Game Theory Assignment Submissions

Compilation of Theory and Numerical Responses

Submission 1: JEA

THEORY QUESTION

Define first-price sealed-bid auctions game.

Suppose vi denotes value player 𝑖 attaches to the object and v1>v2>v3> ......>vn.
Show that in a Nash equilibrium of a first-price sealed-bid auction, the two highest bids are at least v2 and at
most v1.
Show also that any action profile satisfying these conditions is a Nash equilibrium.

THEORY ANSWER

First price sealed bid auction:

Let there be n players and valuation of each of them be vi & its bids be bi.
The player who has the highest bid, he/she wins and pays the amount of their own bid and the payoff of that
player(j) will be,
(vj-bj)
and everyone else gets a payoff of 0.
If 2 or more than 2 players have the same bid, appropriate tie-breaking rule can be applied depending on how
you define such rule.

Given: v1>v2>.....>vn
If highest bid is less than v2 and it is not by P1then P1 has an incentive to deviate.
If highest bid is less than v2 and by P1then P2 has an incentive to deviate.
This proves that highest bid >= v2

If highest bid is greater than v1 winner gets negative payoff for sure, and hence that player will have an
incentive to deviate.
This proves that highest bid <= v1

Suppose 2 highest bids are not the same.

Suppose bi>bj like this.
Player i will have an incentive to bid something slightly higher than bj and he/she can still be the winner and
get a higher payoff.
Such a case should not happen in Nash Equilibrium.
Whenever we talk about Nash, the 2 highest bids should be the same.
This ensures that
v2 <= bi, bj <= v1
in case of Nash Equilibrium.

To prove that whenever

v2 <= bi = bj <= v1
we get a Nash Equilibrium.
let(b1,b2,...,bn)be the action profile satisfying this condition.
The 2 highest bidders must be P1 and P2,since any other possibility won’t work and the winner will have an
incentive to deviate.
We know that
v2 <= b1 = b2 <= v1

Suppose tie-breaking rule is such that in case of equal bids, player with higher valuation wins.
P1 has no incentive to deviate because any higher bid than b1 will reduce the payoff and any lower bid will
make him lose and get 0 payoff.
P2 has no incentive to bid anything higher and is indifferent in bidding anything lower.
Pi such that i = 3,4,..,n no player has any incentive to deviate since winning will mean negative payoff for
any of them and losing will give them the same payoff that they are getting right now.

Hence it is nash equilibrium.

NUMERICAL QUESTION

The question contains a diagram. Thus the pdf solution also includes the question.

NUMERICAL ANSWER

[Link]

Submission 2: [Link]

THEORY QUESTION

What does it mean for a payoff function to represent a player’s preferences? Explain why payoff numbers are
only ordinal and not cardinal.

THEORY ANSWER

A payoff function represents a player’s preferences when it assigns numbers to outcomes in such a way that
outcomes the player prefers more receive higher numbers. In other words, if a player prefers outcome a to
outcome b, then the payoff assigned to a must be greater than the payoff assigned to b. The payoff function
can therefore be considered a numerical representation of the player’s ranking over possible outcomes.
The key idea is that the payoff function captures the order of preferences, not their intensity. For example,
suppose a person prefers ice cream to cake and cake to salad. One way to represent these preferences is to
assign payoffs 3, 2, and 1 respectively. However, assigning 100, 10, and -5 would represent the same thing.
In both cases, the ranking remains the same: ice cream is preferred to cake, and cake is preferred to salad.
Therefore, many different functions can represent the same preference as long as they preserve the ordering.
Thus, the payoff functions are ordinal rather than cardinal. Ordinal means that only the ranking of outcomes
matters, whereas cardinal would imply that the numerical differences between payoffs are meaningful, which
is not the case for payoff functions.

NUMERICAL QUESTION

Two friends are spending two days at the beach and must decide independently whether to swim on the first
day. Each believes that the water is infested with sharks with probability π; if sharks are present, anyone who
swims is attacked and receives payoff -c, while not swimming yields 0, and safe swimming yields 1 per day.
If an attack occurs on the first day, both infer that sharks are present and do not swim on the second day; if
no attack occurs, they retain their prior belief and will swim on the second day only if -πc + (1 - π) >= 0. A
player who swims on the first day receives an expected payoff of -πc + 2(1 - π), regardless of the other
player’s action. Model this situation as a strategic game in which each player chooses whether to swim on
the first day, determine the mixed-strategy Nash equilibria as functions of c and π, and assess whether the
presence of a friend makes an individual more or less likely to swim on the first day.

NUMERICAL ANSWER

[Link]

Submission 3: Ipsum lorem

THEORY QUESTION

Explain the concept of iterated strict dominance. How is it used to simplify a game? Illustrate your answer
with an example and state its relation to Nash equilibrium.

THEORY ANSWER

Iterated strict dominance is a method used to simplify a game by repeatedly removing strategies that are
strictly dominated.
A strategy is strictly dominated if there exists another strategy that gives a player a higher payoff no matter
what the other player does. Since a rational player would never choose such a strategy, it can be eliminated
from the game.
Concept
Suppose player 1 has two strategies, S1 and S2. If S2 gives a strictly higher payoff than S1 for every possible
action of the opponent, then S1 is strictly dominated by S2. The same idea applies to all players.
In iterated strict dominance, we do not stop after removing one dominated strategy. After one round of
elimination, new strategies may become dominated in the reduced game. So we keep removing strictly
dominated strategies step by step until no more can be removed.
How it simplifies a game
This procedure reduces the number of strategies to consider and helps identify the strategically relevant
outcomes. It is often used to solve games more easily, especially when the original payoff matrix is large.
Example
Consider the following game:
(A,X)=(3,2), (A,Y)=(1,1), (A,Z)=(0,0), (B,X)=(2,3), (B,Y)=(2,2), (B,Z)=(1,1), (C,X)=(0,0), (C,Y)=(1,2),
(C,Z)=(4,4)
Step 1: Check for strictly dominated strategies for Player 1
Compare Player 1’s strategies:
A=(3,1,0)
B=(2,2,1)
C=(0,1,4)
No strategy is strictly dominated by another at this stage, because each one is better in some column.
Step 2: Check for strictly dominated strategies for Player 2
Player 2’s payoffs are:
X=(2,3,0)
Y=(1,2,2)
Z=(0,1,4)
Again, no strategy is strictly dominated at the beginning.
So in this example, no strategy can be removed in the first round.
Now consider a slightly modified game where one strategy is dominated:
(U,L)=(2,2), (U,R)=(1,1), (D,L)=(3,3), (D,R)=(0,0)
Here, for Player 1:
D gives payoffs 3 against L and 0 against R
U gives payoffs 2 against L and 1 against R
Neither strictly dominates the other, so no elimination occurs.
But if we change the matrix to:
(U,L)=(2,2), (U,R)=(2,1), (D,L)=(1,3), (D,R)=(0,0)
then D is strictly dominated by U, because:
against L: 2>1
against R: 2>0
So D is removed. After removing D, the reduced game may reveal further dominated strategies.
Relation to Nash equilibrium
Every strategy that is part of a Nash equilibrium must be a best response to the others’ strategies. A strictly
dominated strategy can never be a best response, so:
No strictly dominated strategy can be part of a Nash equilibrium.
Therefore, iterated elimination of strictly dominated strategies helps narrow down possible Nash equilibria.
However, the converse is not always true:
Not every surviving strategy after iterated elimination must be in a Nash equilibrium.
Also, a game may have Nash equilibria even when some strategies are not eliminated.
Important point
If a strategy is strictly dominated, it is irrational to play it, because there is always a better choice regardless
of what others do. That is why strict dominance is a very powerful solution concept.

Iterated strict dominance is a systematic way to simplify games by repeatedly removing strategies that are
always worse than another strategy. It reduces the game to strategically relevant options and often makes it
easier to find Nash equilibria. It is especially useful in games where many strategies can be ruled out by
rationality considerations.

NUMERICAL QUESTION

Two players find themselves in a legal battle over a patent. The patent
is worth 20 to each player so the winner would receive 20 and the loser 0.
Given the country's norms, it is common to bribe the judge hearing a case.
Each player can offer a bribe secretly, and the one whose bribe is the highest
will be awarded the patent. If both choose not to bribe, or if the bribes are the
same amount, then each has an equal chance of being awarded the patent. If
a player does bribe, then the bribe can be valued at either 9 or 20. Any other
number is considered very unlucky, and the judge would surely rule against a
party who offered a different number.
a. Find the unique pure-strategy Nash equilibrium for this game.
b. If the norm were different, so that a bribe of 15 were also acceptable,
is there a pure-strategy Nash equilibrium?
c. Find the symmetric mixed-strategy Nash equilibrium for the game
with possible bribes of 9, 15, and 20.

NUMERICAL ANSWER

[Link]

Submission 4: Mavericks

THEORY QUESTION

In the study of Perfect Bayesian Equilibrium (PBE), specifically within signaling games like the "Gift
Game" , we often encounter multiple equilibria, including pooling equilibria (where different types of
players send the same signal) and separating equilibria (where they send different signals)
A significant limitation of PBE is that it can support "unreasonable" equilibria based on arbitrary beliefs
about off-equilibrium path actions . Explain the concept of the "Intuitive Criterion" (by Cho and Kreps) as a
refinement of PBE. How does it use the logic of rationality to "prune" or eliminate certain PBE that rely on
"uncredible" beliefs?

THEORY ANSWER

A Perfect Bayesian Equilibrium requires that players' strategies are sequentially rational given their beliefs,
and that these beliefs are updated via Bayes' rule wherever possible. In signaling games like the Gift Game,
this often leads to multiple equilibria, specifically pooling equilibria (where different types of senders choose
the same action) and separating equilibria (where they choose different actions).
The "limitation" mentioned arises because PBE allows for nearly any belief to be held about off-equilibrium
path actions (actions that no type of player is supposed to take in equilibrium). If the receiver (e.g., the
person receiving a gift) holds a "threat" belief-such as assuming any unexpected action comes from a "bad"
type-they can sustain equilibria that might not make sense if players are truly rational.

The Concept of the "Intuitive Criterion"

Intuitive Criterion as a method to "prune" these unreasonable equilibria by applying a deeper level of the
rationality assumption.

The logic works as follows:

1. Identify a Deviation: Suppose a player deviates from the equilibrium path by sending a signal that was not
expected.

2. Evaluate Incentives: We look at each possible "type" of the sender. If a certain type (Type
A) would get a lower payoff by deviating regardless of how the receiver responds compared to what they get
in the current equilibrium, then it is irrational for Type A to have deviated.

3. Refine Beliefs: The Intuitive Criterion argues that the receiver should not believe the deviator is Type A.
Instead, the receiver should focus their beliefs on types who could potentially benefit from the deviation if
the receiver reacts favorably.

4. Eliminate the Equilibrium: If the original PBE only "works" because the receiver was supposed to believe
the deviator was Type A, that equilibrium is eliminated as it fails the Intuitive Criterion.

Application to the "Gift Game"

In the context of the Gift Game, the Intuitive Criterion would be used to see if a pooling equilibrium (where
everyone gives the same gift) is robust. If a "High-Quality" sender could distinguish themselves by giving a
slightly more expensive gift that a
"Low-Quality" sender would find too costly to mimic (even if the receiver thought they were high quality),
then the pooling equilibrium might be discarded in favor of a separating one. This ensures that the
equilibrium reflects common knowledge of rationality.

NUMERICAL QUESTION

Consider a Stackelberg Leadership Model with Incomplete Information.

• Firm 1 (Leader) chooses quantity q1 first.
• Firm 2 (Follower) observes q1 and then chooses q2.
• Inverse Demand: P = 100 - (q1 + q2).
• Costs: Firm 1 has a known marginal cost MC1 = 10. Firm 2's marginal cost is private
information. It is either High (MCH = 30) with probability 0.4, or Low (MCL = 10) with
probability 0.6.

Calculate the Optimal Quantity q1 for the Leader, given that they must maximize their expected profit while
anticipating the different best-responses of Firm 2's types.

