Or Departmental Exam

Decision Analysis
To accompany
Quantitative Analysis for Management, Tenth Edition,
by Render, Stair, and Hanna
Power Point slides created by Jeff Heyl © 2009 Prentice-Hall, Inc.
Learning Objectives
After completing this chapter, students will be able to:
1. List the steps of the decision-

making process
2. Describe the types of decision-
making environments
3. Make decisions under uncertainty
4. Use probability values to make
decisions under risk
5. Develop accurate and useful
decision trees
6. Use computers to solve basic
decision-making problems
Chapter Outline
1 Introduction
2 The Six Steps in Decision
Making
3 Types of Decision-Making
Environments
4 Decision Making under
Uncertainty
5 Decision Making under Risk
6 Decision Trees
Introduction
n What is involved in making a good

decision?
n Decision theory is an analytic and
systematic approach to the study of
decision making
n A good decision is one that is based
on logic, considers all available data
and possible alternatives, and the
quantitative approach described here
The Six Steps in Decision Making
1. Clearly define the problem at hand

2. List the possible alternatives
3. Identify the possible outcomes or states
of nature
4. List the payoff or profit of each
combination of alternatives and
outcomes
5. Select one of the mathematical decision
theory models
6. Apply the model and make your decision
Thompson Lumber Company
Step 1 – Define the problem

n Expand by manufacturing and
marketing a new product, backyard
storage sheds
Step 2 – List alternatives
n Construct a large new plant
n A small plant
n No plant at all
Step 3 – Identify possible outcomes
n The market could be favorable or
unfavorable
Step 4 – List the payoffs

n Identify conditional values for the
profits for large, small, and no plants
for the two possible market conditions
Step 5 – Select the decision model
n Depends on the environment and
amount of risk and uncertainty
Step 6 – Apply the model to the data
n Solution and analysis used to help the
decision making
STATE OF NATURE
FAVORABLE UNFAVORABLE
ALTERNATIVE MARKET ($) MARKET ($)
Construct a large plant 200,000 –180,000
Construct a small plant 100,000 –20,000
Do nothing 0 0
Types of Decision-Making
Environments
Type 1: Decision making under certainty

n Decision maker knows with certainty the
consequences of every alternative or
decision choice
Type 2: Decision making under uncertainty
n The decision maker does not know the
probabilities of the various outcomes
Type 3: Decision making under risk
n The decision maker knows the
probabilities of the various outcomes
Decision Making Under
Uncertainty
There are several criteria for making decisions
under uncertainty
1. Maximax (optimistic)
2. Maximin (pessimistic)
3. Criterion of realism (Hurwicz)
4. Equally likely (Laplace)
5. Minimax regret
Maximax
Used to find the alternative that maximizes
the maximum payoff
n Locate the maximum payoff for each alternative
n Select the alternative with the maximum
number
STATE OF NATURE
FAVORABLE UNFAVORABLE MAXIMUM IN
ALTERNATIVE MARKET ($) MARKET ($) A ROW ($)
Construct a large
plant 200,000 –180,000 200,000
Maximax
Construct a small
100,000 –20,000 100,000
plant
Do nothing 0 0 0
Maximin
Used to find the alternative that maximizes
the minimum payoff
n Locate the minimum payoff for each alternative
n Select the alternative with the maximum
number
STATE OF NATURE
FAVORABLE UNFAVORABLE MINIMUM IN
Construct a large
plant 200,000 –180,000 –180,000
Construct a small
plant 100,000 –20,000 –20,000
Do nothing 0 0 0
Maximin
Using QM to Solve Decision
Theory Problems
Theory Problems
Criterion of Realism (Hurwicz)
A weighted average compromise between
optimistic and pessimistic
n Select a coefficient of realism a
n Coefficient is between 0 and 1
n A value of 1 is 100% optimistic
n Compute the weighted averages for each
alternative
n Select the alternative with the highest value
Weighted average = a(maximum in row)

+ (1 – a)(minimum in row)
Criterion of Realism (Hurwicz)
n For the large plant alternative using a = 0.8
(0.8)(200,000) + (1 – 0.8)(–180,000) = 124,000
n For the small plant alternative using a = 0.8
(0.8)(100,000) + (1 – 0.8)(–20,000) = 76,000
STATE OF NATURE
CRITERION
FAVORABLE UNFAVORABLE OF REALISM
ALTERNATIVE MARKET ($) MARKET ($) (a = 0.8)$
Construct a large
200,000 –180,000 124,000
plant
Realism
Construct a small
100,000 –20,000 76,000
plant
Do nothing 0 0 0
Theory Problems
Theory Problems
Equally Likely (Laplace)
Considers all the payoffs for each alternative
n Find the average payoff for each alternative
n Select the alternative with the highest average
STATE OF NATURE
FAVORABLE UNFAVORABLE ROW
ALTERNATIVE MARKET ($) MARKET ($) AVERAGE ($)
Construct a large
plant 200,000 –180,000 10,000
Construct a small
plant 100,000 –20,000 40,000
Eq ually likely
Do nothing 0 0 0
Minimax Regret
Based on opportunity loss or regret, the
difference between the optimal profit and
actual payoff for a decision
n Create an opportunity loss table by determining
the opportunity loss for not choosing the best
alternative
n Opportunity loss is calculated by subtracting
each payoff in the column from the best payoff
in the column
n Find the maximum opportunity loss for each
alternative and pick the alternative with the
minimum number
Minimax Regret
n Opportunity
Loss Tables
STATE OF NATURE
MARKET ($) MARKET ($)
200,000 – 200,000 0 – (–180,000)
200,000 – 100,000 0 – (–20,000)
200,000 – 0 0–0
STATE OF NATURE
ALTERNATIVE MARKET ($) MARKET ($)
Construct a large plant 0 180,000
Construct a small plant 100,000 20,000
Do nothing 200,000 0
Minimax Regret
STATE OF NATURE
FAVORABLE UNFAVORABLE MAXIMUM IN
Construct a large
plant 0 180,000 180,000
Construct a small
100,000 20,000 100,000
plant Minimax
Do nothing 200,000 0 200,000
Decision Making Under Risk
n Decision making when there are several possible
states of nature and we know the probabilities
associated with each possible state
n Most popular method is to choose the alternative
with the highest expected monetary value (EMV)
EMV (alternative i) = (payoff of first state of nature)

x (probability of first state of nature)
+ (payoff of second state of nature)
x (probability of second state of nature)
+ … + (payoff of last state of nature)
x (probability of last state of nature)
EMV for Thompson Lumber
n Each market has a probability of 0.50
n Which alternative would give the highest EMV?
n The calculations are
EMV (large plant) = (0.50)($200,000) + (0.50)(–$180,000)

= $10,000
EMV (small plant) = (0.50)($100,000) + (0.50)(–$20,000)
= $40,000
EMV (do nothing) = (0.50)($0) + (0.50)($0)
= $0
EMV for Thompson Lumber
STATE OF NATURE
ALTERNATIVE MARKET ($) MARKET ($) EMV ($)
Construct a large
plant 200,000 –180,000 10,000
Construct a small
plant 100,000 –20,000 40,000
Do nothing 0 0 0
Probabilities 0.50 0.50
Largest EMV
EMV for Thompson Lumber Using
QM for Windows
Largest EMV
Expected Value of Perfect
Information (EVPI)
n EVPI places an upper bound on what you should
pay for additional information
EVPI = EVwPI – Maximum EMV
n EVwPI is the long run average return if we have
perfect information before a decision is made
EVwPI = (best payoff for first state of nature)

x (probability of first state of nature)
+ (best payoff for second state of nature)
x (probability of second state of nature)
+ … + (best payoff for last state of nature)
x (probability of last state of nature)
Information (EVPI)
n Scientific Marketing, Inc. offers analysis
that will provide certainty about market
conditions (favorable)
n Additional information will cost $65,000
n Is it worth purchasing the information?
Information (EVPI)
1. Best alternative for favorable state of nature is
build a large plant with a payoff of $200,000
Best alternative for unfavorable state of nature is
to do nothing with a payoff of $0
EVwPI = ($200,000)(0.50) + ($0)(0.50) = $100,000
2. The maximum EMV without additional
information is $40,000
EVPI = EVwPI – Maximum EMV
= $100,000 - $40,000
= $60,000
So the maximum Thompson
should pay for the additional
information is $60,000
Information (EVPI)
Expected Opportunity Loss
n Expected opportunity loss (EOL) is the

cost of not picking the best solution
n First construct an opportunity loss table
n For each alternative, multiply the
opportunity loss by the probability of that
loss for each possible outcome and add
these together
n Minimum EOL will always result in the
same decision as maximum EMV
n Minimum EOL will always equal EVPI
Expected Opportunity Loss
STATE OF NATURE
ALTERNATIVE MARKET ($) MARKET ($) EOL
Construct a large plant 0 180,000 90,000
Construct a small
plant 100,000 20,000 60,000
Do nothing 200,000 0 100,000

Probabilities 0.50 0.50
Minimum EOL
EOL (large plant) = (0.50)($0) + (0.50)($180,000)
= $90,000
EOL (small plant) = (0.50)($100,000) + (0.50)($20,000)
= $60,000
EOL (do nothing) = (0.50)($200,000) + (0.50)($0)
= $100,000
Sensitivity Analysis
n Sensitivity analysis examines how our decision
might change with different input data
n For the Thompson Lumber example
P = probability of a favorable market

1 – P = probability of an unfavorable market
EMV(Large Plant) = $200,000P – $180,000)(1 – P)
= $200,000P – $180,000 + $180,000P
= $380,000P – $180,000
EMV(Small Plant) = $100,000P – $20,000)(1 – P)
= $100,000P – $20,000 + $20,000P
= $120,000P – $20,000
EMV(Do Nothing) = $0P + 0(1 – P)
= $0
EMV Values
$300,000
$200,000 EMV (large plant)

