies
where no cell has two underlines, it means there is no "Pure
Strategy" Nash Equilibrium. In that case, players must "mix"
their moves (play randomly) to remain unpredictable.
Let's try a classic Price Competition
problem. In this scenario, two firms (Firm A and Firm B) are deciding whether
to set a High Price or a Low Price.
If they both set a High Price, they split the market and make
good profits. If one undercuts the other with a Low Price, they steal all the
customers. If both set a Low Price, they have a price war and make very little.
The Practice Matrix
The payoffs (Profit A, Profit B) are as follows:
|
Firm B: High Price
|
Firm B: Low Price
|
|
Firm
A: High Price
|
(100, 100)
|
(0, 120)
|
|
Firm
A: Low Price
|
(120, 0)
|
(20, 20)
|
Your Turn: The "Underlining"
Challenge
Follow these three steps to find the Nash Equilibrium:
1. Check Firm A's Best Response: * If Firm B goes High, does Firm A
want 100 (High) or 120 (Low)?
o If Firm B goes Low, does Firm A
want 0 (High) or 20 (Low)?
2. Check Firm B's Best Response: * If Firm A goes High, does Firm B
want 100 (High) or 120 (Low)?
o If Firm A goes Low, does Firm B
want 0 (High) or 20 (Low)?
3. Find the "Double Underline": Which cell has both players choosing their
best response?
What does the result tell us?
This is a version of the Prisoner's Dilemma.
You will likely find that the Nash Equilibrium is (Low Price, Low Price).
The Interpretation:
Even though both firms would be much richer if they both stayed
at a High Price ($100$ each), the
individual incentive to "steal the market" by dropping the price ($120$) is too strong. Because both firms think this way,
they "trap" themselves in the Low Price outcome ($20$ each).
This is why governments often have Anti-Trust laws—to
prevent firms from talking to each other (colluding) to stay at the High Price
equilibrium!
Since you’ve nailed the basic "Price War" logic, let’s
add a layer of realism: Asymmetric Costs.
In the real world, firms aren't always equal. One firm might
have a "Competitive Advantage" (lower costs), which completely shifts
the Nash Equilibrium and can even drive a competitor out of the market.
The Scenario: The Low-Cost Leader
·
Firm
A (The Giant): Has a massive factory
and low costs.
·
Firm
B (The Boutique): Has higher production
costs.
If they both play Low Price, Firm A
still makes a small profit, but Firm B actually loses money (a
negative payoff).
The "Asymmetric" Matrix
|
Firm B: High Price
|
Firm B: Low Price
|
|
Firm
A: High Price
|
(100, 80)
|
(0, 100)
|
|
Firm
A: Low Price
|
(150, 0)
|
(50, -20)
|
Step-by-Step Underlining
1. Firm A’s Best Responses (Columns)
·
If Firm B goes High: Firm A compares 100 vs. 150. Underline 150 (Low Price).
·
If Firm B goes Low: Firm A compares 0 vs. 50. Underline 50 (Low Price).
Insight: Firm A has a Dominant Strategy. No matter what B does, A will always
pick "Low Price."
2. Firm B’s Best Responses (Rows)
·
If Firm A goes High: Firm B compares 80 vs. 100. Underline 100 (Low Price).
·
If Firm A goes Low: Firm B compares 0 vs. -20. Underline 0 (High Price).
Insight: Firm B does not have a dominant strategy. Their best move depends
entirely on A.
The Nash Equilibrium: (Low Price, High Price)
The only cell where both are playing their best response is (150, 0).
The Interpretation (The "So What?")
1. Predatory Pricing: Firm A knows that by choosing a Low Price,
they force Firm B into a corner.
2. Market Exit: In this equilibrium, Firm B makes $0$ profit. If this
continues, Firm B will likely exit the industry.
3. The "Strong" survives: This explains why companies like Amazon or
Walmart can maintain low prices—it’s a Nash Equilibrium that smaller
competitors cannot profitably "match" without going into the negative
($-20$).
Next Level: Sequential Games
In all these examples, we assumed players moved at the same time. But what if Firm A moves first, and Firm B
watches them before deciding?
This is called a Stackelberg Model (or
a Sequential Game).
Let’s look at the Sequential Game
(also known as the Stackelberg Model in economics).