NUMERICAL ANSWER

[Link]

Submission 5: DTP

THEORY QUESTION

Subgame Perfect Equilibrium: Concept, Backward Induction, and the Problem of Non-Credible Threats

Unit III & IV

Game theory distinguishes between Nash Equilibrium and Subgame Perfect Equilibrium (SPE). While Nash
Equilibrium is a powerful solution concept for static games, it is often considered insufficient for analysing
dynamic games — games that unfold over time with sequential decisions. The concept of Subgame Perfect
Equilibrium was introduced precisely to address this inadequacy.

(a) What is an Extensive Form Game? How does it differ from a Normal Form Game? Explain why
extensive form games are better suited for capturing the structure of sequential decision-making.

(b) Define Subgame Perfect Equilibrium. What additional requirement does SPE impose over and above
Nash Equilibrium? Explain the concept of a "subgame" and why SPE requires rational behaviour not just on
the equilibrium path but also at every off-equilibrium-path node.
(c) What is a non-credible threat? Explain clearly — using a real-life or economic example of your choice
— how a non-credible threat can sustain a Nash Equilibrium that is not subgame perfect, and why backward
induction rules it out.

(d) Critics argue that Subgame Perfect Equilibrium, while theoretically elegant, is too demanding as a
solution concept in practice. Do you agree? Discuss the assumptions SPE rests on and the conditions under
which it is a reasonable or unreasonable predictor of real behaviour.

THEORY ANSWER

Part (a) — Extensive Form vs Normal Form Game

A Normal Form Game represents a strategic situation in a matrix where players choose strategies
simultaneously without knowing others’ actions. It is suitable for static games where timing does not matter.

An Extensive Form Game represents the game as a tree, showing the order of moves, players’ decisions,
information available at each stage, and resulting outcomes.

The key advantage of extensive form is that it captures sequential decision-making. In many real-world
situations (e.g., entry decisions, bargaining), who moves first and what is known at each step directly affects
outcomes. Hence, extensive form is better suited for analysing dynamic games.

Part (b) — Subgame Perfect Equilibrium (SPE)

A subgame is a part of an extensive form game that starts at a decision node and includes all subsequent
actions, forming a complete “mini-game.”

A Subgame Perfect Equilibrium (SPE) is a strategy profile that is a Nash Equilibrium in every subgame, not
just in the overall game.

Unlike Nash Equilibrium, which requires optimality only along the equilibrium path, SPE imposes sequential
rationality — players must act optimally at every decision node, including off-equilibrium paths.

SPE is obtained using backward induction, ensuring that strategies are credible and rational at every stage.

Part (c) — Non-Credible Threats and SPE

A non-credible threat is a threat that a player would not rationally carry out if the situation actually arises, as
it would harm their own payoff.

Example: An incumbent firm threatens a price war to deter entry. If the entrant stays out, the threat is never
tested, so the outcome can appear as a Nash Equilibrium.
However, if entry occurs, starting a price war is costly and irrational. The incumbent would prefer to
accommodate rather than incur losses. Thus, the threat is not credible.

Using backward induction, we analyse the final stage first and see that the firm will not fight. Anticipating
this, the entrant chooses to enter.

Hence, SPE eliminates equilibria based on non-credible threats and retains only those with credible
strategies.

Part (d) — Is SPE Too Demanding?

SPE assumes full rationality and common knowledge of rationality, meaning all players act optimally and
know others do the same.

In reality, players may be boundedly rational, influenced by emotions, fairness, or limited reasoning, and
may not follow backward induction (e.g., in long or complex games).

However, SPE remains a useful benchmark for analysing strategic behaviour, especially in structured
environments like markets and auctions.

Thus, while SPE may be demanding in practice, it is a powerful theoretical tool for eliminating non-credible
strategies and understanding strategic interactions.

NUMERICAL QUESTION

Two firms compete in a Cournot duopoly. The inverse demand function is P = 100 − Q, where Q = q1 + q2.
Both firms initially have constant marginal cost c = 10 and no fixed costs. Before the Cournot stage, Firm 1
has the option to invest in a cost-reducing technology at a cost K > 0. If Firm 1 invests, its marginal cost
drops to c1 = 5, while Firm 2's cost remains at c2 = 10. The game proceeds in two stages:

Stage 1: Firm 1 decides whether to Invest (I) or Not Invest (NI).

Stage 2: Both firms simultaneously choose quantities in Cournot competition.

(a) Solve for the Cournot Nash Equilibrium quantities and profits when Firm 1 does NOT invest (both firms
have c = 10).
(b) Solve for the asymmetric Cournot Nash Equilibrium when Firm 1 DOES invest (c1 = 5, c2 = 10).
(c) Find the critical investment cost threshold K* such that Firm 1 invests in the unique Subgame Perfect
Equilibrium (SPE) if and only if K < K*. Use backward induction.
(d) Construct a Nash Equilibrium of the full game (not SPE) in which Firm 1 does NOT invest even when K
< K*. Explain clearly why this equilibrium is Nash but not subgame perfect.
NUMERICAL ANSWER

[Link]

Submission 6: The Focus Point

THEORY QUESTION

Explain Coordination Game

THEORY ANSWER

A coordination game is a game in which players benefit by choosing the same or mutually consistent
strategies.

In such games, there are usually multiple Nash equilibria, and the best outcome depends on players
coordinating their actions.

For example, in the Stag Hunt game, both players get a higher payoff if they cooperate (hunt stag together)
rather than act individually.

Thus, coordination games emphasize the importance of mutual agreement and strategic alignment.

NUMERICAL QUESTION

Consider a Cournot duopoly with inverse demand function:

𝑃=100−𝑄
where
𝑄=𝑞1+𝑞2

Assume both firms have zero cost.

Find:

Reaction functions of both firms

Cournot–Nash equilibrium output and price

NUMERICAL ANSWER

[Link]

Submission 7: Pariket
THEORY QUESTION

All Subgame Perfect Equilibria (SPE) are Nash Equilibria, but not all Nash Equilibria are Subgame Perfect
Equilibria. Justify this statement with an example.

THEORY ANSWER

Why all SPE are NE: Since an SPE requires players to make rational moves in every possible part of the
game, it automatically ensures they are making rational moves in the game as a whole.

Why NOT all NE are SPE: A standard NE allows players to use "non-credible threats" (empty threats). This
means a player can threaten a bad outcome to scare their opponent into taking a different path. However,
SPE does not allow empty threats. SPE demands that if the player actually had to carry out the threat, it must
be their most logical and profitable move. If it hurts them to carry out the threat, it is not an SPE.

Example: The Market Entry Game

Players: {Entrant, Incumbent}
Strategy for Entrant : {Stay Out, Enter}
Strategy for Incumbent : {Fight, Accommodate}

The Rules & Payoffs (Entrant, Incumbent):

1. Stage 1: The Entrant chooses to Enter or Stay Out.
o If they Stay Out, the game ends. Payoffs: (0, 3).

2. Stage 2: If the Entrant Enters, the Incumbent must choose to Fight (start a price war) or Accommodate
(share the market).
o If they Fight, Payoffs: (-1, 1).
o If they Accommodate, Payoffs: (1, 2).

If we look at the game normally, there are two Nash Equilibria:

1. (Enter, Accommodate): The Entrant gets 1, Incumbent gets 2.
2. (Stay Out, Fight): The Entrant gets 0, Incumbent gets 3.

To find the SPE, we use backward induction and look at the "subgame" at Stage 2.
• Assume the Entrant has entered.
• The Incumbent now faces a choice: Fight (1) or Accommodate (2).
• Since getting 2 is better than 1, a rational Incumbent will always choose to Accommodate.
Because the strategy (Stay Out, Fight) relies on an irrational empty threat, it fails the subgame test. It is a
Nash Equilibrium, but NOT a Subgame Perfect Equilibrium. The only strategy that is rational in every single
subgame is (Enter, Accommodate).

NUMERICAL QUESTION
Question is in Answer pdf

NUMERICAL ANSWER

[Link]

Submission 8: Team Gambit

THEORY QUESTION

Explain the process of Iterated Elimination of Dominated Strategies (IEDS). Using a First-Price Sealed-Bid
auction as an example, explain why bidding an amount higher than your own valuation is a "weakly
dominated strategy." How does this elimination simplify the strategy set for a rational player?

THEORY ANSWER

IEDS is used to simplify a game by systematically removing "dominated" strategies.

Dominance: A strategy is strictly dominated if there is another strategy that always yields a higher payoff,
regardless of what the other players do.
The Process: Since a rational player will never choose a dominated strategy, we remove it from their strategy
set. Once removed, the "new" smaller game might reveal new strategies for other players that have now
become dominated. We repeat this process until no more strategies can be removed.

2. Bidding in a First-Price Auction, the highest bidder wins and pays exactly what they bid. Let’s say our true
valuation of the item is V.

Strategy: Bidding more than what you think the item is worth is a weakly dominated strategy because if we
lose: our payoff is [Link] we win: We pay B. Since B > V, our payoff is V - B, which is a negative value (a
loss).If we had bid exactly V or less, our payoff would either be 0 (if we lost) or greater than 0(if we won).
Since 0 is always better than a negative payoff, any bid B > V is dominated by a bid of B = V.

[Link] simplifies the strategy set for a rational player by reducing the infinite number of possible bids down
to a specific range:
Step 1: Eliminate all bids B > V. No rational person wants to win an item only to lose money on the
transaction. The player is left with the range [0, V].
Step 2 : In a First-Price auction, bidding exactly V results in a payoff of 0 (even if we win). Therefore, to get
a positive payoff, a rational player must "shade" their bid, choosing a strategy in the range [0, V).

NUMERICAL QUESTION
In a village, N farmers (where N=10) decide how many cows to graze on a common field. Each farmer i
chooses g_i > 0. The total number of cows is G = sum of g_i. The value of grazing a cow is V(G) = 100 - G.
The cost of buying and tending to a cow is C = 10.

Find the Nash Equilibrium number of cows each farmer will graze.
Calculate the Socially Optimum number of total cows (G) that maximizes the village's total profit.
Numerically demonstrate the "Tragedy of the Commons" by comparing the total cows in equilibrium vs. the
social optimum.

NUMERICAL ANSWER

[Link]

Submission 9: The CoderZ

THEORY QUESTION

Explain the concept of a Mixed Strategy Nash Equilibrium (MSNE). Why does a mixed strategy equilibrium
exist even in games where no pure strategy Nash equilibrium exists? Use the Matching Pennies game to
support your answer.

THEORY ANSWER

Pure Strategy vs. Mixed Strategy

In a pure strategy, a player always chooses the same action with certainty. In a mixed strategy, a player
randomises over their available actions by assigning a probability to each. The idea is that by mixing, a
player becomes unpredictable and prevents the opponent from exploiting a fixed pattern.
Definition of MSNE
A Mixed Strategy Nash Equilibrium is a set of mixed strategies, one for each player, such that each player’s
strategy is a best response to the other player’s strategy. No player has any incentive to change their
probability distribution unilaterally.
The key condition for an MSNE is the indifference condition: a player will only mix between strategies if
each strategy in the mix gives the same expected payoff. This condition is used to solve for the equilibrium
probabilities.
Why Does MSNE Always Exist?
According to Nash’s Existence Theorem (1950), every finite game with a finite number of players and
strategies has at least one Nash Equilibrium in mixed strategies. So even if no pure strategy equilibrium
exists, a mixed strategy equilibrium is always guaranteed to exist.
The intuition is that when players can’t settle on a pure strategy without wanting to deviate, randomising
creates a kind of stability. By making the opponent indifferent between their options, neither player has a
reason to deviate from their mixing probabilities.
Illustration: Matching Pennies
Two players simultaneously show a coin (Heads or Tails). Player 1 wins if the outcomes match; Player 2
wins if they differ. The payoff matrix is:
P2: Heads P2: Tails
P1 Heads (+1, −1) (−1, +1)
Tails(−1, +1) (+1, −1)

There is no pure strategy Nash Equilibrium in this game. In every cell, at least one player wants to switch —
so there is no stable pure outcome.
Finding the MSNE:
Let p = probability that Player 1 plays Heads. For Player 2 to be willing to mix, they must be indifferent
between Heads and Tails:
Expected payoff if P2 plays Heads = −1·p + 1·(1−p) = 1 − 2p
Expected payoff if P2 plays Tails = 1·p + (−1)·(1−p) = 2p − 1
Setting them equal: 1 − 2p = 2p − 1 ⟹ p = 1/2
By symmetry, Player 2 also plays Heads with probability q = 1/2.
MSNE: Both players play Heads with probability 1/2 and Tails with probability 1/2. The expected payoff for
both players is 0.
Conclusion
Mixed Strategy Nash Equilibria arise in games where pure strategy equilibria do not exist, particularly in
zero-sum or strictly competitive settings. By mixing optimally, players reach a stable state where no one
benefits from deviating. Nash’s theorem guarantees such equilibria always exist in any finite game.