Point 2
$100,000 EMV (small plant)

Point 1
0 EMV (do nothing)

.167 .615 1
–$100,000 Values of P
–$200,000
Point 1:
EMV(do nothing) = EMV(small plant)
20,000
0 = $120,000P - $20,000 P= = 0.167
120,000
Point 2:
EMV(small plant) = EMV(large plant)
$120,000P - $20,000 = $380,000P - $180,000
160,000
P= = 0.615
260,000
BEST RANGE OF P
ALTERNATIVE VALUES
Do nothing Less than 0.167
EMV Values Construct a small plant 0.167 – 0.615
$300,000 Construct a large plant Greater than 0.615
$200,000 EMV (large plant)

Point 2
$100,000 EMV (small plant)

Point 1
0 EMV (do nothing)

.167 .615 1
–$100,000 Values of P
–$200,000
Decision Trees
n Any problem that can be presented in a
decision table can also be graphically
represented in a decision tree
n Decision trees are most beneficial when a
sequence of decisions must be made
n All decision trees contain decision points
or nodes and state-of-nature points or
nodes
n A decision node from which one of several
alternatives may be chosen
n A state-of-nature node out of which one state
of nature will occur
Five Steps to
Decision Tree Analysis
1. Define the problem

2. Structure or draw the decision tree
3. Assign probabilities to the states of
nature
4. Estimate payoffs for each possible
combination of alternatives and states of
nature
5. Solve the problem by computing
expected monetary values (EMVs) for
each state of nature node
Structure of Decision Trees
n Trees start from left to right
n Represent decisions and outcomes in
sequential order
n Squares represent decision nodes
n Circles represent states of nature nodes
n Lines or branches connect the decisions
nodes and the states of nature
Thompson’s Decision Tree
A State-of-Nature Node
Favorable Market
A Decision Node
1
Unfavorable Market
Favorable Market
Construct
Small Plant
2
Unfavorable Market
Thompson’s Decision Tree
EMV for Node = (0.5)($200,000) + (0.5)(–$180,000)
1 = $10,000 Payoffs
Favorable Market (0.5)
$200,000
Alternative with best
EMV is selected 1
Unfavorable Market (0.5)
–$180,000

$100,000
Construct
Small Plant
2
–$20,000
EMV for Node = (0.5)($100,000)

2 = $40,000 + (0.5)(–$20,000)
$0
Thompson’s Complex
Decision Tree
Suppose John Thompson has two decisions to
make, with the second decision dependent on the
outcome of the first. Before deciding about
building a new plant, John has the option of
conducting his own marketing research survey at
a cost of $10,000.
n There is a 45% chance that the survey results will
indicate a favorable market (hence, 0.55 is the
probability that the survey results will be
negative).
n There is a 78% chance of a favorable market
given a favorable result from the market survey.
n There is a 27% chance of a favorable market
given that the survey results are negative.
Thompson’s Complex Decision Tree
First Decision Second Decision Payoffs
Point Point
Favorable Market (0.78) $190,000
2 Unfavorable Market (0.22) –$190,000
Small $90,000
Plant 3 Unfavorable Market (0.22)
–$30,000
No Plant
–$10,000
1 Favorable Market (0.27) $190,000
4 Unfavorable Market (0.73) –$190,000
Small $90,000
Plant 5 Unfavorable Market (0.73)
–$30,000
No Plant
–$10,000

$200,000
6 Unfavorable Market (0.50)
–$180,000
Small $100,000
Plant 7 Unfavorable Market (0.50) –$20,000
No Plant
$0
1. Given favorable survey results,

EMV(node 2) = EMV(large plant | positive survey)
= (0.78)($190,000) + (0.22)(–$190,000) = $106,400
EMV(node 3) = EMV(small plant | positive survey)
= (0.78)($90,000) + (0.22)(–$30,000) = $63,600
EMV for no plant = –$10,000
2. Given negative survey results,
EMV(node 4) = EMV(large plant | negative survey)
= (0.27)($190,000) + (0.73)(–$190,000) = –$87,400
EMV(node 5) = EMV(small plant | negative survey)
= (0.27)($90,000) + (0.73)(–$30,000) = $2,400
EMV for no plant = –$10,000
3. Compute the expected value of the market survey,

EMV(node 1) = EMV(conduct survey)
= (0.45)($106,400) + (0.55)($2,400)
= $47,880 + $1,320 = $49,200
4. If the market survey is not conducted,
EMV(node 6) = EMV(large plant)
= (0.50)($200,000) + (0.50)(–$180,000) = $10,000
EMV(node 7) = EMV(small plant)
= (0.50)($100,000) + (0.50)(–$20,000) = $40,000
EMV for no plant = $0
5. Best choice is to seek marketing information
First Decision Second Decision Payoffs
Point Point
$106,400 Favorable Market (0.78)
$190,000
Unfavorable Market (0.22) –$190,000
$106,400
Small $90,000
Plant –$30,000
No Plant
$49,200 –$10,000
–$87,400 Favorable Market (0.27)
$190,000
$2,400
Small $90,000
Plant –$30,000
No Plant
–$10,000
$49,200

$200,000
–$180,000
$40,000

Small $100,000
Plant
No Plant
$0
Expected Value of Sample Information
n Thompson wants to know the actual value

of doing the survey
Expected value Expected value
with sample of best decision
EVSI = information, assuming – without sample
no cost to gather it information
= (EV with sample information + cost)

– (EV without sample information)
EVSI = ($49,200 + $10,000) – $40,000 = $19,200
Is the $10,000–market survey worthwhile?

Exercises
1. Jerry is thinking about opening a bicycle shop in his
hometown. Jerry loves to take his own bike on 50-mile
trips with his friends, but he believes that any small
business should be started only if there is a good chance
of making a profit. Jerry can open a small shop, a large
shop, or no shop at all. Because there will be a 5-year
lease on the building that Jerry is thinking about using,
he wants to make sure that he makes the correct
decision. Jerry is also thinking about hiring his old
marketing professor to conduct a marketing research
study. If the study is conducted, the results could be
either favorable or unfavorable. Develop a decision tree
for Jerry.
Exercises
2. Jerry has done some analysis about the profitability if the
bicycle shop. If he builds the large bicycle shop, he will
earn P2.4 million if the market is favorable, but he will
lose P1.6 million if the market is unfavorable. The small
shop will return a P1.2 million profit in a favorable market
and a P400,000 loss in an unfavorable market. At the
present time, he believes that there is a 50-50 chance that
the market will be favorable. His old marketing professor
will charge him P200,000 for the marketing research. It is
estimated that there is a 0.6 probability that the survey
will be favorable. Furthermore, there is a 0.9 probability
that the market will be favorable given a favorable
outcome of the study. However, the marketing professor
has warned Jerry that there is only a probability of 0.12 of
a favorable market if the marketing research results are
not favorable. Jerry is confused. What should he do?
n How sensitive are the decisions to

changes in the probabilities?
n How sensitive is our decision to the
probability of a favorable survey result?
n That is, if the probability of a favorable
result (p = .45) where to change, would we
make the same decision?
n How much could it change before we would
make a different decision?
p = probability of a favorable survey result
(1 – p) = probability of a negative survey result
EMV(node 1) = ($106,400)p +($2,400)(1 – p)

= $104,000p + $2,400
We are indifferent when the EMV of node 1 is the

same as the EMV of not conducting the survey,
$40,000
$104,000p + $2,400 = $40,000

$104,000p = $37,600
p = $37,600/$104,000 = 0.36
Utility Theory
To accompany
Quantitative Analysis for Management, Eleventh Edition,
Power Point slides created by Brian Peterson
Learning Objective
Understand the importance and use of

utility theory in decision making.
3-2
Utility Theory
n Monetary value is not always a true
indicator of the overall value of the
result of a decision.
n The overall value of a decision is called
utility.
n Economists assume that rational
people make decisions to maximize
their utility.
3-3
Utility Theory
Your Decision Tree for the Lottery Ticket
$2,000,000
Accept
Offer
$0
Heads
Reject (0.5)
Offer
Tails
(0.5)
EMV = $2,500,000
$5,000,000
3-4
Utility Theory
n Utility assessment assigns the worst outcome a
utility of 0, and the best outcome, a utility of 1.
n A standard gamble is used to determine utility
values.
n When you are indifferent, your utility values are
equal.
Expected utility of alternative 2 = Expected utility of alternative 1

Utility of other outcome = (p)(utility of best outcome, which is 1)
+ (1 – p)(utility of the worst outcome,
which is 0)
Utility of other outcome = (p)(1) + (1 – p)(0) = p
3-5
Standard Gamble for Utility
Assessment
(p)
Best Outcome
Utility = 1
(1 – p) Worst Outcome
Utility = 0
Other Outcome
Utility = ?
3-6
Investment Example
n Jane Dickson wants to construct a utility curve
revealing her preference for money between $0
and $10,000.
n A utility curve plots the utility value versus the
monetary value.
n An investment in a bank will result in $5,000.
n An investment in real estate will result in $0 or
$10,000.
n Unless there is an 80% chance of getting $10,000
from the real estate deal, Jane would prefer to
have her money in the bank.
n So if p = 0.80, Jane is indifferent between the bank
or the real estate investment.
3-7
Investment Example
p = 0.80 $10,000
U($10,000) = 1.0
(1 – p) = 0.20 $0
U($0.00) = 0.0
$5,000
U($5,000) = p = 0.80
Utility for $5,000 = U($5,000) = pU($10,000) + (1 – p)U($0)

