UNIT-IV

Game Theory & Simulation

Game Theory
Game theory applications have been developed for situations in which the competing players
are teams, companies, political candidates, armies, and contract bidders.

Each player selects a strategy independently without knowing in advance the strategy of the
other player or players.

Game theory: The study of decision situations in which two or more players compete as
adversaries. The combination of the competing strategies provides the value of the game to
the players.

Two-Person, zero-sum game: A game with two players in which the gain to one player is
equal to the loss to the other player, so that total sum is Zero.

Pure Strategy: A game solution that provides a single best strategy for each player. In other
words, the player will select only one particular strategy and ignores the other player’s
strategy.
➢ The objective of the players is to maximize their gains or minimize their losses.
➢ If a particular probability Pj = 1 (j = 1,2,…, n) and all others are zero, the player is
said to select pure strategy j.

Mixed Strategy: A set of strategies that a player chooses on a particular move of the game
with some fixed probability are called mixed strategies. Thus, there is a probabilistic situation
and objective of each player is to maximize their gains or minimize their losses by making
the choice among the pure strategies with fixed probabilities.
➢ The sum of the probabilities is equal to one, but each probability value is less than 1.

Payoff matrix: The payoffs (a quantitative measure of satisfaction that a player gets at the
end of the play) in terms of gains or losses, when players select their particular strategies
(course of action), can be represented in the form of a matrix, called the payoff matrix.
Player A’s Strategies Player B’s Strategies
B1 B2 ……… Bn
A1 a11 a12 ……… a1n
A2 a21 a22 ……… a2n
….. ………
…….

…….

…….

An am1 am2 ……… am2

This pay-off matrix is drawn from A’s point of view, assuming that the player A is gainer,
therefore he wishes to gain as large a payoff aij as possible, player B on the other hand would
do his best to reach as small a value of aij as possible.
Note: (i) If the pay-off matrix outcome is positive then it is gain to player A and loss to
player B.
(ii) If any negative outcome then it is loss to player A and gain to player B.

Maximin Principle: maximizes the minimum guaranteed gains of player A.

Minimax Principle: minimizes the maximum losses of player B.

Saddle Point: it is the payoff value that represents both the minimum of the column maxima
and the maximum of the row minima.

Optimal Strategy: if the maximin value equals the minimax value, then the game is said to
have a saddle point and the corresponding strategies are called optimal strategies.

Value of the game (V): The payoff amount in the saddle-point position is also called value
of the game.

➢ A game may have more than one saddle points. A game with no saddle point is solved
by choosing strategies with fixed probabilities.
➢ The value of the game, in general, satisfies the equation: maximum value ≤ V ≤
minimax value.
➢ A game is said to be a fair game if the lower (maximin) and upper (minimax) values
of the game are equal and both equals zero.
➢ A game is said to be strictly determinable if the lower (maximin) and upper
(minimax) values of the game are equal and both equal the value of the game.
Simulation
• A simulation model uses the mathematical expressions and logical relationships of a
real system.

• Values for the probabilistic inputs to a simulation are randomly generated based on
historical information.

• A quantity that is difficult to measure with certainty is called a probabilistic input.

• A value for probabilistic input from a discrete probability distribution is given by

matching the probabilistic input with an interval of random numbers.

• In order to verify a simulation model, be sure that the procedures for calculations are
logically correct.

• A simulation model used in situations where the state of the system at one point in
time does not affect the state of the system at future points in time is called a static
simulation model.
1. ABC Plumbing and heating maintains a stock of 35-gallon hot water heaters that it sells to
home-owners and installs for them. The owner likes the idea of having large supply on
hand so as to meet all customer demand, but he also recognizes that it is expensive to do
so. He examines hot water heater sales over the past 54 weeks and notes the following:

Hot water heater Number of weeks this

sales per week number was sold
8 4
6 8
5 12
7 10
9 6
11 9
12 5
Total 54
Using the random numbers given below, simulate demand for 10 weeks and answer the
following questions:
(i) If ABC maintains a constant supply of 10 hot water heaters in any given week, how many
times will he be out-of-stock during the 10-week simulation period?
(ii) What is the average number of heaters demanded per week over the 10-week interval?
Random Numbers: 15, 28, 05, 35, 28, 60, 95, 34, 50, 45.
Solution:
First allocate random numbers 00-99 in proportion to the probabilities associated with
demand of the heaters per day, as shown in the table 1:
Table 1: Allocation of Random numbers
Demand per No. of weeks Probability Cumulative Random
week Probability number
interval
8 4 0.07 0.07 00-06
6 8 0.15 0.22 07-21
5 12 0.22 0.44 22 - 43
7 10 0.19 0.63 44- 62
9 6 0.11 0.74 63- 73
11 9 0.17 0.91 74- 90
12 5 0.09 1.00 91 -99
Total 54 1.00