In the previous examples, players moved at the same time. But in
the real world, one firm often moves first (the Leader), and the
other firm observes that move before deciding what to do (the Follower).
1. The Tool: The Game Tree (Extensive Form)
To solve these, we don't use a matrix; we use a Game Tree. We solve it using Backward
Induction (working from the end of the game back to the beginning).
The Scenario:
·
Firm
A (Leader) chooses to produce a Large or Small quantity.
·
Firm
B (Follower) sees Firm A's choice
and then chooses Large or Small.
The Payoffs:
1. If both produce Large: Market is
flooded. Payoffs: (20, 20)
2. If A produces Large and B produces
Small: A takes the market. Payoffs: (80, 40)
3. If A produces Small and B produces
Large: B takes the market. Payoffs: (40, 80)
4. If both produce Small: High prices,
niche market. Payoffs: (60, 60)
2. Solving with Backward Induction
Step 1: Look at the Follower (Firm B)
We go to the "leaves" of the tree.
·
If
Firm A chose Large: Firm B compares 20
(Large) vs. 40 (Small). B will choose Small.
·
If
Firm A chose Small: Firm B compares 80
(Large) vs. 60 (Small). B will choose Large.
Step 2: Look at the Leader (Firm A)
Firm A is smart. They know how Firm B will
react.
·
If A chooses Large, they know B will choose Small $\rightarrow$ A gets 80.
·
If A chooses Small, they know B will choose Large $\rightarrow$ A gets 40.
The Result:
Firm A chooses Large. Firm B is
forced to choose Small.
The Nash Equilibrium is (Large, Small) with
payoffs (80, 40).
3. The Interpretation: First-Mover Advantage
In sequential games, being the "Leader" often yields a
higher payoff.
·
Commitment: By moving first and picking
"Large," Firm A commits to a strategy.
·
Forcing
the Hand: Firm B would love to produce Large too, but they won't, because
they'd rather have 40 (Small) than 20 (War).
4. Subgame Perfect Nash Equilibrium (SPNE)
In advanced microeconomics, this specific type of equilibrium is
called Subgame Perfect. It means the equilibrium is
"credible."
A "Non-Credible" Threat: If Firm B said, "If you produce Large, I
will also produce Large and ruin us both!", Firm A would ignore them. Why?
Because once Firm A actually produces Large, it is no longer in Firm B's best
interest to commit suicide (20 is worse than 40).
Comparison Table: Simultaneous vs. Sequential
|
Feature
|
Simultaneous (Nash)
|
Sequential (Subgame Perfect)
|
|
Information
|
Players move in the dark.
|
Follower has perfect info of
Leader's move.
|
|
Logic
|
Best Response to a guess.
|
Backward Induction (Thinking
ahead).
|
|
Outcome
|
Often symmetrical.
|
Often favors the Leader
(First-Mover).
|
The Public Goods Game is a classic
microeconomic problem that explains why we often have dirty air, crowded parks,
and underfunded national defenses. It is essentially a "multi-player
Prisoner's Dilemma."
1. The Setup
Imagine a group of 4 neighbors. Each has $100. They can either:
·
Keep the money for themselves (Private benefit).
·
Contribute to a "Common Fund" to build a park
(Public benefit).
The Catch:
For every $1 put into the Common Fund, the "value" grows to $2
(social benefit), but that $2 is split equally among all 4
neighbors, regardless of who paid.
2. The Payoff Math
Let’s look at the return on a $1 contribution:
·
The $1 becomes $2.
·
The $2 is split 4
ways.
·
Result: You get $0.50 back for every
$1 you put in.
3. Finding the Nash Equilibrium
We use the "No Regrets"
test again:
·
If
I contribute $1: I lose $1 and gain
$0.50 back. My net loss is -$0.50.
·
If
I keep my $1: I have my $1, plus I
still get $0.50 for every $1 my neighbors put in.
The Strategy:
Regardless of what the neighbors do, your best individual move is to contribute $0.
If everyone follows their individual Nash Equilibrium, the
Common Fund gets $0. This is called The Tragedy of the Commons or the Free Rider Problem.
4. Interpretation: Individual vs. Social
Optimum
·
Nash
Equilibrium (Individualism): Everyone contributes $0. Total group wealth = $400.