= (0.8)(1) + (0.2)(0) = 0.8
3-8
Investment Example
n We can assess other utility values in the same way.
n For Jane these are:
Utility for $7,000 = 0.90

Utility for $3,000 = 0.50
n Using the three utilities for different dollar amounts,

she can construct a utility curve.
3-9
Utility Curve
1.0 – U ($10,000) = 1.0
0.9 – U ($7,000) = 0.90
0.8 – U ($5,000) = 0.80
0.7 –
0.6 –
Utility
0.5 – U ($3,000) = 0.50
0.4 –
0.3 –
0.2 –
0.1 –
U ($0) = 0
| | | | | | | | | |
$0 $1,000 $3,000 $5,000 $7,000 $10,000
Monetary Value
3-10
Utility Curve
n Jane’s utility curve is typical of a risk avoider.
n She gets less utility from greater risk.
n She avoids situations where high losses might occur.
n As monetary value increases, her utility curve increases
at a slower rate.
n A risk seeker gets more utility from greater risk

n As monetary value increases, the utility curve increases
at a faster rate.
n Someone with risk indifference will have a linear

utility curve.
3-11
Preferences for Risk
Risk
Avoider
Utility
Risk
Seeker
Figure 3.10
Monetary Outcome
3-12
Utility as a
Decision-Making Criteria
n Once a utility curve has been developed

it can be used in making decisions.
n This replaces monetary outcomes with
utility values.
n The expected utility is computed instead
of the EMV.
3-13
Utility as a
n Mark Simkin loves to gamble.

n He plays a game tossing thumbtacks in
the air.
n If the thumbtack lands point up, Mark wins
$10,000.
n If the thumbtack lands point down, Mark
loses $10,000.
n Mark believes that there is a 45% chance
the thumbtack will land point up.
n Should Mark play the game (alternative 1)?
3-14
Utility as a
Decision Facing Mark Simkin

Tack Lands
Point Up (0.45)
$10,000
Tack Lands
Point Down (0.55)
–$10,000
Figure 3.11 Mark Does Not Play the Game

$0
3-15
Utility as a
n Step 1– Define Mark’s utilities.

U (–$10,000) = 0.05
U ($0) = 0.15
U ($10,000) = 0.30
n Step 2 – Replace monetary values with

utility values.
E(alternative 1: play the game) = (0.45)(0.30) + (0.55)(0.05)
= 0.135 + 0.027 = 0.162
E(alternative 2: don’t play the game) = 0.15
3-16
Utility Curve for Mark Simkin
1.00 –
0.75 –
Utility
0.50 –
0.30 –
0.25 –
0.15 –
0.05 –
0 |– | | | |
–$20,000 –$10,000 $0 $10,000 $20,000
Monetary Outcome
3-17
Utility as a
Using Expected Utilities in Decision Making
E = 0.162 Tack Lands Utility

Point Up (0.45)
0.30
Tack Lands
Point Down (0.55)
0.05
Don’t Play
0.15
3-18
IE 12 2
OPERATIONS RESEARCH 2
M od u l e 2 – Pa r t 1
Deterministic Dynamic Programming (DP)
Contents
I. Nature of Computations
II. Resource Allocation Problem
III. Workforce Size Problem
IV. Cargo Loading/ Knapsack Problem
Puzzle
Given the two containers above, fill the 8- liter container with exactly 7 liters of
water.
5-liter container 8-liter container

Why Dynamic Programming?
It is a practical method for finding solutions to extremely

complicated problems.
It has the ability to attack difficult problems that other

optimization tools fail to solve.
Brief History
The term dynamic programming was invented by Richard Ernest
Bellman in 1953 while he was working at RAND Corp. under the
Secretary of Defense
Bellman was interested in planning, decision making and thinking

but decided to use the word programming
He used the term dynamic since the nature of his work was
dynamic, multi- stage, and time-varying.
Richard Ernest Bellman (1920-1984)
• received his PhD at Princeton University at the age of 25
• associate professor of mathematics at Stanford University
• published 619 papers and 39 books
• the Father of Dynamic Programming
General Description
• an optimization procedure that is particularly applicable to
problems requiring a sequence of interrelated decisions
• designed to enhance the computational efficiency of solving

particular mathematical programs
Characteristics of DP Applications
1. The problem can be divided into stages with a decision required
at each stage.
o In many DP problems, the stage is the amount of time that
has elapsed since the beginning of the problem.
o In some situations, decisions are not required at every stage.
2. Each stage has a number of states associated with it. State is the
information needed to any stage to make an optimal decision.
3. The decision chosen at any stage describes how the state at the
current stage is transformed into the state at the next stage.
4. Principle of optimality
Given the current state, the optimal decision for each of the
remaining stages must not depend on previously reached states or
previously chosen decisions.
5. If the states for the problem have been classified into one of T
stages, there must be a recursion that relates the cost or reward
earned during stages t, t+1, t+2, …, T to the cost or reward earned
from stages t+1, t+2, …, T.
General Process
1. Divide the bigger or complex problems into a number of smaller
problems (or subproblems).
2. Solve recursively each subproblem optimally.
3. Use these optimal solutions to construct an optimal solution for
the original problem.
Applications
• Capital budgeting
• Resource allocation
• Cargo loading / Knapsack problem
• Traveling salesman
• Reliability
• Production scheduling
• Inventory control
• Etc.
Resource Allocation
Example 1: Budget Co. company is tasked to optimize the
maximum return from allocating a total 8 million pesos budget to
one or more projects among 1, 2, 3 and 4. The table below shows
the return per budget allocated in each project.
𝒙𝒊 𝒓𝟏 (𝒙𝟏 ) 𝒓𝟐 (𝒙𝟐 ) 𝒓𝟑 (𝒙𝟑 ) 𝒓𝟒 (𝒙𝟒 )
0 0 0 0 0
1 3 1 2 1
2 7 2 4 3
3 10 4 6 6
4 12 8 8 9
5 13 13 13 12
6 14 17 17 14
7 14 19 19 16
8 14 20 20 17
Workforce Size Model
Example 1: A construction contractor estimates that the size of the
work force needed over the next 5 weeks to be 5, 7, 8, 4, and 6
workers, respectively. Excess labor kept on the force will cost
P3,000 per worker per week, and new hiring in any week will incur
a fixed cost of P4,000 plus P2,000 per worker per week.
HW: Resource Allocation
The owner of a chain of four grocery stores has purchased six
crates of fresh strawberries. The table below gives the estimated
total expected profit (in ten thousands of pesos) at each store
when it is allocated various number of crates. For administrative
reasons, the owner does not wish to split crates between stores.
However, he is willing to distribute zero crates to any of his stores.
Find the allocation of six crates to four stores as to maximize the
expected profit.
HW: Resource Allocation
The table below gives the estimated total expected profit (in ten
thousands of pesos) at each store when it is allocated various
number of crates.
HW: Workforce Size Model
Luxor Travel arranges 1-week tours to southern Egypt. The agency
is contracted to provide tourist groups with seven, four, seven, and
eight rental cars over the next four weeks, respectively. Luxor
Travel subcontracts with a local car dealer to supply rental needs.
The dealer charges a rental fee of $320 per car per week, plus a
flat rate of $400 for any rental transaction. Luxor, however, may
elect not to return the rental cars at the end of the week, in which
case the agency will be responsible only for the weekly rental
($320). What is the best way for Luxor Travel to handle the rental
situation?
Cargo Loading/Knapsack Problem
Consider the following data:
Variable Item, j Value, 𝒗𝒋 Weight, 𝒘𝒋

𝒙𝟏 1 65 2
𝒙𝟐 2 80 3
𝒙𝟑 3 30 1
Max Weight W= 5
Markov Analysis
To accompany
Quantitative Analysis for Management, Tenth Edition,
Power Point slides created by Jeff Heyl © 2009 Prentice-Hall, Inc.
Learning Objectives
After completing this chapter, students should be able to:
1. Determine future states or conditions by

using Markov analysis
2. Compute long-term or steady-state
conditions by using only the matrix of
transition probabilities
3. Understand the use of absorbing state
analysis in predicting future conditions
Chapter Outline
1. Introduction
2. States and State Probabilities
3. Matrix of Transition Probabilities
4. Predicting Future Market Share
5. Markov Analysis of Machine Operations
6. Equilibrium Conditions
7. Absorbing States and the Fundamental
Matrix: Accounts Receivable Application
Introduction
n Markov analysis is a technique that deals with

the probabilities of future occurrences by
analyzing presently known probabilities
n It has numerous applications in business
n Markov analysis makes the assumption that
the system starts in an initial state or
condition
n The probabilities of changing from one state
to another are called a matrix of transition
probabilities
n Solving Markov problems requires basic
matrix manipulation
Introduction
n This discussion will be limited to Markov

problems that follow four assumptions
1. There are a limited or finite number of
possible states
2. The probability of changing states
remains the same over time
3. We can predict any future state from the
previous state and the matrix of transition
probabilities
4. The size and makeup of the system do
not change during the analysis
States and State Probabilities
n States are used to identify all possible conditions
of a process or system
n It is possible to identify specific states for many
processes or systems
n In Markov analysis we assume that the states are
both collectively exhaustive and mutually
exclusive
n After the states have been identified, the next
step is to determine the probability that the
system is in this state
n The information is placed into a vector of state
probabilities
p (i) = vector of state probabilities
for period i
= (p1, p2, p3, … , pn)
where
n = number of states
p1, p2, … , pn = probability of being in state 1,
state 2, …, state n
n In some cases it is possible to know with
complete certainty what state an item is in
n Vector states can then be represented as
p (1) = (1, 0)
where
p (1) = vector of states for the machine
in period 1
p1 = 1 = probability of being in the
first state
p2 = 0 = probability of being in the
second state
The Vector of State Probabilities for
Three Grocery Stores Example
n States for people in a small town with three
grocery stores
n A total of 100,000 people shop at the three
groceries during any given month
n Forty thousand may be shopping at American
Food Store – state 1
n Thirty thousand may be shopping at Food Mart –
state 2
n Thirty thousand may be shopping at Atlas Foods
– state 3
n Probabilities are as follows
State 1 – American Food Store: 40,000/100,000 = 0.40 = 40%
State 2 – Food Mart: 30,000/100,000 = 0.30 = 30%
State 3 – Atlas Foods: 30,000/100,000 = 0.30 = 30%
n These probabilities can be placed in the following

vector of state probabilities
p (0) = (0.4, 0.3, 0.3)
where p (0) = vector of state probabilities for the three
grocery stores for initial period
p1 = 0.4 = probability that person will shop at
American Food, state 1
Food Mart, state 2
Atlas Foods, state 3
n The probabilities of the vector states represent
the market shares for the three groceries
n Management will be interested in how their
market share changes over time
n The tree diagram on the next slide shows the
market shares in the next month
n Tree diagram for three grocery stores example
0.8 #1 0.32 = 0.4(0.8)