NUMERICAL QUESTION

Two firms compete in a market with the inverse demand function P = 120 − 2Q, where Q = q₁ + q₂. Firm 1
has a marginal cost of ₹20 (TC₁ = 20q₁) and Firm 2 has a marginal cost of ₹30 (TC₂ = 30q₂).
Answer the following parts:
(a) Write the profit function for each firm and derive their Best Response Functions.
(b) Solve the Best Response Functions simultaneously to find the Cournot–Nash Equilibrium quantities q₁*
and q₂*.
(c) Find the equilibrium market price P*, total output Q*, and profits π₁ and π₂.
(d) Now assume Firm 1 is the Stackelberg leader and Firm 2 is the follower. Find the Stackelberg
equilibrium quantities, price, and profits.
(e) Compare both equilibria. Which model gives higher total output and lower price? What does this mean
for consumers?

NUMERICAL ANSWER

[Link]
 Submission 10: teamGTA6

THEORY QUESTION

A group project is assigned to A and B, each worth 10 points toward their final class grades. A and B submit
independent bids to work hard (W) or ride along (F) to complete the project while waiting for the group to
succeed or fail based on each member's individual effort.

When both A and B invest time and energy in completing their project, they will receive 8 points each for
their joint efforts, creating an outstanding project with a high level of quality and no extra cost for either
student.

Both students will receive 5 points each for producing a mediocre quality project because they both chose to
ride along rather than exert themselves. Alternatively, A will receive only 4 points if he or she chooses to
work while B chooses to ride, and vice versa for B, because only one of them chose to put in a considerable
amount of their time and energy into the project.

1. Identify and define the players, their respective strategy sets, their payoff structures (i.e., assume each
possible outcome).
2. Use the matrix developed in question 1 to find each player's dominant strategy. Consider each player's
perspective as they determine which strategy (i.e. work hard or free ride) to employ for various possible
outcomes.
3. Find a Nash equilibrium and determine whether or not the resulting payoffs are Pareto optimal.
4. Describe the common game structure between this game and the well-known prisoner's dilemma game.
Explain the commonality of the scenario between the prisoner's dilemma and this game.
5. Suggest two changes the instructor could implement in order to change the pay structure and create the
equilibrium of both A and B working hard.

THEORY ANSWER

(a) Game has 2 players, A and B, and their strategies are Sₐ = S_b = {Work Hard (W), Free Ride (F)}. The
following functions define each of their payoffs in accordance with different combinations of strategies
played.

Payoffs will be shown in the format (A's Payoff, B's Payoff).

u(W,W) = (8,8)
u(W,F) = (4,9)
u(F,W) = (9,4)
u(F,F) = (5,5)

(b) Since A's best choice is always to Free Ride, we can make some assumptions regarding A and B's
potential for coordinating their desires by analyzing the payoff matrix for both players.

If B chooses W, A will receive a payoff of 9 from F and 8 from W. Therefore A will choose to F.

If B chooses F, A will receive a payoff of 5 from F and 4 from W. Therefore A will have to F.

Thus, A has a strictly dominant strategy and will always choose to F regardless of what B chooses.

(c) The only Nash equilibrium of this game is (F,F) = (5,5). Since neither player can increase their respective
payoffs by selecting a different strategy in this Nash equilibrium, they have only one best response available
to them (which is to continue to play F). In this case, A can remain at 5 (given that B continues F) or drop to
4 (given that A switches to W while B plays F). The same holds true for B. Therefore, (F,F) is a Nash
equilibrium.

This Nash equilibrium is NOT Pareto optimal. The (W,W) combination, on the other hand, yields both
students higher payoffs than the (F,F) combination (8,8) > (5,5). A clear possibility exists for a Pareto
improvement for both players, but an inability to reach that Pareto optimum exists due to the rational self-
interests of both players.

(d) This game represents the Prisoner's Dilemma, or PD, form of game wherein the players have two
alternatives, C = Cooperate or D = Defect, and their payoffs satisfy the following order:

T > R > P > S; where T = Temptation, R = The reward for mutual cooperation, P = The reward for mutual
defection; S = The penalty for not cooperating.

NUMERICAL QUESTION

In a two-leg football final:

Team A first decides whether to play safe (S) or attack (A) in the home leg.
If Team A chooses attack, Team B observes this and then chooses Defend (D) or Counter (C).

The payoffs are:

(A, D) = (3, 1)
(A, C) = (1, 3)
(S) = (2, 2)
Questions
1. Draw the extensive form game tree.
2. Find the subgame perfect equilibrium using backward induction.
3. Compare it with the normal-form Nash equilibrium.
Team A moves first at the root (blue node). If they choose S (Play Safe), the game ends immediately with
payoff (2, 2). If they choose A (Attack), Team B observes this and responds at their decision node (coral),
choosing either D (Defend) → (3, 1) or C (Counter) → (1, 3).
Answer 2 : Subgame Perfect Equilibrium (Backward Induction)
We solve backwards : starting from the last decision node and working to the root.
Step 1 Team B's decision (the subgame after Attack):
Team B compares their own payoffs:
Choose D (Defend): Team B gets 1
Choose C (Counter): Team B gets 3
→ Team B will rationally choose Counter (C), since 3 > 1.
So if Team A attacks, the outcome will be (1, 3) Team A gets only 1.
Step 2 : Team A's decision at the root:
Now Team A folds this forward:
Choose S (Play Safe): Team A gets 2
Choose A (Attack): Team A anticipates Team B counters → Team A gets 1
Team A will rationally choose Play Safe (S), since 2 > 1.
Subgame Perfect Equilibrium (SPE):
{Team A: Play Safe (S); Team B: Counter if Attack (C)}
Equilibrium outcome: (2, 2)
This is the unique SPE found by backward induction.
Answer 3 : Normal-Form Nash Equilibrium (Comparison)
In normal form, we list all strategies. Team A has 2 strategies: S, A. Team B has 2 strategies: D, C (but these
are contingent plans for what they'd do if attacked).
Payoff Matrix:

Team B: Defend (D)

Team B: Counter (C)
Team A: Play Safe (S)
(2, 2)
(2, 2)
Team A: Attack (A)
(3, 1)
(1, 3)

Finding Nash Equilibria (best-response analysis):

NE 1 (S, D):
If Team B plays D, Team A's best response: S gives 2, A gives 3 → Team A prefers A (not S). This is NOT a
NE.
NE 2 (S, C):
If Team B plays C, Team A's best response: S gives 2, A gives 1 → Team A prefers S
If Team A plays S, Team B's payoff is 2 regardless of D or C → Team B is indifferent (any strategy is a best
response)
(S, C) is a Nash Equilibrium with outcome (2, 2)
NE 3 (A, D):
If Team A plays A, Team B's best response: D gives 1, C gives 3 → Team B prefers C (not D). This is NOT a
NE.

NUMERICAL ANSWER

[Link]

Submission 11: Grand-Theft-Auto (GTA)

THEORY QUESTION

Consider a third-price sealed-bid auction, which differs from a first and a second-price auction only in that
the winner (the person who submits the highest bid) pays the third highest price. (Assume that there are at
least three bidders.)
(a) Show that for any player 𝑖 , the bid of 𝑣(𝑖) weakly dominates any lower bid, but does not weakly
dominate any higher bid. (To show the latter, for any bid 𝑏(𝑖) > 𝑣(i) , find bids for the other players such that
player 𝑖 is better off bidding 𝑏(𝑖) than bidding 𝑣(i).
(b) Show that the action profile in which each player bids her valuation is not a Nash equilibrium.
(c) Find a Nash equilibrium. (There are ones in which every player submits the same bid.)

THEORY ANSWER

(a) The argument that a bid of 𝑣(i) weakly dominates any lower bid is the same as for a second-price
auction. Now compare bids 𝑏(i) > 𝑣(i) and 𝑣(i) . Suppose that one of the other players bids is between 𝑣(i)
and 𝑏(i) , and all the remaining bids are less than 𝑣(i) . If player 𝑖 bids 𝑣(i) , she loses and obtains payoff 0. If
she bids 𝑏(i) , she wins and pays the third highest bid, which is less than 𝑣(i).Thus she is better off bidding
𝑏(i) than bidding 𝑣(i) .

(b) Each player’s bidding her valuation is not a Nash equilibrium because player 2 can deviate and bid more
than 𝑣(1) and obtain the object at the price 𝑣(3) instead of not obtaining the object.

(c) Any action profile in which every player bids b , where 𝑣(2) ≤ 𝑏 ≤ 𝑣(1) , is a Nash equilibrium. If player
1 raises her bid, the outcome does not change. If she lowers her bid, her payoff becomes zero, which is no
higher than her payoff in the action profile. If any other player raises her bid, then she wins and pays 𝑏,
obtaining a nonpositive payoff. If any other player lowers her bid, the outcome does not change.

The set of all Nash equilibria is given as follows. (Note that the question asks only for one equilibrium, not
all equilibria.) A profile of bids is a Nash equilibrium if and only if it satisfies one of the following two
conditions: Player 1 wins, at least two players bid at least 𝑣(2) , and the third highest bid is at least 𝑣(k) ,
where k is the player who submits the second highest bid. Player 𝑘 wins for some 𝑘 with 2 ≤ 𝑘 ≤ 𝑛 − 1, at
least two players bid at least 𝑣(1) , the third highest bid is at most 𝑣 (k) , and the index of the player
submitting the second highest bid is greater than 𝑘.

NUMERICAL QUESTION

Consider the game when the inverse demand function is given by 𝑃(𝑄) = 𝛼 − 𝑄 for 𝑄 ≤ 𝛼 and 𝑃(𝑄) = 0 for
𝑄>𝛼.
For values of 𝑐(H) and 𝑐(L) close enough that there is a Nash equilibrium in which all outputs are positive,
find this equilibrium.
Compare this equilibrium with :
a. The Nash equilibrium of the game in which firm 1 knows that firm 2’s unit cost is 𝑐(L) .
b. The Nash equilibrium of the game in which firm 1 knows that firm 2’s unit cost is 𝑐(H) .

NUMERICAL ANSWER

[Link]

Submission 12: Nivadi

THEORY QUESTION

Explain Subgame Perfect Nash Equilibrium (SPNE) and Bayesian Nash Equilibrium (BNE). Compare their
roles in analyzing extensive form games with perfect and imperfect information.