Based on the given random numbers, simulate the demand for the heaters for next 10 weeks,
as shown in table 2:
Table 2: Simulation Worksheet
Day Random number Demand
1 15 6
2 28 5
3 05 8
4 35 5
5 28 5
6 60 7
7 95 12
8 34 5
9 50 7
10 45 7
Total 67

From Table2:
(i) Number of weeks he is expected to be out-of-stock = 1
Total demand 67
(ii) Average demand for heaters per week = = = 6.7
Number of weeks 10

2. A company manufactures 100 toy vehicles per day. The sale of toy vehicles depends upon
demand which has the following distribution:
Sale of toy vehicles (Nos.) Probability
97 0.10
98 0.15
99 0.20
100 0.35
102 0.15
103 0.05

There is no carry-over of inventory. The following details are given below:

 Variable production cost per toy vehicle ₹ 20
Selling Price per toy vehicle ₹ 25
Penalty attracted per unsold toy vehicle ₹2
Penalty attracted per unit of demand not met ₹ 1

Random numbers to be used: 10, 99, 20, 25, 58, 85, 66, 70, 05, 40.
Estimate the profit/loss for the next ten days using the above random numbers and
assuming 100 toy vehicles are produced per day.

Solution:
First allocate random numbers 00-99 in proportion to the probabilities associated with
demand of the toy vehicles produced per day, as shown in the table 1:

Table 1: Allocation of Random numbers

Demand Probability Cumulative Probability Random number interval

97 0.10 0.10 00-09
98 0.15 0.25 10-24
99 0.20 0.45 25-44
100 0.35 0.80 45-79
102 0.15 0.95 80-94
103 0.05 1.00 95-99

Based on the given random numbers and assuming 100 toy vehicles produced per day
simulate the demand for the toy vehicles for next 10 weeks, as shown in table 2, based on the
information relating to the toy vehicles produced and sold determine the unsatisfied demand
and the number of toy vehicles unsold, in order to calculate the total cost.

Table 2: Simulation Worksheet

Day Random Demand No. of toy No. of toy Unsatisfied No. of toy
No. vehicles vehicles demand vehicles
produced sold unsold
1 10 98 100 98 0 2
2 99 103 100 100 3 0
3 20 98 100 98 0 2
4 25 99 100 99 0 1
5 58 100 100 100 0 0
6 85 102 100 100 2 0
7 66 100 100 100 0 0
8 70 100 100 100 0 0
9 05 97 100 97 0 3
10 40 99 100 99 0 1
Total 996 1,000 991 5 9

The calculation of company’s profit/loss is given in table 3:

 Table 3: Calculation of company’s profit/loss
Number Unit rate (₹) Amount (₹)
Sale of toy vehicle 991 25 991*25 = 24,775
Less: Variable 1,000 20 1,000*20 = 20,000
production cost
Less: Penalty on 5 1 5*1 = 5
unsatisfied demand
Less: Penalty on 9 2 9*2 = 18
unsold toy vehicles
Profit/loss 4,752

Thus, the simulated profit for 10 day’s operation is ₹ 4,752.

3. A dentist schedules all his patients for 30-minute appointments. Some of the patients take
more 30 minutes some less, depending on the type of dental work to be done. The
following summary shows the various categories of work, their probabilities and time
actually, needed to complete the work:
Category of service Time Required Number of patients
(minutes)
Filling 45 35
Crown 60 25
Cleaning 15 55
Extraction 45 25
Checkup 15 55
Simulate the dentist’s clinic for four hours and determine the average waiting time for the
patients as well as the idleness of the doctor. Assume that all the patients show up at the
clinic at exactly their scheduled arrival time starting at 9.00 a.m. Use the following random
numbers for handling the above problem: 65, 77, 25, 8, 33, 21, 89,10.