American Food #1 0.1
#2 0.04 = 0.4(0.1)
0.4 0.1
#3 0.04 = 0.4(0.1)
0.1 #1 0.03
Food Mart #2 0.7
#2 0.21
0.3 0.2
#3 0.06
0.2 #1 0.06
Atlas Foods #3 0.2
#2 0.06
0.3 0.6
#3 0.18
Matrix of Transition Probabilities
n The matrix of transition probabilities allows us to

get from a current state to a future state
Let Pij = conditional probability of being in state j

in the future given the current state of i
n For example, P12 is the probability of being in

state 2 in the future given the event was in state 1
in the period before
Matrix of Transition Probabilities
n Let P = the matrix of transition probabilities
P11 P12 P13 … P1n

P21 P22 P23 … P2n
P= …
…
Pm1 … Pmn
n Individual Pij values are determined empirically

n The probabilities in each row will sum to 1
Transition Probabilities for the
Three Grocery Stores
n We used historical data to develop the following
matrix
0.8 0.1 0.1
P = 0.1 0.7 0.2
0.2 0.2 0.6
Row 1
0.8 = P11 = probability of being in state 1 after being in state 1 in the
preceding period
preceding period
preceding period
Transition Probabilities for the
Three Grocery Stores
0.8 0.1 0.1
P = 0.1 0.7 0.2
0.2 0.2 0.6
Row 2
preceding period
preceding period
preceding period
Row 3
preceding period
preceding period
preceding period
Predicting Future Market Shares
n One of the purposes of Markov analysis is to

predict the future
n Given the vector of state probabilities and the
matrix of transitional probabilities, it is not very
difficult to determine the state probabilities at a
future date
n This type of analysis allows the computation of
the probability that a person will be at one of the
grocery stores in the future.
n Since this probability is equal to market share, it
is possible to determine the future market shares
of the grocery stores
n When the current period is 0, the state
probabilities for the next period 1 are determined
as follows
p (1) = p (0)·P
n For any period n we can compute the state

probabilities for period n + 1
p (n + 1) = p (n) ·P
n The computations for the next period’s market

share are
p (1) = p (0) ·P
0.8 0.1 0.1

= (0.4, 0.3, 0.3) 0.1 0.7 0.2
0.2 0.2 0.6
= [(0.4)(0.8) + (0.3)(0.1) + (0.3)(0.2),
(0.4)(0.1) + (0.3)(0.7) + (0.3)(0.2),
(0.4)(0.1) + (0.3)(0.2) + (0.3)(0.6)]
= (0.41, 0.31, 0.28)
p (0) = (0.4, 0.3, 0.3)
p (1) = (0.41, 0.31, 0.28)
n The market share for American Food and Food

Mart have increased and the market share for
Atlas Foods has decreased
n We can determine if this will continue by looking
at the state probabilities will be in the future
n For two time periods from now
p (2) = p (1)·P
n Since we know that
p (1) = p (0)P
n We have
p (2) = p (1)P = [p (0)P]P = p (0)PP = p (0)P2
n In general
p (n) = p (0)Pn
n The question of whether American and Food Mart
will continue to gain market share and Atlas will
continue to loose is best addressed in terms of
equilibrium or steady state conditions
Markov Analysis of
Machine Operations
n The owner of Tolsky Works has recorded the
operation of his milling machine for several years
n Over the past two years, 80% of the time the
milling machine functioned correctly for the
current month if it had functioned correctly
during the preceding month
n 90% of the time the machine remained incorrectly
adjusted if it had been incorrectly adjusted in the
preceding month
n 10% of the time the machine corrected to
problems and operated correctly when it had
been operating incorrectly
Markov Analysis of
Machine Operations
n The matrix of transition probabilities for this
machine is
0.8 0.2
P=
0.1 0.9
where
P11 = 0.8 = probability that the machine will be correctly functioning
this month given it was correctly functioning last month
P12 = 0.2 = probability that the machine will not be correctly
functioning this month given it was correctly functioning
last month
P21 = 0.1 = probability that the machine will be correctly functioning
this month given it was not correctly functioning last
month
P22 = 0.9 = probability that the machine will not be correctly
functioning this month given it was not correctly
functioning last month
Markov Analysis of
Machine Operations
n What is the probability that the machine will be
functioning correctly one and two months from
now?
p (1) = p (0)P
0.8 0.2
= (1, 0)
0.1 0.9
= [(1)(0.8) + (0)(0.1), (1)(0.2) + (0)(0.9)]
= (0.8, 0.2)
Markov Analysis of
Machine Operations
n What is the probability that the machine will be
functioning correctly one and two months from
now?
p (2) = p (1)P
0.8 0.2
= (0.8, 0.2)
0.1 0.9
= [(0.8)(0.8) + (0.2)(0.1), (0.8)(0.2) + (0.2)(0.9)]
= (0.66, 0.34)
Equilibrium Conditions
n It is easy to imagine that all market shares will
eventually be 0 or 1
n But equilibrium share of the market values or
probabilities generally exist
n An equilibrium condition exists if state
probabilities do not change after a large number
of periods
n At equilibrium, state probabilities for the next
period equal the state probabilities for current
period
n Equilibrium state probabilities can be computed
by repeating Markov analysis for a large number
of periods
n It is always true that
p (next period) = p (this period)P
n Or
p (n + 1) = p (n)P
n At equilibrium
p (n + 1) = p (n)
n So at equilibrium
p (n + 1) = p (n)P = p (n)
n Or
p = pP
n For Tolsky’s machine
p=pP
0.8 0.2
(p1, p2) = (p1, p2)
0.1 0.9
n Using matrix multiplication
(p1, p2) = [(p1)(0.8) + (p2)(0.1), (p1)(0.2) + (p2)(0.9)]

n The first and second terms on the left side, p1 and
p2, are equal to the first terms on the right side
p1 = 0.8p1 + 0.1p2
p2 = 0.2p1 + 0.9p2
n The state probabilities sum to 1

p1 + p2 + … + pn = 1
n For Tolsky’s machine

p1 + p2 = 1
n We arbitrarily decide to solve the following two
equations
p2 = 0.2p1 + 0.9p2
p1 + p2 = 1
n Through rearrangement and substitution we get
0.1p2 = 0.2p1
p 2 = 2p1
p1 + p2 = 1
p1 + 2p1 = 1
3p1 = 1
p1 = 1/3 = 0.33333333
p2 = 2/3 = 0.66666667
Exercises A
1. During the day, the traffic on North Monroe Street
in Quincy is fairly steady, but the traffic conditions
can vary considerably from one hour to the next
due to slow drivers and traffic accidents. As one
driver said, "The traffic conditions on Monroe can
be either fair, tolerable, or miserable." If the traffic
conditions are fair in one hour, there is a 20%
chance that they will be tolerable in the next hour
and a 10% chance that they will be miserable. If the
traffic conditions are tolerable, there is a 20%
chance that they will be fair in the next hour and a
5% chance that they will be miserable. If addition, if
the traffic conditions are miserable, there is a 60%
chance that they will remain that way and a 30%
chance that they will be fair in the next hour. If the
traffic conditions are miserable at this time, what is
the probability that they will be fair in two hours?
What is the probability that they will be tolerable in
two hours?
Exercises A
2. Greg Cracker, mayor of Quincy, is alarmed about the
traffic conditions on Monroe Street (see Problem #1). In
the long run, what percent of the time will traffic conditions
be fair, tolerable, and miserable on Monroe Street?
3. The tiger minnow, which can be found in Lake Jackson
and in Lake Bradford, is a small meat-eating fish. At the
present time, there are 900 tiger minnows in Lake Jackson
and 100 tiger minnows in Lake Bradford, but a new 10-
foot-wide canal between these two lakes will soon change
these numbers. Because tiger minnows eat other fish and
themselves, the total population remains about the same.
Bob Brite, an Eagle Scout from Troop B, has done nothing
but watch the tiger minnows going through the canal.
During the past month Bob has observed 90 tiger minnows
go from Lake Jackson to Lake Bradford, and he has
observed 5 tiger minnows go from Lake Bradford to Lake
Jackson. Assuming that these migration patterns will
remain the same, how many tiger minnows will be in each
lake in the long run?
Exercises A
4. The residents of Lake Bradford are angry about

