Game Theory: Nash Equilibria, Strategic Dilemmas, Dynamic Games, and Information Asymmetry | The Journal

Game Theory is the mathematical study of strategic interaction where the outcome for each participant depends critically on the choices of all other participants. Whether analyzing nuclear deterrence during the Cold War, auction design in spectrum licensing, corporate pricing wars, or algorithmic trading competition, game theory provides the formal vocabulary to dissect multi-agent decision systems.

This guide explores the foundational solution concepts, classic behavioral dilemmas, dynamic extensive-form games, and information asymmetries that govern strategic reality.

1. Core Solution Concepts

What is a Nash Equilibrium?

The Core Definition: Formulated by Nobel laureate John Nash in 1950, a Nash Equilibrium is a profile of strategies $(s_1^*, s_2^*, \dots, s_n^*)$ in which no player has an incentive to unilaterally deviate from their chosen strategy:

$u_i(s_i^*, \mathbf{s}_{-i}^*) \ge u_i(s_i, \mathbf{s}_{-i}^*) \quad \forall s_i \in \mathcal{S}_i, \quad \forall i$

Where:

→ $u_i$ is player $i$ 's payoff function.
→ $s_i^*$ is player $i$ 's equilibrium strategy.
→ $\mathbf{s}_{-i}^*$ represents the strategies chosen by all other players.

In plain language: given what everyone else is doing, each player is making the absolute best decision they can. It represents a state of strategic stability.

Crucial Caveat: A Nash equilibrium does not imply that the outcome is optimal, fair, or efficient for the players. As demonstrated by the Prisoner's Dilemma, a Nash equilibrium can be Pareto-inferior to unchosen cooperative outcomes.

What is a Dominant Strategy vs. a Dominated Strategy?

→
Strictly Dominant Strategy: A strategy that yields a strictly higher payoff than any other alternative strategy, regardless of what actions other players choose: $u_i(s_i^{\text{dom}}, \mathbf{s}_{-i}) > u_i(s_i', \mathbf{s}_{-i}) \quad \forall s_i' \ne s_i^{\text{dom}}, \quad \forall \mathbf{s}_{-i}$ Strategic Rule: If a player has a strictly dominant strategy, a rational actor must always play it.
→
Strictly Dominated Strategy: A strategy that always performs worse than another strategy, no matter what opponents do: $\exists \tilde{s}_i \quad \text{such that} \quad u_i(\tilde{s}_i, \mathbf{s}_{-i}) > u_i(s_i^{\text{undesirable}}, \mathbf{s}_{-i}) \quad \forall \mathbf{s}_{-i}$ Strategic Rule: Rational agents will never play a dominated strategy. The systematic elimination of these strategies—known as Iterated Elimination of Strictly Dominated Strategies (IESDS)—is a primary technique for simplifying and solving complex normal-form games.

What is the difference between Pure and Mixed Strategies?

→
Pure Strategy: A deterministic decision where a player selects a single specific action from their strategy set with 100% certainty (e.g., always choosing "Defect" or always playing "Rock").
→
Mixed Strategy: A probabilistic strategy where a player randomizes across two or more pure strategies according to a specific probability distribution $\sigma_i \in \Delta(\mathcal{S}_i)$ : $\sigma_i = (p_1, p_2, \dots, p_k), \quad \sum_{j=1}^k p_j = 1, \quad p_j \ge 0$ Why Randomize? In zero-sum competitive games without a pure Nash equilibrium (like Rock-Paper-Scissors or penalty kicks in soccer), playing a predictable pure strategy allows an opponent to exploit you. Randomizing choices keeps the opponent indifferent between their options, preventing exploitation.
- →Nash’s Existence Theorem: Every finite game (with a finite number of players and finite pure strategies) has at least one Nash Equilibrium, either in pure or mixed strategies.

2. Famous Dilemmas & Strategic Games

What is the Prisoner's Dilemma, and why is it significant?

The Prisoner’s Dilemma (framed by Merrill Flood, Melvin Dresher, and Albert Tucker) is the quintessential model of the tension between individual rationality and collective welfare.

The Scenario: Two suspects are arrested and placed in separate interrogation cells. Each suspect has two choices: Cooperate (stay silent) or Defect (confess and betray the partner).

Payoff Matrix (Years in Prison):

	Player 2 Cooperates	Player 2 Defects
Player 1 Cooperates	(-1, -1) [Mutual Cooperation]	(-10, 0) [Sucker's Payoff]
Player 1 Defects	(0, -10) [Temptation]	(-5, -5) [Nash Equilibrium]

→
The Analytical Logic:
- →If Player 2 stays silent, Player 1 gets 0 years by defecting vs. 1 year by staying silent $\implies$ Defect.
- →If Player 2 defects, Player 1 gets 5 years by defecting vs. 10 years by staying silent $\implies$ Defect.
- →Defection is a strictly dominant strategy for both players.
→The Paradox: Both players defect, serving 5 years each. Yet, if both had cooperated, they would have served only 1 year each.
→Significance: It mathematically proves that rational individuals pursuing self-interest will often arrive at an outcome that leaves everyone strictly worse off. It explains the tragedy of the commons, cartel breakdowns, price wars, and global carbon emission failures.