Solution:
First allocate random numbers 00-99 in proportion to the probabilities, as shown in the table 1:
Table 1: Allocation of Random numbers
Service Time
Required Random
Category of (minutes) No. of Cumulative Numbers
service patients Probability Probability Interval
Filling 45 35 0.18 0.18 00-17
Crown 60 25 0.13 0.31 18-30
Cleaning 15 55 0.28 0.59 31-58
Extraction 45 25 0.13 0.72 59-71
Checkup 15 55 0.28 1.00 72-99
Total =
195
 Table 2: Arrival Pattern and Nature of Service
Service Time
Needed
Patient Number Random Number Category of service (minutes)
1 65 Extraction 45
2 77 Check up 15
3 25 Crown 60
4 8 Filling 45
5 33 Cleaning 15
6 21 Crown 60
7 89 Check up 15
8 10 Filling 45

Table 3: Computation of Arrivals, Service end time, waiting of patients, Idle time of the
Doctor
Service
Patient Service duration Service Waiting Idle Time
Number Arrival start (in minutes) Ends (in minutes) (in minutes)
1 9.00 a.m 9.00 45 9.45 a.m 0 0
2 9.30a.m 9.45 15 10.00 a.m 15 0
3 10.00a.m 10.00 60 11.00 a.m 0 0
4 10.30a.m 11.00 45 11.45 a.m 30 0
5 11.00a.m 11.45 15 12.00 noon 45 0
6 11.30a.m 12.00 60 1.00 p.m 30 0
7 12.00noon 1.00 15 1.15 p.m 60 0
8 12.30p.m 1.15 45 2.00 p.m 45 0
Total time 225 0
Average
time 28.12

The dentist was not idle even once during the entire simulated period.
The total waiting time of the patients = 225.
The average waiting time = 28.12 minutes.

4. Inter-arrival and service durations studied over past few years for a single channel queuing
system revealed the following patterns:
Inter-arrival time Service time
Minutes Probability Minutes Probability
2 0.19 1 0.15
4 0.22 3 0.28
6 0.32 5 0.30
8 0.17 7 0.17
10 0.10 9 0.10
Use the following series of random numbers, simulate the queue behavior for a
period of 60 minutes.
Arrivals: 19, 32, 59, 81, 27, 45, 26, 52, 77, 46.
Service: 08, 27, 74, 96, 48, 07, 65, 78, 92,49.
Use the above simulation to estimate:
 (i) the mean time spent by the customer waiting for service.
(ii) the probability of the server being idle.

Solution:
First allocate random numbers 00-99 in proportion to the probabilities of inter-arrival time &
service time, as shown in the table 1 & 2:

Table 1: Allocation of Random numbers for Inter-arrival time

Inter-arrival time Cumulative

(Minutes) Probability Probability Random Number Interval
2 0.19 0.19 00 - 18
4 0.22 0.41 19 - 40
6 0.32 0.73 41 - 72
8 0.17 0.90 73 - 89
10 0.1 1.00 90 -99

Table 2: Allocation of Random numbers for Service time

Cumulative Random Number
Service time (Minutes) Probability Probability Interval
1 0.15 0.15 00 - 14
3 0.28 0.43 15 - 42
5 0.30 0.73 43 - 72
7 0.17 0.90 73 - 89
9 0.10 1.00 90 - 99
The simulation worksheet developed for the given problem is shown in table 3.
Table 3: Simulation Worksheet
Inter-
arrival Arrival Service service
Arrival Random time time Random Service Starts finish Waiting Idle
Number number (Minutes) (Minutes) number Time (Minutes) (Minutes) time time
1 19 4 4 8 1 4 5 0 4
2 32 4 8 27 3 8 11 0 3
3 59 6 14 74 7 14 21 0 1
4 81 8 22 96 9 22 31 0 0
5 27 4 26 48 5 31 36 5 0
6 45 6 32 7 1 36 37 4 0
7 26 4 36 65 5 37 42 1 0
8 52 6 42 78 7 42 49 0 1
9 77 8 50 92 9 50 59 0 0
10 46 6 56 49 5 59 64 3 0
Total 13 9

(i) Average waiting time = 13/10 = 1.3

(ii) Probability of ideal time = 9/64 = 0.140

Author: Dr. Y L P Thorani