THEORY ANSWER

1. Subgame Perfect Nash Equilibrium (SPNE):

Subgame Perfect Nash Equilibrium is a refinement of Nash Equilibrium used in extensive form games with
perfect information.
• A strategy profile is SPNE if it forms a Nash Equilibrium in every subgame
• It is used to analyze sequential decision-making
• It is obtained using backward induction
This ensures that players make optimal decisions at every stage of the game
2. Bayesian Nash Equilibrium (BNE):
Bayesian Nash Equilibrium is used in games with imperfect or incomplete information, where players do not
have full knowledge about other players.
• Each player has beliefs (probabilities) about unknown information
• Players choose strategies that maximize expected payoff
• Equilibrium occurs when strategies are optimal given beliefs
3. Comparison of SPNE and BNE:
Feature SPNE BNE
Type of Game Perfect InformationImperfect Information
Nature Sequential Strategic under uncertainty
Method Backward Induction Expected payoff maximization
Focus Optimal decisions at every stage Decision-making with beliefs

4. Role in Game Analysis:

• SPNE helps analyze games where players move one after another and all information is known
• BNE helps analyze situations where players must act without full information

SPNE and BNE are important solution concepts used to analyze different types of games.
• SPNE applies to perfect information and sequential games
• BNE applies to imperfect information and uncertainty-based games
Both extend Nash Equilibrium to provide more accurate predictions of player behavior.
• Pure Nash Equilibrium: No pure strategy Nash equilibrium exists

NUMERICAL QUESTION

Consider the following normal form game:

Player B: Left (L) Player B: Right (R)
Player A: Up (U) (3, 2) (1, 4)
Player A: Down (D) (2, 3) (4, 1)
(a) Find if any player has a dominant strategy.
(b) Find all pure strategy Nash equilibria.

NUMERICAL ANSWER

[Link]

Submission 13: Group 13

THEORY QUESTION

(a) Define a first-price sealed-bid auction [Link] vi denotes the value that player
i attaches to the object, and v1>v2>v3>⋯>[Link] that in a Nash equilibrium of a first-price sealed-bid
auction, The two highest bids are at least 𝑣2 and at most 𝑣1. Also, show that any action profile satisfying
these conditions is a Nash equilibrium.

THEORY ANSWER

Definition: First-price sealed-bid auction:

In a first-price sealed-bid auction, each player submits a bid for the object without knowing the bids of other
players. The highest bidder wins the object and pays the amount equal to their bid.
Given:

Let 𝑣𝑖 be the valuation of player 𝑖 such that ,

v1>v2>v3>⋯>vn.
Part 1: Show that in Nash Equilibrium, the two highest bids satisfy,
v2 ≤ highest bids <= v1

(i) Upper Bound: Highest bid ≤ 𝑣1

In a first-price auction, the winner pays their own [Link] player 1 bids more than 𝑣1 then their payoff
becomes negative: Payoff=v1−bid<0

Hence, no rational player will bid more than their valuation.

Therefore, Highest bid ≤ 𝑣1

(ii) Lower Bound: Highest bid ≥ 𝑣2

Assume that the highest bid is less than 𝑣2. Then player 2 can slightly increase their bid above the current
highest bid and win the auction.
Since the bid is still less than 𝑣2, player 2 gets a positive payoff:

Payoff = 𝑣2 − bid >0

This means player 2 has an incentive to deviate, so the situation cannot be a Nash equilibrium.

Hence, Highest bid≥v2

Conclusion (Part 1):

Combining both results: 𝑣2 ≤ two highest bids ≤ 𝑣1:

Part 2: Any such action profile is a Nash Equilibrium:

Consider any bidding profile where the two highest bids lie between 𝑣2 and 𝑣1.

Player 1: Cannot gain by increasing bid (would reduce payoff) or decreasing bid (may lose).
Player 2: Cannot profitably outbid since they would have to bid at least 𝑣2, giving zero or negative payoff.
Other players: Their valuations are lower than 𝑣2, so bidding higher would give negative payoff.
Thus, no player can improve their payoff by unilaterally deviating.

NUMERICAL QUESTION

In a strategic game with perfect information, two people have 12 pieces of muffins to divide between
themselves. They use the following procedure. Each person names an integer between 0 and 12 and if the
sum total of the two announced numbers is at most 12 then each get muffins equal to the number they had
called. If the sum of the numbers that the people announced exceeds 12 and the numbers are different then
the person who named the smaller number receives the number of muffins equal to the announced number
and the other person gets the remaining muffins. If the sum of the numbers that the people called exceeds 12
and the numbers are the same, then each person receives 6 muffins each. Determine the best response of each
player to each of the other player's actions. find the Nash equilibria of the game.

NUMERICAL ANSWER

[Link]

Submission 14: Nashty

THEORY QUESTION

Discuss the concept of rationality and common knowledge in game theory. Explain how these assumptions
affect players’ strategic decisions.

THEORY ANSWER

In game theory, the concepts of rationality and common knowledge are foundational assumptions used to
predict how players behave in strategic situations. These concepts help explain why players choose certain
strategies and how outcomes such as equilibria are reached.
Concept of Rationality
Rationality refers to the idea that each player:
• Has well-defined preferences over outcomes.
• Aims to maximize their own payoff (utility).
• Chooses strategies that are optimal given their beliefs about other players.
In simple terms, a rational player will always select the strategy that provides the highest expected benefit.
Key features of rationality:
• Consistency in decision-making
• Utility maximization
• Ability to evaluate consequences of actions
Concept of Common Knowledge
Common knowledge goes beyond individual rationality. A fact is common knowledge if:
1. All players know it.
2. All players know that all players know it.
3. This mutual awareness continues infinitely.
In game theory, it is often assumed that:
• All players are rational.
• This rationality is common knowledge among all players.
Relationship Between Rationality and Common Knowledge
Rationality alone is not sufficient for predicting outcomes in strategic games. Players must also believe that
others are rational, and that others believe the same about them. This infinite chain of beliefs (common
knowledge) allows players to anticipate each other’s actions.
Impact on Strategic Decisions
(a) Prediction of Opponent Behaviour
• If a player knows that opponents are rational, they can eliminate strategies that are clearly suboptimal
(dominated strategies).
• This leads to more accurate predictions of others’ moves.
(b) Iterated Elimination of Dominated Strategies
• With common knowledge of rationality, players repeatedly eliminate inferior strategies.
• This process narrows down the possible outcomes of the game.
(c) Nash Equilibrium Formation
• Rational players, with common knowledge, tend to choose strategies that form a Nash equilibrium—
where no player has an incentive to deviate unilaterally.
• Without these assumptions, equilibrium may not be reached or predicted correctly.
(d) Strategic Thinking and Anticipation
• Players think multiple steps ahead (“I know that you know that I know…”).
• This deep reasoning influences choices in games like the prisoner’s dilemma, auctions, and bargaining
situations.
(e) Coordination and Cooperation
• In coordination games, common knowledge helps players align expectations and choose mutually
beneficial outcomes.
• Without it, coordination failures may occur.
Limitations of These Assumptions
• In real life, players may not be perfectly rational.
• Information may not be common knowledge.
• Cognitive limitations and uncertainty can affect decisions.
Conclusion
Rationality and common knowledge are central assumptions in game theory that significantly shape strategic
decision-making. Rationality ensures players aim to maximize their payoffs, while common knowledge
allows them to anticipate others’ actions. Together, these concepts enable the prediction of stable outcomes
such as Nash equilibrium, although their strict assumptions may not always hold in real-world scenarios.

NUMERICAL QUESTION

Construct a custom 3 X 3 payoff matrix representing a market entry game between two competing firms
(Firm A and Firm B). Design the payoffs such that Firm A has a strictly dominant strategy, but Firm B does
not. Using the method of iterated strict dominance, mathematically prove step-by-step how your game
resolves to a unique pure strategy Nash equilibrium.

NUMERICAL ANSWER

[Link]
 Submission 15: JohnNash

THEORY QUESTION

Two firms simultaneously decide whether to advertise or not. Neither knows what the other will do, yet one
strategy seems better regardless. What kind of strategic reasoning is at work here, and what are its
limitations?

THEORY ANSWER

This situation involves a strictly dominant strategy, which is a type of strategic reasoning. In game theory, a
dominant strategy is when a player can get the most benefit from one specific action, no matter what their
opponent does. This strategy mathematically guarantees a better outcome no matter what the other company
does, so there is no need to guess, predict, or analyze what the other company will do. The process of making
decisions becomes completely self-contained because a purely rational actor will always choose the option
that gives them the most utility, no matter what else is going on.

This reasoning means that both of these companies will have to run their advertising campaigns. Firm A will
figure out that they will always make more money if they advertise than if they don't, no matter what Firm B
does. Since Firm B is dealing with the same maths, both companies will logically come to the same
conclusion on their own. So, both companies will use their best strategy, which will lead to a predictable
Nash Equilibrium where both companies are actively advertising. This is because neither can afford to take
the chance of being the only one left behind.

However, using dominant strategy reasoning has big problems, the biggest of which is that it often leads
players to a bad outcome that is similar to the classic Prisoner's Dilemma. When companies follow their own
dominant strategies, they often spend a lot of money on advertising that cancels out the effects of the other
companies' ads. This leads to lower net profits than if the companies had all agreed not to advertise at all.
Additionally, this logic presupposes a simplified "one-shot" game with entirely rational participants. It
doesn't take into account how complicated real-world business can be, like when companies work together to
avoid ad wars or how people act in ways that don't make sense.

NUMERICAL QUESTION

In a highly competitive tech market, two cybersecurity firms, TechSecure and CyberGuard, are preparing to
launch their new network defense software. Each firm must decide whether to adopt an "Aggressive"
marketing strategy (heavy advertising, undercutting prices) or a "Passive" marketing strategy (standard
advertising, maintaining premium prices).

a. Construct and explain a 2x2 payoff matrix for the given scenario, assuming your own payoff values.
b. Identify if any strictly dominant strategy exists for either firm. Justify your reasoning.

c. Find all the Pure-Strategy Nash equilibria for this game.

d. Calculate the Mixed-Strategy Nash equilibrium, showing your step-by-step mathematical working and
reasoning.

NUMERICAL ANSWER

[Link]

Submission 16: Game On

THEORY QUESTION

Why Don't India and Pakistan Just… Strike First?

India and Pakistan have coexisted in a state of nuclear deterrence since 1998 — two neighbours, both with
nukes, both with grievances, and somehow neither has pressed the button. A one-shot game would suggest
"Strike First" dominates. Yet here we are. Why?

Consider a Strike or Deter game between India and Pakistan with the following Game Table :
India(D/S)-Pakistan(D/S) - [3,3][0,5][5,0][-100,-100]

(a) Identify the Nash equilibrium of this one-shot game. Is it Pareto efficient? Explain why the one-shot
outcome is alarming.
(b) "Burning the bridge" in game theory refers to making a threat credible by eliminating your own options.
How does a country's publicly announced "No First Use" (NFU) nuclear policy relate to this concept? Is an
NFU declaration strategically rational or irrational? Argue both sides.
(c) How does imperfect information (e.g., mistaking a missile test for an attack, as nearly happened in 1983
with the Soviet Petrov incident) affect the sustainability of deterrence as an SPE(Subgame Perfect
Equilibrium)? What modification to the trigger strategy might help?

THEORY ANSWER

(a) Nash equilibrium & Pareto efficiency

Two Nash equilibria: (Strike, Deter) and (Deter, Strike) since the striker gets 5 and neither player wants to
deviate unilaterally.
(Deter, Deter)(Deter,Strike)(Strike,Deter)are all pareto efficient as neither gives a better option of increasing
self payoff without reducing other player's payoff.
Each country has an incentive to pre-emptively strike if it expects the other to deter, creating instability and
risk of catastrophe.
b) strategically rational : NFU commits a country to retaliation only, increasing credibility of second-strike
deterrence and reducing escalation risk.
Potentially irrational : it removes the option of a pre-emptive strike, which might weaken bargaining power if
the opponent doubts retaliation.
c)False alarms or misinterpreted signals could trigger retaliation under grim trigger, destroying cooperation
even without a real attack.A forgiving trigger strategy (e.g., tit-for-tat with verification or delayed
punishment) helps maintain deterrence despite accidental signals.

NUMERICAL QUESTION

The Nuclear Inspector's Dilemma — US vs Iran

The IAEA suspects Iran might be secretly enriching uranium. The US must decide whether to conduct an
expensive inspection. Iran, meanwhile, decides whether to enrich covertly or comply with the treaty. Neither
player knows the other's move in advance. Sound familiar? It should — it's on the news every other week.