·
Social
Optimum (Cooperation):
Everyone contributes $100. The $400 becomes $800. Total group wealth = $800.
The gap between $400 and $800 is the "cost of
non-cooperation."
5. Why do Public Goods exist at all?
Since the Nash Equilibrium is to "Free Ride," why do
we have parks and roads? Microeconomics offers three main solutions:
1. Government Intervention (Taxation): The government makes contribution mandatory.
You are "forced" to cooperate so everyone can reach the $800 outcome.
2. Social Norms / Reputation: In small groups, free riders are shamed or
excluded from future benefits (a form of the "Grim Trigger").
3. Altruism: Some people derive "warm glow" utility simply from
the act of giving, which changes their personal payoff matrix.
Summary Table: Market Failures
|
Term
|
Definition
|
Economic Result
|
|
Free
Rider
|
Someone who benefits without
paying.
|
Under-provision of the good.
|
|
Non-Excludable
|
You can't stop someone from
using it.
|
Leads to the Free Rider
problem.
|
|
Non-Rival
|
My use doesn't stop your use.
|
Marginal cost of an extra
user is $0$.
|
Here is a comprehensive Advanced Microeconomics Game
Theory Cheat Sheet. This is designed to help you quickly identify
the "type" of question on an exam and the mathematical logic required
to solve it.
1. Static Games (Simultaneous Move)
Key Tool:
Payoff Matrix & The Underlining Method.
·
Nash
Equilibrium (NE): A state where no
player can benefit by changing their strategy unilaterally.
o Method: Underline the best response for Player A in each column;
underline the best response for Player B in each row. Cells with two underlines
are NE.
·
Dominant
Strategy: A strategy that is
better than any other strategy, regardless of what the opponent does.
·
Prisoner’s
Dilemma: A game where the Nash
Equilibrium leads to a worse outcome for all players than mutual cooperation.
o Logic: $T > R > P > S$.
2. Dynamic Games (Sequential Move)
Key Tool:
Game Tree (Extensive Form) & Backward Induction.
·
Subgame
Perfect Nash Equilibrium (SPNE): An equilibrium that represents a Nash Equilibrium in every
"subgame" (branch) of the larger game.
·
Backward
Induction: Start at the end of
the tree. Determine what the "Follower" will do in every scenario,
then move back to the "Leader" to see which path they will choose
based on those reactions.
·
First-Mover
Advantage: The ability of the
leader to commit to a strategy that forces the follower into a less-optimal
(for them) position.
3. Repeated Games (Infinite Horizon)
Key Tool:
The Discount Factor ($\delta$).
·
Grim
Trigger Strategy: A strategy that
starts with cooperation but punishes a single defection with infinite
retaliation.
·
Stability
Condition: Cooperation is stable
if:
$$\delta \ge \frac{T - R}{T - P}$$
o If $\delta$ is high (close
to 1), players value the future and will cooperate.
o If $\delta$ is low (close
to 0), players are "impatient" and will defect.
4. Market Failures & Public Goods
Key Tool:
Social vs. Private Marginal Benefit.
·
The
Free-Rider Problem: In a public goods
game, the Nash Equilibrium is for everyone to contribute $0$, even though the social optimum is for everyone to
contribute fully.
·
Tragedy
of the Commons: Individual users,
acting independently according to their own self-interest, deplete a shared
resource (like overfishing).
·
Condition
for Public Good Provision:
(The sum of everyone's marginal willingness to pay must equal the marginal cost
of production).
5. Oligopoly Models (Cheat Table)
When games are applied specifically to firm competition:
|
Model
|
Choice Variable
|
Strategy
|
Equilibrium Outcome
|
|
Cournot
|
Quantity
|
Simultaneous
|
Between Monopoly and Perfect
Competition.
|
|
Bertrand
|
Price
|
Simultaneous
|
Price = Marginal Cost (Same
as Perfect Competition).
|
|
Stackelberg
|
Quantity
|
Sequential
|
Leader produces more and
earns more than Follower.
|
Final Tip:
If the Nash Equilibrium is not the best possible outcome for the
group (e.g., in a Prisoner's Dilemma or Public Goods game), highlight the Deadweight Loss or Coordination Failure.
0 Comments