the canal between Lake Bradford and Lake
Jackson. This canal has allowed too many tiger
minnows into Lake Bradford, and, as a result,
the value of the lake property on Lake Bradford
has gone down considerable. One solution
would be to place a one-way dam in the canal.
This would only reduce the fraction of the tiger
minnows migrating from Lake Jackson to Lake
Bradford. (See Problem 3 for details.) In other
words, the dam would have the effect of
reducing the probability of a tiger minnow
migration from Lake Jackson to Lake Bradford.
What would this probability have to be to
restore the original number of tiger minnows in
each lake?
Exercises A
5. John Jones of Bayside Laundry has been
providing cleaning and linen service for rental
condominiums on the Gulf coast for over 10
years. Currently, John is servicing 26
condominium developments. John’s two
major competitors are Cleanco, which
currently services 15 condominium
developments, and Beach Services, which
performs laundry and cleaning services for
11 condominium developments.
Exercises A
Recently, John contacted Bay Bank about a
loan to expand his business operations. To
justify the loan, John has kept detailed records of
his customers and the customers that he
received from his two competitors. From the past
year, he was able to keep 18 of his original 26
customers. During the same period, he was able
to get 1 new customer from Cleanco and 2 new
customers from Beach Services. Unfortunately,
John lost 6 of his original customers to Cleanco
and 2 of his original customers to Beach Services
during the same year. John has also learned that
Cleanco has kept 80% of its current customers.
He also knows that Beach Services will keep at
least 50% of its customers
Exercises A
For John to get the loan from Bay Bank,

he needs to show the loan officer that he
will maintain an adequate share of the
market. The officers of Bay Bank are
concerned about the recent trends for
market share, and they have decided not to
give John a loan unless he will keep at least
35% of the market share in the long run.
What types of equilibrium market shares
can John expect? If you were an officer of
Bay Bank, would you give John a loan?
Why or why not?
Absorbing States and the
Fundamental Matrix
n Accounts Receivable example
n The examples so far assume it is possible to go
from one state to another
n This is not always possible
n If you must remain in a state it is called an
absorbing state or transient state
n If you are certain to return to a particular state it
is called a recurrent state
n An accounts receivable system normally places
accounts in three possible states
State 1 (p1): paid, all bills
State 2 (p2): bad debt, overdue more than three months
State 3 (p3): overdue less than one month
State 4 (p4): overdue between one and three months
Fundamental Matrix
n The matrix of transition probabilities of this
problem is
NEXT MONTH
BAD <1 1 TO 3
THIS MONTH PAID DEBT MONTH MONTHS
Paid 1 0 0 0
Bad debt 0 1 0 0
Less than 1 month 0.6 0 0.2 0.2
1 to 3 months 0.4 0.1 0.3 0.2
n Thus
1 0 0 0
0 1 0 0
P=
0.6 0 0.2 0.2
0.4 0.1 0.3 0.2
Fundamental Matrix
n To obtain the fundamental matrix, it is necessary
to partition the matrix of transition probabilities
as follows
I 0
1 0 0 0 1 0 0 0
I= 0=
0 1 0 0 0 1 0 0
P=
0.6 0 0.2 0.2 0.6 0 0.2 0.2
0.4 0.1 0.3 0.2 A= B=
0.4 0.1 0.3 0.2
A B
where
I = an identity matrix
0 = a matrix with all 0s
Fundamental Matrix
n The fundamental matrix can be computed as
F = (I – B)–1
–1
1 0 – 0.2 0.2
F=
0 1 0.3 0.2
–1
0.8 –0.2
F=
–0.3 0.8
–1 d –b
The inverse a b a b r r
is c =
of the matrix c d d –c a
r r
where
r = ad – bc
Fundamental Matrix
n To find the matrix F we compute
r = ad – bc = (0.8)(0.8) – (–0.3)(–0.2) = 0.64 – 0.06 = 0.58
n With this we have
0.8 –(–0.2)
–1
0.8 –0.2 0.58 0.58 1.38 0.34
F= = =
–0.3 0.8 –(–0.3) 0.8 0.52 1.38
0.58 0.58
• The values in the fundamental matrix represent the average number of times
the process visits each of the nonabsorbing states before ultimately
reaching some absorbing state.
• For example,the value of 1.38 in the first column and first row of the
fundamental matrix tells us that if the process starts in state 3 (account is
overdue less than one month), it will return to state 3 an average of 1.38 times.
• Similarly, if the process starts in state 3, it will visit state 4 (account is overdue
between one and three months) an average of 0.34 times.
• If the process starts in state 4, it will visit state 3 an average of .52 times and
revisit state 4 an average of 1.38 times.
Limiting Matrix
n Limiting matrix L gives the long-run probabilities

of reaching each absorbing state from each
transient state by multiplying the fundamental
matrix F by the submatrix A: L = FA
or
n Limiting transition matrix L gives the eventual
likelihood that the process will move from each
transient state to each absorbing state.
n When matrix L is multiplied by the initial state

vector, we get the probabilities that the Markov
process will eventually reach each absorbing
state.
Fundamental Matrix
n We can use the limiting matrix L = FA to answer
questions such as how much of the debt in the
less than one month category will be paid back
and how much will become bad debt
M = (M1, M2, M3, … , Mn)
where
n = number of nonabsorbing states
M1 = amount in the first state or category
M2 = amount in the second state or category
Mn = amount in the nth state or category
Fundamental Matrix
n If we assume there is $2,000 in the less than one
month category and $5,000 in the one to three
month category, M would be
M = (2,000, 5,000)
Amount paid and

= MFA
amount in bad debts 0.97 0.03
= (2,000, 5,000) 0.86
0.14
= (6,240, 760)
n Out of the total of $7,000, $6,240 will eventually be

paid and $760 will end up as bad debt
Exercise B
n Professor Orlina gives 2-month computer

programming courses during the summer term.
Students must pass a number of exams to pass
the course, and each student is given 3 chances
to take the exams. The following states describe
the possible situations that could occur:
n State 1. Pass all of the exams and pass the course
n State 2. Do not pass all of the exams by the third
attempt and flunk the course
n State 3. Fail an exam in the first attempt
n State 4. Fail an exam in the second attempt
Exercise B
n After observing several classes, Professor Orlina was
able to obtain the following matrix of transition
probabilities:
S1 S2 S3 S4
S3é0.6 0 0.1 0.3ù
S1êê 1 0 0 0ú
ú
S4ê0.3 0.3 0.2 0.2ú
S2êë 0 1 0 0 úû
n At the present time there are 50 students who did not
pass all exams on the first attempt, and there are 30
students who did not pass all remaining exams on the
second attempt. How many students in these two
groups will pass the course, and how many will fail the
course?
Exercise C
n A study of facial tissue consumption revealed that a
Markov process can be used to model tissue brand
selection.
n The four principal companies manufacturing facial tissue
in the United States are Kimberly-Clark (KC), Scott (a
subsidiary of Kimberly-Clark) (S), Proctor & Gamble (PG),
and James River (JR).
n The transition matrix describing tissue selection is
believed to be:
Exercise C
n a. Chris Willis is currently purchasing Scott tissues. What is the
probability that Chris will buy Scott tissues each of her next five
purchases?
n b. A survey of 100 consumers living in a particular apartment building
asked respondents which brand of tissue they last purchased. The
data were was as follows:
Brand Number of Consumers
Kimberly-Clark 20
Scott 26
Proctor & Gamble 23
James River 27
Other 4
n i. Among this group, what is the expected number who will purchase
each of the four major brands for their next tissue purchase?
n ii. Among this group, what is the expected number who will purchase
each of the four major brands for the purchase after that?
n c. On the basis on the above transition matrix, determine the market

share of the four major brands.
Sources
n http://wps.prenhall.com/bp_render_qam_10
n Hillier, F.S. and G.J. Lieberman. 2001. Introduction
to Operations Research. New York: McGraw-Hill
Co. Inc.
IE 12 2
OPERATIONS RESEARCH 2
M od u l e 2 – Pa r t 3
Markov Processes in Economic Analyses
Fast-food restaurant selection
• The town of Sandpoint, Idaho has three fast-food restaurants: RallyBurger, Burger Barn,
and Caesar’s.
• We have observed that, among Sandpoint’s population, the choice of a fast-food
restaurant is influenced solely by the last fast-food restaurant visited. In particular, the
probability that a customer who has last eaten at RallyBurger returns to RallyBurger on his
next visit to a fast-food restaurant is .70; the probability that the customer will next choose
Burger Barn is .20, and the probability the customer will next choose Caesar’s is .10.
• A customer who has last eaten fast food at Burger Barn has a .35 probability that his next
fast-food restaurant visit will be to RallyBurger, a .50 probability that he will return to Burger
Barn, and a .15 probability that he will go to Caesar’s next time.
• Finally, if a customer last ate fast food at Caesar’s, there is a .25 probability that he will go
to RallyBurger, a .30 probability that he will choose Burger Barn, and a .45 probability that
he will return to Caesar’s on his next visit to a fast-food restaurant.
• Caesar’s management would like to estimate its overall market share of the fastfood
business in Sandpoint. To aid in this analysis, it wishes to construct a transition matrix that
describes how customers select fast-food restaurants.
• The steady state probabilities are
• 𝝅𝟏 = 0.51111 for Rally Burger
• 𝝅𝟐 = 0.31111 for Burger Barn
• 𝝅𝟑 = 0.17778 for Ceasar’s
Mean Recurrence Time for Fast-Food restaurant
• The average time required for the process to return to a given state
• It is the inverse of the steady state value: 1/ 𝝅𝒊
• For the fast-food restaurant example, the steady state value corresponding to Rally Burger
is .51111. Thus, a customer who is currently eating at Rally Burger will return to Rally Burger
on the average of every 1/.5111 = 1.9565 visits to a fast- food restaurant.
• Similarly, a customer will return to Burger Barn every 1/.3111 = 3.2143 visits to a fast-food
restaurant and to Caesar’s every 1/.17778 = 5.6250 visits.
Rolley’s Rentals
• Rolley’s Rentals, located in Lahaina, Hawaii, rents bicycles for daily use. The company
estimates that its profit is a function of the weather and has classified weather conditions
into three states: sunny, cloudy, and rainy.
• Based on meteorological studies for Lahaina, Rolley’s estimates that if today is sunny, there
is a 75% chance of sunny weather tomorrow, a 20% chance of cloudy weather tomorrow,
and a 5% chance of rainy weather tomorrow.
• If today is cloudy, the likelihood of sunny weather tomorrow is 45%, of cloudy weather 40%,
and of rainy weather 15%.
• If today is rainy, there is a 35% chance of sunny weather tomorrow, a 45% chance of
cloudy weather, and a 20% chance of rainy weather.
Rolley’s Rentals
• Mr. Rolley would like to determine a transition matrix to assist in estimating his expected
daily profit.
Rolley’s Rentals
• Rolley’s Rentals has estimated that it earns an average profit of $120 on sunny days and
$40 on cloudy days, but it suffers an average loss of $200 on rainy days.
• Mr. Rolley would like to determine his expected profit over the upcoming week if the
weather in Lahaina today is sunny.
• Solution:
• If the weather today is sunny, the probability that tomorrow’s weather will be sunny,
cloudy, or rainy is .75, .20, and .05, respectively.
• Since Rolley’s earns an expected profit of $120 on a sunny day and $40 on a cloudy day
and loses $200 on a rainy day, its expected profit tomorrow is: 0.75($120) + 0.20($40) +
0.05(-$200) = $88
Rolley’s Rentals
• For day 2, 𝜋(2) = (.75, .20, .05)xP = (.6700, .2525, .0775).
• Thus, the expected profit for day 2 is: .6700($120) + .2525($40) + .0775(-$200) = $75
• For day 3, 𝜋(3) = (.6700 .2525 .0775)xP = (.6432, .2699, .0869)