Game Table : US(Inspect/Dont Inspect)- Iran(Enrich/Comply)- [10,8][-3,2][-10,10][0,0]

(a) Show that no pure strategy Nash equilibrium exists in this game.
(b) Let p = probability Iran plays Enrich, and q = probability US plays Inspect. Find the mixed strategy Nash
equilibrium (MSNE).
(c) Interpret the equilibrium probabilities in plain English. What does it say about how often the US should
realistically inspect?
(d) If Iran's cost of being caught increases (the −8 becomes −14), how does the MSNE change? Who
benefits?

NUMERICAL ANSWER

[Link]

Submission 17: Vertex

THEORY QUESTION

"In an extensive form game, a threat is only strategically meaningful if it is credible — that is, if the
threatening player would actually carry it out when the time comes."

With reference to the concept of Subgame Perfect Equilibrium (SPE) and Backward Induction:
(a) Formally explain why Nash Equilibrium alone is insufficient to analyse dynamic games, and how SPE
refines it by eliminating non-credible threats. Illustrate with the Ultimatum Game.
(b) Analyse the "Burning the Bridge" scenario as a strategic commitment device. How does a player
voluntarily restricting their own future choices strengthen their bargaining position? Connect this to the
concept of first-mover advantage in the Stackelberg Model.

THEORY ANSWER

Answer (a)

Limitation of Nash Equilibrium in Dynamic Games:

In a static game, Nash Equilibrium requires each player's strategy to be a best response to others. However,
in extensive form (sequential) games, NE allows for equilibria sustained by non-credible threats — threats
that a rational player would never actually carry out when the decision node is reached.
A strategy in an extensive form game must specify actions at every information set, including those off the
equilibrium path. NE places no rationality requirement on these off-path nodes, which allows implausible
equilibria to survive.
Subgame Perfect Equilibrium (SPE) refines NE by requiring that strategies constitute a Nash Equilibrium in
every subgame, not just the overall game. This is operationalised through backward induction — solving
from the final nodes backward to the root, ensuring no player relies on a threat they wouldn't rationally
execute.
Illustration — The Ultimatum Game:

Player 1 proposes a split (x,1−x)(x, 1-x)

(x,1−x) of ₹100.

Player 2 either accepts or rejects (both get 0 if rejected).

A NE can support Player 2 threatening to reject any offer below ₹50, which keeps Player 1 from offering
less. However, rejection is never rational for Player 2 once reached — any positive amount is better than
zero. Backward induction eliminates this threat: Player 2 accepts any x>0x > 0
x>0, so Player 1 offers the minimum positive amount. This is the unique
SPE.

Answer (b)

"Burning the bridge" refers to a situation where a player voluntarily eliminates their own retreat option,
making it common knowledge that they cannot back down. Paradoxically, having fewer choices strengthens
your strategic position — because it removes the opponent's incentive to hold out, expecting you to
eventually concede.
Formally, by eliminating the option to retreat, the player converts what was previously a non-credible threat
("I will fight") into a credible commitment — because the backward induction now leaves them with only
one feasible action.
Connection to Stackelberg First-Mover Advantage:
In the Stackelberg Duopoly, Firm 1 (the leader) publicly and irreversibly commits to a large output quantity
q1∗q_1^*
q1∗ before Firm 2 decides. This commitment is the bridge being burned.

Firm 2, observing q1∗q_1^*

q1∗, best-responds with a lower q2∗q_2^*
q2∗.

Firm 1 anticipates this and strategically overproduces relative to Cournot levels.

The key is irreversibility — if Firm 1 could revise its output after Firm 2 responds, the commitment would
be non-credible and the Stackelberg outcome would unravel back toward Cournot. The first-mover
advantage therefore depends entirely on the commitment being binding, which is precisely the same logic as
burning the bridge.

NUMERICAL QUESTION

A government auctions a single radio spectrum licence to two risk-neutral bidders. Each bidder's private
valuation viv_i
vi is drawn
independently and uniformly from [0, 1].
Assume a symmetric BNE where each bidder uses the strategy b(v)=αvb(v) = \alpha v
b(v)=αv.

(a) Set up Bidder 1's expected profit as a function of their bid b1b_1
b1, given Bidder 2 plays b2=αv2b_2 = \alpha v_2
b2=αv2.
(1 mark)
(b) Maximise expected profit to derive α. Hence write the equilibrium bidding strategy explicitly. (2 marks)
(c) Compute the seller's expected revenue under this First-Price auction. Compare it with the expected
revenue under a Second-Price auction (state without re-deriving). What theorem does this confirm? (2
marks)

NUMERICAL ANSWER

[Link]

Submission 18: Team Strong

THEORY QUESTION

Subgame Perfect Equilibrium and Backward Induction

Answer the following:

(a) Define a Subgame Perfect Equilibrium (SPE). How does it refine the concept of Nash Equilibrium in
extensive-form games?
(b) Explain the method of Backward Induction and describe how it is used to find the SPE.
(c) Apply backward induction to the Ultimatum Game: Player 1 proposes a split of ₹100 between themselves
and Player 2. Player 2 either accepts (both get the proposed amounts) or rejects (both get ₹0). Find the
Subgame Perfect Equilibrium.
(d) Why does the SPE prediction of the Ultimatum Game often fail in real-life experiments? What does this
tell us about the assumptions of game theory?

THEORY ANSWER

(a) Subgame Perfect Equilibrium

A Subgame Perfect Equilibrium is a strategy profile in an extensive-form game such
that the strategies form a Nash Equilibrium in every subgame of the original game, not
just the game as a whole.
A subgame is any portion of the game tree that
(i) starts at a single decision node,
(ii)includes all successors of that node, and
(iii) does not split any information set.
SPE refines Nash Equilibrium by ruling out non-credible threats – threats a rational
player would never actually carry out if the moment arose. By requiring rational behaviour
at every node (on and off the equilibrium path), SPE eliminates Nash equilibria that rely
on such threats

(b) Backward Induction

Backward Induction is the standard procedure for computing the SPE in finite, perfect information games:
1. Start at the terminal nodes and record their payoffs.
2. Move one step up the tree; at each node find the action that maximises the current
player’s payoff.
3. Replace that node with the resulting payoff vector.
4. Repeat steps 2–3 moving toward the root.
5. The sequence of optimal actions found constitutes the SPE.

(c) SPE of the Ultimatum Game

Setup: Player 1 offers a split (x, 100 − x). Player 2 observes the offer and accepts or
rejects it.
Step 1 – Player 2’s decision (bottom of the tree):
Accepting any offer ε > 0 yields ε; rejecting yields 0. A rational Player 2 accepts any
positive offer.
Step 2 – Player 1’s decision (root):
Knowing Player 2 accepts anything positive, Player 1 offers the smallest positive amount
to Player 2 and keeps the rest. In the continuous case this approaches ε → 0.
SPE: Player 1 offers (100 − ε, ε) ≈ (Rs. 100, Rs. 0). Player 2 accepts.

(d) Why SPE Fails in Practice – Behavioural Critique

Experiments consistently find that:
• Player 1 typically offers 40–50% of the stake, not near zero.
• Player 2 often rejects offers below 20–30%, accepting a personal loss to punish unfair
behaviour.
This divergence exposes the limits of classical rationality:
• Inequity Aversion: Players care about fairness, not just their own payoff (Fehr
& Schmidt, 1999).
• Emotions and Norms: Anger and social norms override narrow self-interest.
• Bounded Rationality: Players do not always perform full backward induction.
SPE is a powerful theoretical benchmark, but real behaviour incorporates social preferences and emotions.
This gap motivates Behavioural Game Theory, which extends
classical game theory to model these factors.

NUMERICAL QUESTION

Cournot Duopoly – Nash

Equilibrium
Two firms, Firm 1 and Firm 2, compete in a market selling a homogeneous good. The
inverse market demand function is:
P = 100 − Q, Q = q1 + q2
Each firm has a constant marginal cost c = 10 and zero fixed costs. Both firms choose
their quantities simultaneously and independently.
(a) Derive the best-response function for each firm.
(b) Find the Cournot–Nash Equilibrium quantities q1*
and q2*.
(c) Compute the equilibrium price P* and each firm’s profit.
(d) Compare the Cournot outcome with the monopoly and perfectly competitive outcomes.

NUMERICAL ANSWER

[Link]
 Submission 19: The Equilibrium Crew

THEORY QUESTION

Explain the concept of Bayesian Nash Equilibrium in games with incomplete information. Using the
example of a first-price sealed-bid auction, describe how equilibrium bidding strategies are determined for
risk-neutral bidders. Also explain how the equilibrium changes when bidders are risk-averse.

THEORY ANSWER

In many real-world strategic situations, players do not have complete information about others, such as their
preferences, costs, or valuations. Bayesian games provide a framework to analyze such situations by
introducing the concept of types, where each player has some private information and forms beliefs about the
types of others. These beliefs are typically represented using probability distributions. The appropriate
solution concept in such settings is the Bayesian Nash Equilibrium, in which each player chooses a strategy
that maximizes their expected payoff given their beliefs about other players’ types and strategies.

A classic application of Bayesian games is the first-price sealed-bid auction. In this auction format, each
bidder submits a bid without knowing the bids of others, and the highest bidder wins while paying their own
bid. Since each bidder has a private valuation of the item, the situation naturally involves incomplete
information. To determine the equilibrium strategy, we assume that bidders follow a symmetric strategy
where the bid is a function of their valuation. For risk-neutral bidders with valuations uniformly distributed
between 0 and 1, the equilibrium strategy involves “bid shading,” meaning that bidders submit bids lower
than their true valuations. This occurs because bidding too high reduces profit, while bidding too low reduces
the probability of winning, so bidders balance these two effects.

In equilibrium, for 𝑛 bidders, the optimal bid is given by a fraction of the valuation, specifically
𝑏(𝑣)=(𝑛−1/𝑛)*𝑣.
This result shows that as competition increases, bidders bid more aggressively. The reasoning behind this
equilibrium comes from maximizing expected payoff, which is the product of the probability of winning and
the surplus obtained if the bidder wins.

When bidders are risk-averse, their behavior changes significantly. Unlike risk-neutral bidders who focus on
maximizing expected payoff, risk-averse bidders prefer a higher probability of winning even if it means a
lower payoff. As a result, they bid more aggressively, meaning their bids move closer to their true valuations.
This reduces bid shading and increases the seller’s expected revenue. Thus, risk preferences play a crucial
role in shaping equilibrium outcomes in auctions.

Overall, Bayesian Nash Equilibrium provides a powerful way to analyze strategic interactions under
uncertainty, and auction models illustrate how private information and beliefs influence decision-making in
competitive environments.
NUMERICAL QUESTION

Consider a Cournot duopoly where two firms simultaneously choose quantities

𝑞1 and 𝑞2. The inverse demand function is:

𝑃=120−(𝑞1+𝑞2)

Each firm has a cost function:

𝐶(𝑞𝑖)=20𝑞𝑖

Find the profit functions, derive the best response functions, compute the Cournot-Nash equilibrium, and
determine the equilibrium price and profits. Also explain how the result reflects strategic interdependence
between firms.

NUMERICAL ANSWER

[Link]

Submission 20: XQC

THEORY QUESTION

An incumbent firm (I) faces a potential entrant (E). The game proceeds as follows:
• Stage 1: E decides: Enter (En) or Stay Out (SO)
• Stage 2: If E enters, I decides: Accommodate (A) or Fight (F) (wage a price war)
Payoffs (E, I):
E's Move I's Response Payoff (E, I)
Enter Fight (−2, −1)
Enter Accommodate (4, 2)
Stay Out — (0, 6)

(a) Draw the extensive-form game tree. Convert to normal form and identify all Nash
Equilibria of the game. For each NE, state whether or not it is subgame perfect and why.
(b) Derive the Subgame Perfect Equilibrium (SPE) via backward induction. Explain precisely
why I's threat to "Fight if entered" is non-credible, and why it cannot be sustained in a SPE
even though it supports a Nash Equilibrium.
(c) Now introduce a Stage 0: Before E decides, I can choose to make an irreversible
investment in excess capacity at cost K = 2. This investment changes I's post-entry payoffs as
follows:
• Fight becomes profitable: I's Fight payoff changes from −1 → +1
• Accommodate becomes wasteful: I's Accommodate payoff changes from 2 → 0
• Monopoly payoff (if E stays out) becomes 6 − 2 = 4 (net of investment)
Draw the new three-stage extensive-form game tree. Find the new SPE using backward
induction. Does I choose to invest? Carefully compare I's equilibrium payoff with and
without the investment.
(d) Using your analysis, explain the "Burning the Bridge" principle: How does an agent
improve their strategic outcome by voluntarily destroying one of their own options? Why
does this work specifically because of the sequential and observable nature of the
commitment? Support your answer with one non-military real-world analogy from
economics or business.