What is the Hawk-Dove Game (Game of Chicken)?

The Hawk-Dove Game (developed by John Maynard Smith and George R. Price in evolutionary biology) models escalating conflict and brinkmanship over a contested resource of value $V$ , with injury cost $C$ (where $C > V$ ).

→Hawk Strategy: Fight aggressively until injured or the opponent retreats.
→Dove Strategy: Display peacefully, but retreat immediately if the opponent escalates.

Payoff Matrix:

	Opponent plays Dove	Opponent plays Hawk
Player plays Dove	$(\frac{V}{2}, \frac{V}{2})$	$(0, V)$
Player plays Hawk	$(V, 0)$	$(\frac{V - C}{2}, \frac{V - C}{2})$ [Mutual Destruction]

→The Strategic Logic: If the opponent plays Dove, you should play Hawk (win entire resource $V$ ). If the opponent plays Hawk, playing Hawk leads to catastrophic injury ( $\frac{V - C}{2} < 0$ ), so you should play Dove.
→Equilibria: There are two asymmetric pure Nash equilibria: $(\text{Hawk}, \text{Dove})$ and $(\text{Dove}, \text{Hawk})$ , plus a mixed-strategy equilibrium where players fight with probability $p = \frac{V}{C}$ .
→Real-World Application: Brinkmanship, sovereign debt negotiations, and nuclear standoffs. The winning strategy often involves visibly destroying your own steering wheel (tying your own hands) to credibly force the opponent into playing Dove.

3. Dynamic & Informational FAQs

What is a Subgame Perfect Nash Equilibrium (SPNE)?

In sequential (dynamic) games, players make moves over time in a game tree (extensive-form game), observing previous moves.

A standard Nash equilibrium in an extensive-form game can involve incredible threats—threats that a player would never actually carry out if the game node were genuinely reached.

The Refinement (Reinhard Selten, 1965): A strategy profile is a Subgame Perfect Nash Equilibrium (SPNE) if it induces a Nash equilibrium in every single subgame of the overarching game tree.

How to Solve for SPNE: Backward Induction

→Start at the terminal leaf nodes of the game tree.
→Determine the optimal rational move for the player acting at that final stage.
→Prune suboptimal branches, replacing that subgame with its terminal payoff.
→Work backward through each decision stage to the root node.

Backward induction filters out bluffs and guarantees dynamic credibility.

What is Information Asymmetry: Moral Hazard vs. Adverse Selection?

Information asymmetry occurs when one party in an economic transaction holds material information that the other party lacks.

Concept	Timing	Nature of Asymmetry	Canonical Example	Market Solution
Adverse Selection	Pre-Contractual (Ex-ante)	Hidden Characteristics / Types: Unobservable quality prior to signing an agreement.	Market for Lemons (Akerlof): Low-quality used cars dominate the market because buyers cannot verify quality and only offer average prices, driving out good cars. Life Insurance: High-risk individuals are more motivated to purchase insurance.	Signaling (e.g., educational degrees, warranties) and Screening (e.g., insurance deductibles).
Moral Hazard	Post-Contractual (Ex-post)	Hidden Actions / Effort: Unobservable behavior or lack of care after a contract is executed.	Car Insurance: Drivers take greater risks or park carelessly knowing total losses are covered by the insurer. Executive Bailouts: Banks taking excessive risks believing they are "too big to fail."	Incentive Alignment: Deductibles, copays, performance-based bonuses, and direct monitoring.

4. Structural Comparison: Types of Strategic Games

Game Type	Core Feature	Strategic Dynamic	Canonical Example
Zero-Sum	One player’s gain is exactly equal to another’s loss: $\sum u_i = 0$ . Wealth is strictly redistributed.	Pure conflict. No mutual gain is possible. Minimax theorem applies.	Poker, Chess, Futures trading.
Non-Zero-Sum	Payoffs do not sum to zero. Players can win together (win-win) or lose together (lose-lose).	Cooperation, coordination, and mutual destruction are possible.	International trade, Climate treaties, R&D alliances.
Simultaneous	Players make their moves at the exact same moment without knowing competitors' choices.	Represented via payoff matrices (normal form). Hidden intentions.	Rock-Paper-Scissors, Sealed-bid auctions.
Sequential	Players take turns making moves, observing preceding choices before acting.	Represented via game trees (extensive form). Analyzed via backward induction.	Tic-Tac-Toe, Chess, Corporate market entry decisions.

Strategic Takeaways for Practitioners

→Anticipate Counter-Responses: In competitive environments, never evaluate your move in isolation; evaluate how competitors will rationally counter your move.
→Commitment Over Flexibility: Counterintuitively, limiting your future options (credible commitments, burning boats) often increases bargaining power in brinkmanship games.
→Align Incentives Against Information Asymmetry: When designing contracts, structure mechanisms (deductibles, equity vesting) that neutralize both pre-contract adverse selection and post-contract moral hazard.