• and the expected profit is: .6432($120) + .2699($40) + .0869(-$200) = $70.60
Rolley’s Rentals
• Therefore, Rolley’s expected profit over the upcoming seven-day period is:
• $88.00 + $75.00 + $70.60 + $69.14 + $68.64 + $68.49 + $68.43 = $508.30
Rolley’s Rentals
• Mr. Rolley is considering selling his business. A business consultant has told him that
businesses such as his are usually valued at six times their expected yearly profit.
• Mr. Rolley would like to determine a fair asking price for the business, should he decide
to sell it.
• Solution:
• The following steady state probability values for weather in Lahaina are obtained:
• Sunny: 0.6298
• Cloudy: 0.2786
• Rainy: 0.0916
• Therefore, Rolley’s expected daily profit is: .6298 ($120) + .2786 ($40) + .0916 (-$200) =
$68.40
Rolley’s Rentals
• Since Rolley’s expected yearly profit should be 365($68.40) = $24,966, the business is
worth 6($24966) = $149,796.
• Because businesses generally sell for between 5% and 10% below their asking price, Mr.
Rolley has decided to ask $160,000 for the business.
GAME THEORY
Mr. Casiano DC. Jaurigue

Outline
l Chapter 1 Introduction
l Brief History
l Game Theory Defined
l Importance and Applications of Game
Theory
l Methodology
l Classification of Game Models
l Elements of a Game
2
Outline
l Chapter 2 Two-Person, Zero-Sum Game
l Conservative Strategies
l Minimax Criterion
l Nash Equilibrium
l Dominance
l Pure Strategy Game
l Mixed Strategy Game
l Analytical Method
l Graphical Method
l Linear Programming Solution
3
Outline
l Chapter 3 Two-Person, Non-constant-

Sum Game
l Chapter 4 N-Person Game
l Chapter 5 Applications
4
INTRODUCTION
Brief History of Game Theory
l 1913 - E. Zermelo provided the first theorem of

game theory; asserts that chess is strictly
determined
l 1928 - John von Neumann proved the minimax
theorem
l 1944 - John von Neumann & Oskar Morgenstern
wrote "Theory of Games and Economic
Behavior”
l 1950-1953 - John Nash describes Nash
equilibrium
l 1972 - John Maynard Smith wrote “Game Theory
and The Evolution of Fighting”
Source: Xiao, Thomas & Westwell, 6
Evolutionary Game Theory
Brief History of Game Theory
l In 1994, John Harsanyi, John Nash, and Reinhard

Selten jointly received the Nobel Prize in
Economics from the Royal Swedish Academy of
Sciences for developing the notion of
noncooperative game theory.
l John Nash developed the concepts of the Nash

equilibrium and the Nash bargaining problem
which are the cornerstones of modern game
theory.
7
What is Game Theory?
Game theory deals with decision making

under conflict or competition where 2 or
more decision makers are involved (with
conflicting objectives) and the
consequences (payoff) to each depends
on the course of action taken by all, and
each party is usually trying to maximize
its overall welfare at the expense of the
others.
8
Why Study Game Theory?
l It stimulates us to think about conflicts in

a novel way.
l Helps us to analyze situations more
rationally and formulate an acceptable
alternative with respect to circumstance.
l It is applicable to almost any type of
conflict (military, economic, political,
social, etc.)
9
Because the press tells us to …
“ Managers have much to learn from game theory —

provided they use it to clarify their thinking, not as a
substitute for business experience.”
The Economist, 15 June 1996
“ Game Theory, long an intellectual pastime, came into

its own as a business tool.”
Forbes, 3 July 1995
“ Game theory is hot.”

The Wall Street Journal, 13 February 1995
Source: Mike Shor. Game Theory & Business Strategy 10 10

Because recruiters tell us to …
l “Game theory forces you to see a business situation

over many periods from two perspectives: yours and
your competitor’s.”
l Judy Lewent – CFO, Merck
l “Game theory can explain why oligopolies tend to be

unprofitable, the cycle of over capacity and
overbuilding, and the tendency to execute real
options earlier than optimal.”
l Tom Copeland – Director of Corporate Finance, McKinsey
Source: Mike Shor. Game Theory & Business Strategy 11 11

Where is game theory currently used?
– Ecology
– Networks
– Economics
Applications
l Marketing strategies
l Labor-management negotiations
l Potential mergers
l International military conflicts
l Competitive bidding
l Advertising
l Etc.
13
Limitations & Problems
l Assumes players always maximize their

outcomes
l Some outcomes are difficult to provide a
utility for
l Not all of the payoffs can be quantified
l Not applicable to all problems
Methodology
l Present the managerial situation,

problem, or conflict in game format.
l View the decision makers as players.
l Determine the optimal playing strategies
for the participants.
15
Strategy
l a comprehensive plan of action
l a decision rule or set of instructions

about which actions a player should take
following all possible histories of play
16
Classification of Game Models
l by the number of players

l by the sum of all payoffs
l by the number of strategies employed
17
Elements of a Game
l Players
l Set of possible actions
l Information available
l Payoff consequences (in matrix form)
l Players’ preference over payoffs
18
Payoff Matrix
Brilliant Lights
Use radio Use newspaper
Use radio 3 5
Bright Lights
Use
newspaper 1 -2
l Payoffs are shown only for the row player (Bright Lights).
Brilliant Light’ payoffs are just the negative of each number.
l A player’s payoff is the amount that the player wins or loses in

a particular situation in a game.
19
Illustration 1
Consider two lighting fixture stores, namely,

Bright Lights and Brilliant Lights. The
respective market shares have been stable up
until now, but the situation may change. The
owner of Bright Lights has developed two
distinct advertising strategies – using radio
spots and newspaper ads. Upon hearing this,
the owner of Brilliant Lights also proceeds to
prepare radio and newspaper ads. The next
slide shows what will happen to current market
shares if both stores begin advertising.
20
Bright Lights’ Payoff Table
Brilliant Lights
Use radio 3 5
Bright Lights
Use
newspaper 1 -2
l Positive entries in the matrix represent wins for the

row player (Bright Lights) but losses for the column
player (Brilliant Lights).
l Negative entries represent losses for the row player
(Bright Lights) but wins for the column player (Brilliant
Lights).
21
Bright Lights’ Payoff Matrix
Brilliant Lights
Use radio 3 5
Bright Lights
Use
newspaper 1 -2
Assumption:
l Row player tries to maximize winnings
while column player tries to minimize
losses.
22
Two-Person, Zero-Sum Game
l There are only two players.

l One player wins a certain amount and
the other player loses the same amount
(i.e., the sum of the gains and losses for
both players is zero).
l Assumption:
l Each player moves to maximize his own
payoff without knowing in advance how the
opponent will move.
23
MINIMAX CRITERION
Conservative Strategies
l Playing safe by adopting the conservative strategy
of assuming the worst:
l The row player chooses the move where
smallest payoff is as large as possible (to guard
himself when his opponent manage to select
the column that will result in the smallest
payoff).
l The column player (wanting to maximize his
payoff) minimizes his opponent’s payoff by
choosing the column where maximum entry is
as small as possible.
25
Minimax Criterion
l an approach to selecting strategies that
will minimize losses for each player
Bad news: Knowing game theory does not
guarantee winning
Good news: There is a framework for thinking

about strategic interaction
l minimization of one’s maximum losses

(or maximization of one’s minimum
gains)
26
Row Player’s Maximin Strategy
l Choose the row whose minimum entry is

the largest of the minimum row entries
Column Player’s Minimax Strategy

l Choose the column whose maximum
entry is the smallest of the maximum
column entries
27
Illustration 2
Consider B Row min
é -1 - 3 5 ù - 3
A ê- 2.5 1 3 ú -2.5
ê ú
ê 4
ë 0 - 2ú -2 û maximin
Column max 4 1 5
minimax
l Player A’s conservative (maximin) strategy is to choose

strategy 3.
l Player B’s conservative (minimax) strategy is to choose
strategy 2.
28
Illustration 2
Consider B Row min
é -1 - 3 5 ù - 3
A ê- 2.5 1 3 ú -2.5
ê ú
ê 4
ë 0 - 2ú -2 ûmaximin
Column max 4 1 5
minimax
l If player A knows that player B is playing conservatively,

will he change his strategy? Why or why not?
l If player B knows that player A is playing conservatively,
29
Illustration 3
Consider B Row min