THEORY ANSWER

Part (a): Game Tree and Nash Equilibria

The extensive-form tree has E at the root, branching into En and SO. The En branch leads to
I's decision node (A or F). Terminal payoffs: (4,2) for (En,A), (−2,−1) for (En,F), and (0,6) for
SO.
Normal Form — I has strategies {A, F} (contingent on entry):

Nash Equilibria:
NE 1 — (Enter, A): If I plays A, E prefers Enter (4 > 0) . If E plays Enter, I prefers A (2 > −1). →
Valid NE.
NE 2 — (Stay Out, F): If I plays F, E prefers Stay Out (0 > −2) . If E plays Stay Out, I is
indifferent between A and F (both yield 6, since no entry occurs). → Valid NE.
NE 1 is subgame perfect (I's strategy is rational at the post-entry subgame). NE 2 is not
subgame perfect — it is sustained by a non-credible threat (see Part b).

Part (b): SPE and Non-Credible Threats

Backward induction:
Step 1 — I's subgame (post-entry): I chooses between A (payoff = 2) and F (payoff = −1).
Rational choice: Accommodate (A).
Step 2 — E's decision: E anticipates I will Accommodate → E compares Enter (payoff = 4) vs.
Stay Out (payoff = 0). Rational choice: Enter.
SPE: (Enter, Accommodate) with payoffs (E = 4, I = 2).
Why Fight is non-credible: In NE 2, I threatens to Fight if E enters. But once entry has
occurred, fighting costs I a payoff of −1 versus +2 from accommodating. No rational agent
inflicts damage on themselves to punish an event that has already happened. The threat is
not sequentially rational — it fails at the post-entry subgame. SPE eliminates precisely such
"empty threats" by requiring rational play at every subgame, not just on the equilibrium
path.

I: Accommodate (A) I: Fight (F)

E: Enter 4, 2 −2, −1
E: Stay Out 0, 6 0, 6

Part (c): The Commitment Game (Three-Stage Tree)

New payoff structure after investment:
E's Move I's Response Payoff (E, I)
Enter Fight (−2, +1)
Enter Accommodate (4, 0)
Stay Out — (0, 4)
New extensive-form tree: I moves first at Stage 0 (Invest / Don't Invest). Each branch leads
to E's decision node, which (if Enter) leads to I's response node.
Backward induction on the "Invest" branch:
Step 1 — I's post-entry decision (having invested): Fight gives I = 1, Accommodate gives I = 0.
→ I chooses Fight.
Step 2 — E's entry decision (knowing I will Fight): Enter gives E = −2, Stay Out gives E = 0. → E
chooses Stay Out.
Step 3 — I's Stage 0 decision: If I invests: outcome is (Stay Out, Fight) → I gets 4. If I does not
invest: SPE from Part (b) applies → I gets 2. Since 4 > 2, I chooses to invest.
New SPE: I invests, E stays out, and I would fight if entry occurred (though it doesn't).
Scenario I's Equilibrium Payoff
No investment (Part b) 2
With investment (Part c) 4
By spending K = 2, I converts a non-credible threat into a credible one, deters entry entirely,
and ends up better off by +2.

Part (d): The "Burning the Bridge" Principle

The paradox: I improves its outcome by destroying its own option to accommodate. Here is
why this works:
1. Commitment changes beliefs: The investment is observable and irreversible. When E
sees I has invested, E rationally updates its belief — it now knows I will fight. The
threat is no longer empty; it is incentive-compatible.
2. Restriction as power: Without commitment, I's best response to entry is
Accommodate (2 > −1), so the threat to Fight is disbelieved. With commitment, I's
best response to entry is Fight (1 > 0). E enters only if it expects accommodation —
so by making accommodation irrational, I eliminates the rational basis for entry.

3. Sequential observability is essential: This only works because the commitment is

made before E moves and is publicly observable. A secret commitment (or a
reversible one) would not shift E's beliefs and would therefore have no deterrence
value.
Real-world analogy — Central Bank Independence: A government that retains the power to
print money cannot credibly promise low inflation — firms and workers expect it to inflate
away debt in a crisis. To solve this, governments "burn the bridge" by granting independence
to a central bank with a strict inflation mandate (e.g., the Reserve Bank of India's MPC
framework). The government voluntarily gives up its own monetary policy tool. This
restriction is precisely what makes the low-inflation commitment credible — wage-setters
and investors believe it — and paradoxically leads to better macroeconomic outcomes (lower
inflation expectations, lower borrowing costs) than if the government retained full control.
(THIS IS JUST AN EXAMPLE OF AN ANALOGY)

NUMERICAL QUESTION

Two firms compete in a market with inverse demand P = 120 − Q, where Q = q1 + q2.
• Firm 1's marginal cost is MC1 = 30 — this is common knowledge.
• Firm 2 has private information about its own cost:
o With probability 1⁄2, Firm 2 is type L (low cost): MC2 = 20
o With probability 1⁄2, Firm 2 is type H (high cost): MC2 = 40

Firm 2 knows its own type. Firm 1 only knows the probability distribution. Both firms choose
quantities simultaneously.
(a) Define a Bayesian Nash Equilibrium (BNE) formally in the context of this game. What is
the key difference from a standard Nash Equilibrium?
(b) Derive Firm 2's best response functions for each type separately.
(c) Derive Firm 1's best response function using expected profit maximization. Solve the
system to find the BNE quantities (q1*, q2L*, q2H*).
(d) Calculate each firm's equilibrium profits for each realized type, and Firm 1's expected
profit. Then find the Cournot NE quantities and Firm 1's expected profit under complete
information (i.e., if Firm 1 knew Firm 2's exact type). Compute the value of information to
Firm 1 and comment on the result.

NUMERICAL ANSWER

[Link]

Submission 21: AKD

THEORY QUESTION

Theory question -Define the concept of "iterated strict dominance." Then, provide a theoretical argument
explaining whether the process of iteratively eliminating strictly dominated strategies can ever result in the
elimination of a pure strategy Nash [Link] by contradiction.

THEORY ANSWER
Iterated Strict Dominance is the process where:

Players repeatedly remove such strictly dominated strategies (i.e., strategies that are always worse than
another).
- eAfter each round of elimination, the game becomes smaller.

Players then re-evaluate the reduced game to eliminate more dominated strategies.

This process continues until no further eliminations are possible.

Pure Strategy Nash Equilibrium

A pure strategy Nash equilibrium is a set of strategies (one for each player) where no player can increase
their own payoff by unilaterally changing their action, while others keep theirs constant.

A Pure Strategy Nash Equilibrium is a stable situation in a game where:

The process of iterated elimination of strictly dominated strategies never eliminates any pure strategy Nash
Equilibrium of the original game.

Proof (By Contradiction)

Step 1: Assumption
Assume, for contradiction, that iterated strict dominance eliminates a Pure Strategy Nash Equilibrium.

Let the equilibrium be:

s* = (s1*, s2*, ..., sn*)

Step 2: First Elimination

Since elimination happens in rounds, there must be a first round where part of the equilibrium is removed.

Let:

Round k be the first round where a strategy si* (belonging to player i) is eliminated.
Step 3: Reason for Elimination

If si* is eliminated in Round k, then it must be strictly dominated by another strategy si′ that is still available.

This means:

ui(si′, s−i) > ui(si*, s−i)

for all s−i belonging to S−i^k

Where:

S−i^k = strategies of other players that survive till round k

Step 4: Deriving the Contradiction

Since Round k is the first elimination affecting equilibrium:

All other equilibrium strategies s−i* must still exist in Round k

Thus:

s−i* belongs to S−i^k

Substituting into the inequality:

ui(si′, s−i*) > ui(si*, s−i*

Step 5: Contradiction with Nash Equilibrium

The above inequality implies:

Player i can get a strictly higher payoff by switching from si* to si′
But this contradicts the definition of Nash Equilibrium, which states:

ui(si*, s−i*) ≥ ui(si, s−i*)

for any alternative strategy si

Conclusion

Since our assumption leads to a contradiction, it must be false.

Therefore:

Iterated elimination of strictly dominated strategies never removes any Pure Strategy Nash Equilibrium.

NUMERICAL QUESTION

Case Study: Packaged Drinking Water Market

A local market has two firms selling the same product, such as packaged drinking water. The market demand
is given by the inverse demand function:

P = 100 - (q1 + q2)

where:

q1 = output of Firm 1 (Leader)

q2 = output of Firm 2 (Follower)

Firm 1 is the market leader and has a constant marginal cost:

c1 = 20
Firm 2 is the follower, but its cost is uncertain:

With probability p = 0.4, marginal cost is low:

cL = 10

-With probability 1 - p = 0.6, marginal cost is high:

cH = 30

Firm 2 knows its own cost before choosing output, but Firm 1 o

QUESTIONS

1. Find Firm 2’s best response in each cost state.

2. Find the Stackelberg–Bayesian equilibrium outputs of both firms.

3. Compute profits of both firms in both cost states.

4. Explain:

Whether the leader always has an advantage

How Firm 1’s output changes if the probability of low cost increases

NUMERICAL ANSWER

[Link]
 Submission 22: Best Response

THEORY QUESTION

Q. "A dominant strategy equilibrium is always a Nash Equilibrium, but a Nash Equilibrium need not involve
dominant strategies." Explain this statement with a self-constructed 2×2 payoff matrix that clearly illustrates
both concepts. Also discuss: what happens to the equilibrium prediction if one player has a dominant strategy
but the other does not?

THEORY ANSWER

Sol: Part 1: Dominant strategy equilibrium ⟹ Nash Equilibrium

A dominant strategy is one that gives a player the highest payoff regardless of what the opponent does. If
both players have dominant strategies, they will play them — and by definition, neither wants to deviate
unilaterally. That's exactly what a Nash Equilibrium is. So the implication holds in one direction
automatically.
Part 2: Nash Equilibrium ⟹ dominant strategies (does not hold)
Consider this 2×2 matrix (Row player / Column player):

L
R
U
3, 3
0, 2
D
2, 0
1, 1

Checking for dominant strategies:

Row: U gives (3, 0), D gives (2, 1) → U is better if Column plays L, but D is better if Column plays R. No
dominant strategy.
Column: Same logic by symmetry. No dominant strategy.
Checking for Nash Equilibrium:
(U, L): Row gets 3 — deviating to D gives 2. No. Column gets 3 — deviating to R gives 2. No. ✅ Nash
Equilibrium.
(D, R): Row gets 1 — deviating to U gives 0. No. Column gets 1 — deviating to L gives 0. No. ✅ Nash
Equilibrium.
So we have two Nash Equilibria with zero dominant strategies — proving the second part.
Part 3: One player has a dominant strategy, the other doesn't
L
R
U
4, 3
4, 1
D
2, 2
3, 3

Row: U gives (4, 4), D gives (2, 3) → U dominates D for Row. Dominant strategy: U.
Column: If Row plays U → L gives 3, R gives 1 → prefer L. If Row plays D → L gives 2, R gives 3 →
prefer R. No dominant strategy for Column.
Since Row's dominant strategy (U) is common knowledge, Column can rationally anticipate Row will play U
and best-respond accordingly — choosing L. The equilibrium resolves to (U, L) through iterated elimination
of dominated strategies, even without Column having a dominant strategy of their own. This is exactly what
iterated strict dominance does — it uses one player's dominance to simplify the other player's decision.