é8 4ù 4 maximin
A êê- 4 0ú -4
ú
êë 5 3úû 3
Column max 8 4 minimax
l If player A knows that player B is playing conservatively,

l If player B knows that player A is playing conservatively,
30
Remark
l There are games where the knowledge

that one’s opponent is playing
conservatively cannot be used to
advantage.
31
PURE STRATEGY GAMES
Pure Strategy
(or Strictly Determined) Games
l neither player can benefit from the
knowledge that the opponent is using the
conservative strategy
l the strategy of each player does not
change regardless of the other player’s
strategy.
l if and only if its payoff matrix has a
saddle point (or minimax point)
33
Saddle Point
l a situation where both players are facing

pure strategies or using conservative
strategies
l an entry in the payoff matrix that is both
the smallest entry in its row and the
largest entry in its column
l also considered as the value of the game
34
Value of the Game
l the numerical value of the saddle point
l average or expected game outcome if

the game is played an infinite number of
times
35
Remark
l Strategies for saddle point games can be

determined without performing any
calculations.
l The game has a saddle point if

maximin = minimax.
36
Illustration 4
Express Airlines and Air Speed are two airlines

that are competing on two particular routes.
They each have two strategies for improving
their relative market position: pricing (cheaper
tickets) and service (more flight attendants,
better food, more comfortable seats). Market
research indicates when these strategies are
employed, the results will be as indicated on
the next slide.
37
Air Speed
pricing service
pricing 25 18
Express Airlines
service -15 15
Entries are number of passengers (in hundreds).
l Check if the game has a saddle point.

l Determine the optimal strategy for each airline.
l Who wins the game?
38
Solution
l Game’s saddle point: 18

l Optimum strategy for
l Express Air: cut fares
l Air Speed: offer improved service
l If both airlines follow this policy, Express

Air will gain 18 units (1,800 passengers)
from Air Speed.
39
Exercise
Find the saddle point/s, the conservative

strategies, and the value of each of the
following strictly determined games:
a) b) é3 54 3ù
é 8 0 - 2ù
ê- 6 4 1 ú ê2 40 - 2ú
ê ú ê ú
êë 4 3 2 úû ê1 - 6 5 0 ú
ê0 ú
ë 2 6 -1 û
40
Illustration 5
Two video rental chains, VCity and Vvid, are thinking
about opening a new store in one of three cities. If
chain VCity locates its new store in city 1, its estimate
earning per year will be PhP 225,000 more than chain
Vvid if Vvid also locates in city 1 and PhP 90,000 more
if Vvid locates in city 3, but it will earn PhP 450,000
less than Vvid if Vvid locates in city 2. If chain Vcity
locates in city 2, it will earn PhP 900,000, PhP
675,000, and PhP 540,000 more than Vvid if Vvid
locates in cities 1, 2, and 3, respectively. Finally, if
chain VCity locates in city 3, it will earn PhP 675,000
and PhP 450,000 more than Vvid if Vvid locates in
cities 1 and 3, respectively, and it will earn PhP
225,000 less than Vvid if Vvid locates in city 2. Where
should each chain locate its new store?
41
Solution
VCity
city 1 city 2 city 3
city 1
Vvid city 2
city 3
42
Dominance
l a procedure used to reduce the size of

games to simpler form that is equivalent
to the original by eliminating strategies
that would never be played
l When is strategy eliminated?

l if all its game’s outcomes are the same or
worse than the corresponding game
outcomes of another strategy
43
Procedure for Finding the Reduced
Form of a Matrix Game
1. Compare the rows in the matrix. Delete any
row that is dominated by another row (i.e.,
each entry is less than the corresponding
entry of the other row).
2. Compare the columns in the matrix. Delete
any column that dominates another column
(i.e., each entry is greater than the
corresponding entry of the other column).
3. Repeat Steps 1 and 2 to eliminate all
redundant rows and columns.
44
Illustration
Consider the following matrix game:
C
é3 1 0 - 3ù
R
ê7 -1 1 6 ú
ê ú
êë2 4 5 1 úû
Does player R have a superior move?

Will player C prefer column 4 to column 1?
45
Reduced form of the game:
R
é-1 6ù
ê 4 1ú
ë û
46
Exercise
Find the reduced form of the ff. matrix games:
1. é 2 -1 4 ù 2. é 3 -2 6 2ù
ê-1 5 ê 4 -1 - 2 0 ú
1ú ê ú
ê ú
ê-1 0 6 - 2ú
ëê-1 3 - 4úû ê2 ú
ë 1 5 0 û
47
Exercise
é6 3 5 2ù
é2 - 3 - 2 -1ù ê1 7 4ú
3. 4.
ê1 ê 9 ú
ê 2 - 2 - 3úú
ê4 - 4 0 5ú
êë 0 -1 0 1 úû ê ú
ë3 - 6 4 4û
48
Exercise
5. é- 3 4 3 2 0ù
ê2 1 4 6 ú3ú
ê
ê2 1 3 5 1ú
ê1 ú
ê 6 - 4 3 1ú
êë 2 1 3 4 1úû
49
Nash Equilibrium (NE)
l Occurs when each player's strategy is optimal,

given the strategies of the other players.
l A player's best response (or best strategy) is

the strategy that maximizes that player's
payoff, given the strategies of other players.
l A Nash equilibrium is a situation in which

each player makes his or her best response.
Source: Harcourt Brace & Company 50

Nash Equilibrium (NE)
l If each player has chosen a strategy and no

player can benefit by changing his or her
strategy while the other players keep theirs
unchanged, then the current set of strategy
choices and the corresponding payoffs
constitute a Nash Equilibrium.
l Simply, players R and C are in Nash
equilibrium if R is making the best decision it
can, taking into account C's decision, and C is
making the best decision it can, taking into
account R's decision.
Source: Harcourt Brace & Company 51
NE History
l named after John Forbes Nash, Jr.
l Antoine Augustine Cournot developed the first concept

of the Nash equilibrium in pure strategies in his
theory of oligopoly (1838).
l Firms choose a quantity of output to maximize their
own profit but the best output for one firm depends
on the outputs of others.
l A Cournot equilibrium occurs when each firm's

output maximizes its profits given the output of the
other firms, which is a pure-strategy NE.
52
NE History
l Modern game-theoretic concept of NE is defined in
terms of mixed-strategies, where players choose a
probability distribution over possible actions.
l John von Neumann and Oskar Morgenstern
introduced the concept of the mixed strategy NE in
their 1944 book The Theory of Games and Economic
Behavior.
l Their analysis was restricted to the very special case of zero-
sum games. They showed that a mixed-strategy NE will exist
for any zero-sum game with a finite set of actions.
l John Forbes Nash in his 1951 article Non-Cooperative
Games defined a mixed strategy NE for any game
with a finite set of actions and prove that at least one
(mixed strategy) NE must exist. the
53
NE History
l John Forbes Nash, Jr. in his 1951 article Non-
Cooperative Games defined a mixed strategy NE for
any game with a finite set of actions and prove that at
least one (mixed strategy) NE must exist.
54
Basic Strategies
1. Plan ahead and look back

2. Use a dominating strategy if
possible
3. Eliminate any dominated strategies
4. Look for any equilibrium
5. Mix up the strategies
Plan ahead and look back
If you have a dominating strategy,
use it
Use strategy 1
Eliminate any dominated strategy
Eliminate strategy
2 as it’s dominated
by strategy 1
Look for any equilibrium
l Dominating Equilibrium
l Minimax Equilibrium
l Nash Equilibrium
The Golden Rule
COMMANDMENT
Never assume that your opponents’
behavior is fixed.
Predict their reaction to your behavior.
Decision Theory
vs. Game Theory
l Ten of you go to a restaurant
l If each of you pays for your own meal…

l This is a decision problem
l If you all agree to split the bill...

l Now, this is a game
Defining the Game
CAVEAT
Predict opponents’ reaction
to your behavior.
BUT
Be sure you understand who your
opponents are.
(Do you know everyone who may
react to your decisions?)
Summary
Defining the Game
l The strategic environment

l Who are the players? (Decision makers)
l What strategies are available? (Feasible actions)
l What are the payoffs? (Objectives)
l Rules of the game

l What is the time-frame for decisions?
l What is the nature of the conflict?
l What is the nature of interaction?
l What information is available?
MIXED STRATEGY GAMES
Recall:
l For pure strategy (or strictly determined

games), each player can play
conservatively without worrying that his
opponent might detect his strategy.
l For non-strictly determined games, if one
player plays conservatively every time,
his opponent may notice this and alter
his strategy to increase his payoff.
65
Marketing Example
Consider two competing companies who are to make a
decision regarding an investment in a new promotional
campaign.
Co. A’s course of actions:
a1: Advertise in all media.
a2: Advertise in newspaper only.
Co. B’s alternatives:
b1: Run a sweepstakes.
b2: Run a big sale.
66
Payoff Matrix
Company B
b1 b2
a1 4 -1
Company A
a2 -2 1
l Is this game a pure strategy game?