NUMERICAL QUESTION

Q. Two firms (A and B) simultaneously choose output levels in a Cournot duopoly. The inverse demand is:
P=100−Q, where Q=qA+qB
Both firms have a constant marginal cost of MC = 10 and zero fixed costs.
(a) Derive each firm's best response function.
(b) Find the Cournot-Nash equilibrium outputs, market price, and each firm's profit.
(c) If both firms colluded to maximize joint profit, what would total output and price be? Why is this
outcome unstable as a Nash Equilibrium?

NUMERICAL ANSWER

[Link]

Submission 23: Game-engers

THEORY QUESTION

Using one real world application explain how-

1. Rationality and common knowledge influence strategic outcomes
2. The failure of these assumptions can change equilibrium predictions

THEORY ANSWER

Application: Oligopoly Pricing in the Airline Industry

- A useful real-world application to understand the role of rationality and common knowledge is pricing
behavior in the airline industry, involving firms such as Delta Air Lines, American Airlines, and United
Airlines. These firms operate in an oligopolistic market where each airline’s pricing decision directly affects
the profits of others. This interdependence makes the situation ideal for analysis through game theory.

1. Influence of Rationality and Common Knowledge on Strategic Outcomes

- In game theory, rationality implies that each player chooses strategies that maximize their own payoff,
assuming consistency and logical decision-making. Common knowledge of rationality extends this idea
further: not only is each player rational, but every player knows that all others are rational, and everyone
knows that everyone knows this, creating an infinite chain of mutual beliefs. These assumptions are critical
in determining equilibrium outcomes.
- In the context of airline pricing, firms face a strategic dilemma similar to the Prisoner’s Dilemma. If all
airlines maintain high ticket prices, they collectively enjoy high profits. However, each airline has an
incentive to slightly reduce prices to attract a larger share of customers. Given rationality, each airline
anticipates that competitors will also attempt to undercut prices to maximize their own profits. Because this
reasoning is common knowledge, no airline expects others to maintain high prices out of cooperation or
goodwill.
- As a result, each airline adopts a defensive strategy by lowering its prices preemptively. This leads to a
Nash equilibrium in which all firms charge relatively low prices, even though this outcome is not collectively
optimal. The equilibrium emerges because no airline can unilaterally increase its price without losing market
share. Thus, rationality and common knowledge drive the system toward a stable but suboptimal outcome,
often referred to as a price war. This demonstrates how strong theoretical assumptions can produce precise
and predictable strategic outcomes, even if those outcomes are inefficient from a collective perspective.

2. Impact of Failure of These Assumptions on Equilibrium Predictions

- When the assumptions of rationality and common knowledge are relaxed or fail to hold, the predicted
equilibrium outcomes can change significantly. In real-world markets, firms do not always behave with
perfect rationality. Instead, they may exhibit bounded rationality, relying on heuristics or simplified decision-
making processes. In such cases, airlines may not immediately anticipate competitors’ reactions or may delay
price adjustments. This can result in temporary price stability or gradual changes rather than an immediate
convergence to the Nash equilibrium predicted by standard theory.
- Furthermore, the absence of common knowledge introduces uncertainty into the strategic environment. If
airlines are unsure about competitors’ cost structures, objectives, or level of rationality, they cannot
confidently predict how others will respond to pricing decisions. This uncertainty may lead firms to
experiment with different pricing strategies or maintain higher prices for longer periods. Consequently,
multiple equilibria may arise, or outcomes may fluctuate over time instead of settling into a single
predictable equilibrium.
- Behavioral factors also play a crucial role in altering equilibrium predictions. Firms may be influenced by
considerations such as reputation, fairness, or long-term strategic positioning. For example, an airline might
avoid aggressive price cuts to signal stability or to discourage future price wars. Such behavior deviates from
the purely profit-maximizing assumption and can sustain higher prices than those predicted by standard Nash
equilibrium analysis.
- Additionally, in a repeated interaction setting—where airlines compete over time rather than in a one-shot
game—the breakdown of static assumptions allows for cooperative outcomes to emerge. Firms may
implicitly coordinate by maintaining high prices and punishing deviations through future price cuts. This
type of tacit collusion is not predicted by one-shot models under strict rationality assumptions but becomes
feasible when players consider long-term payoffs and strategic interdependence over time.
- In summary, when rationality and common knowledge do not fully hold, equilibrium predictions become
less rigid and more reflective of real-world complexity. Outcomes may include sustained cooperation, price
rigidity, or fluctuating strategies, highlighting the importance of incorporating more realistic assumptions
into economic models.

Conclusion
- The assumptions of rationality and common knowledge are foundational in game theory, enabling clear and
precise predictions such as the Nash equilibrium in oligopolistic pricing. However, real-world deviations
from these assumptions—such as bounded rationality, incomplete information, behavioral biases, and
repeated interactions—can significantly alter strategic outcomes. Therefore, while these assumptions provide
a useful analytical baseline, relaxing them is essential for understanding and accurately predicting behavior
in complex economic environments.

NUMERICAL QUESTION

Two players compete for a total of 14 tokens. Each player simultaneously announces an integer between 0
and 14 (inclusive).

The allocation of tokens is determined as follows:

If the sum of the announced numbers is less than or equal to 14, then each player receives the number of
tokens they announced.
If the sum of the announced numbers exceeds 14 and the two numbers are different, then:
The player who announced the smaller number receives that number of tokens.
The other player receives the remaining tokens, equal to
14
14 minus the smaller number.
If the sum of the announced numbers exceeds 14 and the two numbers are equal, then each player receives 7
tokens.

Determine the best response of each player for every possible action of the other player. Identify all pure
strategy Nash equilibria of the game. Explain the intuition behind the equilibrium outcomes.

NUMERICAL ANSWER

[Link]
 Submission 24: Achievers

THEORY QUESTION

In a sequential game, backward induction is used to determine equilibrium.

However, in some cases it fails to give clear results.
Explain any three such situations with examples.

THEORY ANSWER

1. Requires Complete Information

Consider a sequential game where Player A moves first and Player B responds.
If A chooses L → B gets (4,2) or (3,3)
If A chooses R → B gets (2,5) or (1,1)
Backward induction assumes B knows all these payoffs.
Limitation:
If B is uncertain (e.g., doesn’t know exact payoffs), then B’s optimal choice is unclear, and A cannot predict
B’s move.
Conclusion:
Backward induction fails when information is incomplete.

2. Not Applicable to Infinite/Repeated Games

Two firms repeatedly decide prices:

Each day: High price or Low price
Game continues indefinitely (no final round)
Backward induction needs a last stage to start reasoning.
Limitation:
Since there is no endpoint, the method cannot be applied.
Conclusion:
Backward induction works only for finite games, not infinite ones.

3. Indeterminacy when Payoffs are Equal

At the final stage, Player B chooses:

Left → (5,5)
Right → (5,5)
Both options give equal payoff, so B is indifferent.
Limitation:
Backward induction cannot determine which action B will take.
Conclusion:
Leads to multiple possible outcomes (no unique solution).

NUMERICAL QUESTION

Question: Stackelberg Duopoly

Two firms produce a homogeneous product. The market inverse demand function is:

𝑃=100−𝑄, where 𝑄=𝑞1+𝑞2

P=100−Q,where Q=q1+q2

Firm 1 is the leader

Firm 2 is the follower

Both firms have constant marginal cost:

𝑀𝐶1=𝑀𝐶2=20MC1=MC2=20

Find the reaction function of Firm 2 (follower)

Determine the optimal output of Firm 1 (leader)
Find equilibrium outputs
q1,𝑞2,q1,q2, market price, and profits

NUMERICAL ANSWER

[Link]

Submission 25: Dominant Strategy

THEORY QUESTION

[Link]
usp=sharing

THEORY ANSWER

[Link]
usp=sharing

NUMERICAL QUESTION
Three firms — A, B, and C — simultaneously choose a pricing strategy: Low (L), Medium (M), or High (H).
However, Firm B is a conditional entrant: it enters the market and chooses High (H) only if the sum of Firm
A's and Firm C's strategy indices is strictly greater than 2, where L = 1, M = 2, H = 3. Otherwise, Firm B
automatically plays Low (L), regardless of its own intention. Firms A and C choose freely.

The payoff to each firm depends on all three strategies actually played (including B's resolved strategy).
Payoff matrices are given.

Answer the following:

(a) Determine B’s strategy and construct the effective payoff matrix
(b) Identify dominated strategies
(c) Apply IESDS
(d) Find Nash equilibrium

NUMERICAL ANSWER

[Link]

Submission 26: abcd

THEORY QUESTION

QUESTION 2
The Indifference Principle in Mixed Strategy Equilibria
Topics: Mixed Strategies | Indifference Condition | Expected Payoff Calculation

Question
Explain the Indifference Principle as it applies to mixed strategy Nash Equilibria. Provide a clear theoretical
account of why a player's randomized strategy must be constructed so that the opposing player gains no
preference for any one of their own available actions. Support your explanation with a worked example.

THEORY ANSWER

ANSWER — QUESTION 2

1. Understanding Mixed Strategies

In many strategic interactions, no single action guarantees the best outcome regardless of what the opponent
does. When a game contains no pure strategy Nash Equilibrium, players may respond by assigning a
probability distribution across their available actions. This approach is referred to as a mixed strategy.
Rather than locking into one choice, a player using a mixed strategy keeps the opponent uncertain about
what action will be taken. The selection in each instance is governed by pre-set probabilities, and the player's
objective is to identify the probability values that constitute an equilibrium.

2. The Indifference Principle — What It States

When a player uses a mixed strategy at equilibrium, the probabilities assigned to their actions must be
chosen such that the opponent receives equal expected payoffs from all of their own strategies. This
requirement is known as the Indifference Principle.

Put differently, the randomizing player must calibrate their mix so precisely that the opponent has no
computational reason to favor one action over another. Every option in the opponent's strategy set must
appear equally attractive when evaluated in terms of expected returns.

3. Why Indifference Is Required for Equilibrium to Hold

The necessity of this condition becomes clear when we reason through what happens if it is violated:

The Logic of Why Indifference Cannot Be Avoided:

6. Suppose one of the opponent's strategies yields a higher expected return: A rational opponent will
immediately recognize this and allocate probability 1 to that superior strategy, abandoning all others.
7. This changes the environment the first player faces: The first player built their mixed strategy around
the assumption that the opponent would continue randomizing. If the opponent stops, the first player's mix
may no longer be optimal.
8. Equilibrium breaks down: What was supposed to be a stable mixed strategy pair unravels because one
side has deviated.
9. The remedy is indifference: The only way to prevent the opponent from abandoning randomization is to
ensure that no action is more rewarding than any other. Indifference removes any incentive to deviate.

An important conceptual point follows: players do not randomize out of confusion or indecision.
Randomization is a deliberate strategic tool used to control the opponent's behavior by eliminating any
exploitable preference.

4. Worked Example — Coin Side Selection Game

Setup
Consider a two-player game in which each participant privately selects one face of a coin either Head (H) or
Tail (T) without knowing the other's choice.

• Player A wins one unit whenever both players select the same face.
• Player B wins one unit whenever the two players select opposite faces.
• The loser in each case surrenders one unit to the winner.
Payoff Matrix
A vs B B selects Head B selects Tail
A selects Head A gains +1, B loses 1 A loses 1, B gains +1
A selects Tail A loses 1, B gains +1 A gains +1, B loses 1

Why No Pure Strategy Equilibrium Exists: If Player A selects Head, Player B's best response is Tail. But if
Player B selects Tail, Player A prefers Tail as well. And if A selects Tail, B prefers Head and so the cycle
continues indefinitely. No stable pure strategy pair exists; each player always has an incentive to switch.

Applying the Indifference Principle

Assign to Player A the probability p for selecting Head and (1 - p) for selecting Tail. Player B will then
calculate the expected gain from each of their available options:

Player B's Expected Payoff Calculations:

When B selects Head:

E(Head) = (-1) multiplied by p, added to (+1) multiplied by (1 - p)
E(Head) = 1 - 2p

When B selects Tail:

E(Tail) = (+1) multiplied by p, added to (-1) multiplied by (1 - p)
E(Tail) = 2p - 1

For Player B to be genuinely indifferent between Head and Tail, both expressions must produce the same
value:

Setting E(Head) equal to E(Tail):

1 - 2p = 2p - 1
Rearranging: 4p = 2
Therefore: p = 0.5

This tells us that Player A must select Head and Tail with equal probability of 0.5. At this value, Player B's
expected payoff from Head equals their expected payoff from Tail neither option appears superior, and
Player B has no grounds for abandoning randomization.

The symmetric structure of the game ensures that an identical computation applies to Player B's mix. When
Player B also randomizes with probability 0.5, Player A becomes indifferent between their own two choices.

The Mixed Strategy Nash Equilibrium of this game is: Player A plays Head with probability 0.5, Player B
plays Head with probability 0.5.
5. Concluding Remarks
Key Takeaways:

• Randomization is deliberate, not accidental: A player adopts a mixed strategy not because they are
undecided, but because doing so forces the opponent into a position where no action is more profitable than
any other.
• Indifference is the equilibrium anchor: The specific probabilities used in a mixed strategy are derived
precisely by solving for the values that equalize the opponent's expected payoffs.
• Deviation becomes irrational under indifference: Once the opponent is made indifferent, they gain
nothing by deviating from their equilibrium mix. This mutual stability is what makes the outcome a Nash
Equilibrium.
• Pure strategies as boundary cases: A pure strategy is simply a mixed strategy in which one action
receives a probability of one and all others receive zero.

In summary, the Indifference Principle reveals that in mixed strategy Nash Equilibria, the probability
assignments are not guesses or habits — they are mathematically derived values that sustain strategic balance
by removing any exploitable asymmetry in the opponent's expected returns.

NUMERICAL QUESTION

QUESTION 1
Vendor Allocation at a Festival
Topics: Normal Form Games | Dominant Strategy | Nash Equilibrium
Total Marks: 18

Background
At an annual city festival, two competing food vendors Meera and Rajan are each privately informed by the
festival organizer that a single high-footfall corner stall has become available. The organizer contacts both
vendors at the same time and asks each of them to submit their preference either to file a claim for the stall
or to step back and let the other have it. Crucially, neither vendor is told what the other has decided before
submitting their own response.

The outcomes based on their combined choices are structured as follows:

Situation Who Benefits Other Party Final Outcome

Only Meera files claim Secures premium spot Receives nothing Meera gets spot
Only Rajan files claim Receives nothing Secures premium spot Rajan gets spot
Both file claims Assigned to outsider Assigned to outsider Neither benefits
Neither files claim Random draw, 50% chance Random draw, 50% chance Outcome uncertain
It is understood by both parties that this setup and all its consequences are mutually known each vendor is
aware that the other has received identical information and understands the full consequences of every
possible outcome.

Questions

Part (a) [4 Marks]

Break down the three essential structural elements that formally define this interaction as a game. On the
basis of these elements, construct a reasoned argument for why this particular situation qualifies as a Normal
Form Game in the game-theoretic sense.

Part (b) [5 Marks]

Examine whether either vendor possesses a strictly dominant strategy. Invoke the precise theoretical
definition of strict dominance and verify every relevant case methodically. Arrive at a well-supported
conclusion regarding the existence or absence of dominant strategies in this game.

Part (c) [5 Marks]

Identify all pure strategy Nash Equilibria present in this game. For every possible combination of strategies,
examine whether either vendor would gain an advantage by independently changing their decision. Clearly
distinguish stable outcomes from unstable ones.

Part (d) [4 Marks]

Evaluate whether the scenario in which both vendors step back constitutes a Nash Equilibrium. Ground your
answer in theoretical reasoning rather than intuition alone.

NUMERICAL ANSWER

[Link]

Submission 27: Akd

THEORY QUESTION

Define the concept of "iterated strict dominance." Then, provide a theoretical argument explaining whether
the process of iteratively eliminating strictly dominated strategies can ever result in the elimination of a pure
strategy Nash [Link] by contradiction.

THEORY ANSWER

Iterated Strict Dominance is the process where:

Players repeatedly remove such strictly dominated strategies (i.e., strategies that are always worse than
another).
- After each round of elimination, the game becomes smaller.

Players then re-evaluate the reduced game to eliminate more dominated [Link] process continues
until no further eliminations are possible.

Pure Strategy Nash Equilibrium

A pure strategy Nash equilibrium is a set of strategies (one for each player) where no player can increase
their own payoff by unilaterally changing their action, while others keep theirs constant.A Pure Strategy
Nash Equilibrium is a stable situation in a game where:

The process of iterated elimination of strictly dominated strategies never eliminates any pure strategy Nash
Equilibrium of the original game.

Proof (By Contradiction)

Step 1: Assumption
Assume, for contradiction, that iterated strict dominance eliminates a Pure Strategy Nash Equilibrium.

Let the equilibrium be:

s* = (s1*, s2*, ..., sn*)

Step 2: First Elimination

Since elimination happens in rounds, there must be a first round where part of the equilibrium is removed.

Let:

Round k be the first round where a strategy si* (belonging to player i) is eliminated.

Step 3: Reason for Elimination

If si* is eliminated in Round k, then it must be strictly dominated by another strategy si′ that is still available.

This means:

ui(si′, s−i) > ui(si*, s−i)

for all s−i belonging to S−i^k

Where:

S−i^k = strategies of other players that survive till round k

Step 4: Deriving the Contradiction

Since Round k is the first elimination affecting equilibrium:

All other equilibrium strategies s−i* must still exist in Round k

Thus:

s−i* belongs to S−i^k

Substituting into the inequality:

ui(si′, s−i*) > ui(si*, s−i*

Step 5: Contradiction with Nash Equilibrium

The above inequality implies:

Player i can get a strictly higher payoff by switching from si* to si′
But this contradicts the definition of Nash Equilibrium, which states:

ui(si*, s−i*) ≥ ui(si, s−i*)

for any alternative strategy si

Conclusion
Since our assumption leads to a contradiction, it must be false.

Therefore:
Iterated elimination of strictly dominated strategies never removes any Pure Strategy Nash Equilibrium.

NUMERICAL QUESTION

Case Study: Packaged Drinking Water Market

A local market has two firms selling the same product, such as packaged drinking water. The market demand
is given by the inverse demand function:
P = 100 - (q1 + q2)

where:

q1 = output of Firm 1 (Leader)

q2 = output of Firm 2 (Follower)

Firm 1 is the market leader and has a constant marginal cost:

c1 = 20
Firm 2 is the follower, but its cost is uncertain:

With probability p = 0.4, marginal cost is low:

cL = 10

-With probability 1 - p = 0.6, marginal cost is high:

cH = 30

Firm 2 knows its own cost before choosing output, but Firm 1 o

QUESTIONS

1. Find Firm 2’s best response in each cost state.

2. Find the Stackelberg–Bayesian equilibrium outputs of both firms.

3. Compute profits of both firms in both cost states.

4. Explain:

Whether the leader always has an advantage

How Firm 1’s output changes if the probability of low cost increases

NUMERICAL ANSWER

[Link]
 Submission 28: the nash equilibrium

THEORY QUESTION

Explain the concept of Iterated Elimination of Strictly Dominated Strategies (IESDS). Illustrate the process
with an example.

THEORY ANSWER

Iterated Elimination of Strictly Dominated Strategies (IESDS):

IESDS is a method used in game theory to simplify a game by removing strategies that are strictly worse
than another strategy for a player, regardless of what the opponent does.

A strategy is strictly dominated if there exists another strategy that gives a better payoff in every possible
situation.

The elimination is done step-by-step (iteratively) until no more dominated strategies remain.

Steps in IESDS:
1. Identify strictly dominated strategies for any player.
2. Remove those strategies from the game.
3. Re-evaluate the reduced game.
4. Repeat until no dominated strategies remain.

Example:

Consider the payoff matrix:

B1 B2
A1 (2, 2) (1, 3)
A2 (3, 1) (0, 2)

Step 1: Check Player A

• Compare A1 and A2:
• Against B1: A2 (3) > A1 (2)
• Against B2: A1 (1) > A2 (0)
→ No strict dominance → nothing removed
Step 2: Check Player B
• Compare B1 and B2:
• Against A1: B2 (3) > B1 (2)
• Against A2: B2 (2) > B1 (1)

→ B1 is strictly dominated by B2, so eliminate B1

Step 3: Reduced Game

B2
A1 (1, 3)
A2 (0, 2)

Now compare A1 and A2:

• A1 gives better payoff than A2 → eliminate A2

Final Outcome:
Only strategy left is (A1, B2)

Conclusion:
IESDS helps simplify complex games and can lead to a unique solution, but it only works when strictly
dominated strategies exist.

NUMERICAL QUESTION

For each one of the following normal form games find

(a) The pure-strategy Nash equilibria and their payoffs;
(b) The mixed-strategy Nash equilibria and their payoffs;
(c) Are there any equilibria in dominant strategies? If yes, which ones?

NUMERICAL ANSWER

[Link]

Submission 29: SITUS

THEORY QUESTION
In the classic "Battle of the Sexes" game, a couple wishes to go to an event together but prefers
different events (e.g., Player 1 prefers Boxing, Player 2 prefers Ballet). If they choose simultaneously,
there are two pure-strategy Nash Equilibria and a coordination problem.

Suppose we alter the game into an extensive form game with perfect information: Player 1
moves first and has the option to "burn their bridge" by publicly purchasing a non-refundable,
non-transferable ticket to Boxing before Player 2 makes any decisions.

Using the concepts of Subgame Perfect Equilibrium and Forward Induction, explain theoretically
why Player 1 will choose to burn the bridge, and why "limiting one's own options" paradoxically
increases a player's payoff in this scenario.

THEORY ANSWER

In a simultaneous Battle of the Sexes game, neither player has a dominant strategy, leading to
uncertainty and the risk of miscoordination (where both end up at different venues and receive a
payoff of 0).

By changing the game to an extensive form where Player 1 moves first by buying a non-
refundable ticket, Player 1 creates a credible commitment device. This fundamentally alters the
subgames that follow. Once Player 1 has "burned the bridge" (committed to Boxing), their payoff
for choosing Ballet in any subsequent interaction becomes effectively negative (they lost the ticket
money and are at the wrong venue). Player 2 observes this commitment.

When it is Player 2's turn to move, Player 2 knows that Player 1 must go to Boxing to avoid a
massive loss. Therefore, Player 2's subgame reduces to a simple choice: go to Ballet alone (payoff
0) or go to Boxing with Player 1 (payoff > 0, even if it is not their preferred event). A rational
Player 2 will choose Boxing.

By applying backward induction, Player 1 anticipates that if they commit to Boxing, Player 2's
best response will be to follow suit. The Subgame Perfect Equilibrium results in Player 1 playing
"Buy Boxing Ticket" and Player 2 playing "Go to Boxing if Player 1 bought ticket".

Paradoxically, by removing their own ability to compromise, Player 1 forces the other player to
be the one who compromises, effectively eliminating the coordination problem and securing their
most preferred outcome.

NUMERICAL QUESTION

Two firms, Firm 1 and Firm 2, compete in a market by choosing quantities q1 and q2. The inverse
market demand is given by P = 100 − Q, where Q = q1 + q2.

Firm 1 has a constant marginal cost c1 = 10. Firm 1 knows its own cost, but Firm 2’s cost is
a secret. Firm 1 only knows that Firm 2 is equally likely to be a "High Cost" type (cH = 20) or a
"Low Cost" type (cL = 10). Firm 2 knows its own actual cost, as well as Firm 1’s cost.

Find the Bayesian Nash Equilibrium quantities for Firm 1 (q*1) and Firm 2 (qH2 and qL2).

NUMERICAL ANSWER

[Link]