l What are the conservative strategies of Co. A and Co.
B?
l If each company knows that its competitor is playing
conservatively, will it change its strategy?
67
Both players will soon find out that
l It is better to shift from strategy to strategy (mix

the strategy) rather than play the same one all
the time as with a pure strategy approach.
l They should practice maximum secrecy with
their plans so that the opponent will not be able
to guess the next move.
l The average payoff is determined by the
fraction of time (proportion) that each of the
strategies is played, and there is a certain
fraction that is best for each player.
68
Mixed Strategy Game
l a game where there is no saddle point

l its best strategy is a random selection of
strategies which conform, in the long run,
to predetermined proportions (i.e., the
players play each strategy for a certain
percentage of the time)
How to get the time percentage?
69
Consider a 2 x 2 game.
Let P = fraction of the time A plays a1
1- P = fraction of the time A plays a2
Q = fraction of the time B plays b1
1- Q = fraction of the time B plays b2
Q 1–Q
P
1–P
70
Determining A’s best strategy
Multiply P and 1 – P with the corresponding
payoff and solve for P and 1- P by setting
column 1 equal to column 2 in the game.
Q 1–Q
P 4 2
1–P 1 10
4P +1(1- P) = 2P +10(1- P)
P=?
71
Determining B’s best strategy
Multiply Q and 1 – Q with the corresponding
payoff and solve for Q and 1 – Q by setting
column 1 equal to column 2 in the game.
Q 1–Q
P 4 2
1–P 1 10
4Q + 2(1- Q) = 1Q +10(1- Q)
Q=?
72
References
l Jin Xiao, J. Thomas, and J. Westwell. An Introduction
to Evolutionary Game Theory
l http://en.wikipedia.org/wiki/Nash_equilibrium
l Hoffmann, L.D. and G.L. Bradley. 1995. Applied
Finite Mathematics
l Mike Shor. Game Theory & Business Strategy
l Harcourt Brace & Company. 2001. Game Theory
73
Nonlinear Programming
Nonlinear Programming
n The methods seen so far have assumed that the
objective function and constraints are linear.
n Terms such as X13, 1/X2, log X3, or 5X1X2 are not
allowed.
n But there are many nonlinear relationships in the real
world that would require the objective function,
constraint equations, or both to be nonlinear.
n Excel can be used to solve these nonlinear
programming (NLP) problems.
n One disadvantage of NLP is that the solution yielded
may only be a local optimum, rather than a global
optimum.
n In other words, it may be an optimum over a
particular range, but not overall.
10-2
General NLP
Minimize f(x)
s.t. gi(x) (£, ³, =) bi, i = 1,…,m
x = (x1,…,xn)T is the n-dimensional vector of

decision variables
f (x) is the objective function
gi(x) are the constraint functions
bi are fixed known constants
Possible Optimal Solutions to NLPs
(not occurring at corner points)
objective objective
function level function level
curve curve
optimal optimal
solution solution
Feasible Feasible
Region Region
linear objective, nonlinear objective,

nonlinear constraints linear constraints
objective
function level objective
curve function level
curves
optimal
solution optimal
solution
Feasible
Region
Feasible
Region
nonlinear objective, nonlinear objective,
nonlinear constraints linear constraints
Linearly Constrained Convex
Function with Unique Global
Maximum
x2
Maximize f (x) = (x1 – 2)2 + (x2 – 2)2 5
subject to –3x1 – 2x2 ≤ –6 4
–x1 + x2 ≤ 3 3
2
x1 + x2 ≤ 7
1
2x1 – 3x2 ≤ 4
1 3 x1
2 4 5
An NLP Solution Strategy
X2 C D
B E
objective
function level
curves
Feasible
Region
A
(the starting point)
X1
Local vs. Global Optimal
Solutions
X2
Local optimal
solution
C
Local and
F global
E
Feasible G optimal
Region B solution
A
D
X1
Nonlinear Objective Function and
Linear Constraints
n The Great Western Appliance Company sells two
models of toaster ovens, the Microtoaster (X1)
and the Self-Clean Toaster Oven (X2).
n They earn a profit of $28 for each Microtoaster no
matter the number of units sold.
n For the Self-Clean oven, profits increase as more
units are sold due to a fixed overhead.
n The profit function for the Self-Clean over may be
expressed as:
21X2 + 0.25X22
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall 10-8

Linear Constraints
The objective function is nonlinear and there are
two linear constraints on production capacity and
sales time available.
Maximize profit = 28X1 + 21X2 + 0.25X22
subject to X1 + 21X2 ≤ 1,000 (units of production
capacity)
0.5X1 + 0.4X2 ≤ 500 (hours of sales time
available)
X1, X2 ≥ 0
When an objective function contains a squared term

and the problem constraints are linear, it is called a
quadratic programming problem.
Linear Constraints
Excel 2010 Solver Solution for Great Western
Appliance NLP Problem
Program 10.9

Both Nonlinear Objective Function
and Nonlinear Constraints
n The annual profit at a medium-sized (200-400
beds) Hospicare Corporation hospital depends on
the number of medical patients admitted (X1) and
the number of surgical patients admitted (X2).
n The objective function for the hospital is
nonlinear.
n They have identified three constraints, two of
which are nonlinear.
n Nursing capacity - nonlinear
n X-ray capacity - nonlinear
n Marketing budget required

The objective function and constraint equations for
this problem are:
Maximize profit = $13X1 + $6X1X2 + $5X2 + $1/X2
subject to 2X12 + 4X2 ≤ 90 (nursing capacity in thousands
of labor-days)
X1 + X23 ≤ 75 (x-ray capacity in thousands)
8X1 – 2X2 ≤ 61 (marketing budget required in
thousands of $)

Excel 2010 Solution for Hospicare’s NLP Problem
Program 10.10
Linear Objective Function and
Nonlinear Constraints
n Thermlock Corp. produces massive rubber
washers and gaskets like the type used to seal
joints on the NASA Space Shuttles.
n It combines two ingredients, rubber (X1) and oil
(X2).
n The cost of the industrial quality rubber is $5 per
pound and the cost of high viscosity oil is $7 per
pound.
n Two of the three constraints are nonlinear.

The firm’s objective function and constraints are:
Minimize costs = $5X1 + $7X2

subject to 3X1 + 0.25X 2 + 4X + 0.3X 2
≥ 125(hardness
1 2 2
constraint)
13X1 + X13 ≥ 80 (tensile
strength)
0.7X1 + X2 ≥ 17 (elasticity)

Excel 2010 Solution for Thermlock NLP Problem
Program 10.11

Simulation Modeling
To accompany
Quantitative Analysis for Management, Eleventh Edition,
Power Point slides created by Brian Peterson
Learning Objectives
1. Tackle a wide variety of problems by
simulation.
2. Understand the seven steps of conducting a
simulation.
3. Explain the advantages and disadvantages of
simulation.
4. Develop random number intervals and use
them to generate outcomes.
5. Understand alternative simulation packages
available.

Chapter Outline
1. Introduction
2. Advantages and Disadvantages of Simulation
3. Monte Carlo Simulation
4. Simulation and Inventory Analysis
5. Simulation of a Queuing Problem
6. Simulation Model for a Maintenance Policy
7. Other Simulation Issues

Introduction
n Simulation is one of the most widely used
quantitative analysis tools.
n To simulate is to try to duplicate the features,
appearance, and characteristics of a real system.
n We will build a mathematical model that comes as
close as possible to representing the reality of the
system.
n Physical models can also be built to test systems.

Introduction
Using simulation, a manager should:
1. Define a problem.
2. Introduce the variables associated with the
problem.
3. Construct a simulation model.
4. Set up possible courses of action for testing.
5. Run the simulation experiment.
6. Consider the results.
7. Decide what courses of action to take.

Process of Simulation
Define Problem
Introduce Important
Variables
Construct Simulation
Model
Specify Values of
Variables to Be Tested
Conduct the
Simulation
Examine the
Results
Figure 14.1 Select Best Course

of Action

Advantages and Disadvantages
of Simulation
The main advantages of simulation are:
1. It is relatively straightforward and flexible.
2. Recent advances in computer software make
simulation models very easy to develop.
3. Can be used to analyze large and complex
real-world situations.
4. Allows “what-if?” type questions.
5. Does not interfere with the real-world system.
6. Enables study of interactions between
components.
7. Enables time compression.
8. Enables the inclusion of real-world
complications.
Advantages and Disadvantages
of Simulation
The main disadvantages of simulation are:

1. It is often expensive as it may require a long,
complicated process to develop the model.
2. It does not generate optimal solutions; it is a
trial-and-error approach.
3. It requires managers to generate all conditions
and constraints of real-world problem.
4. Each model is unique and the solutions and
inferences are not usually transferable to other
problems.

Monte Carlo Simulation
n When systems contain elements that exhibit
chance in their behavior, the Monte Carlo method
of simulation can be applied.
n Some examples are:
1. Inventory demand.
2. Lead time for inventory.
3. Times between machine breakdowns.
4. Times between arrivals.
5. Service times.
6. Times to complete project activities.
7. Number of employees absent.

Monte Carlo Simulation
n The basis of the Monte Carlo simulation is
experimentation on the probabilistic elements
through random sampling.
n It is based on the following five steps:
1. Establishing a probability distribution for
important variables.
2. Building a cumulative probability distribution
for each variable.
3. Establishing an interval of random numbers
for each variable.
4. Generating random numbers.
5. Actually simulating a series of trials.
Harry’s Auto Tire
n A popular radial tire accounts for a large portion
of the sales at Harry’s Auto Tire.
n Harry wishes to determine a policy for managing
this inventory.
n He wants to simulate the daily demand for a
number of days.
Step 1: Establishing probability distributions

n One way to establish a probability distribution for a
given variable is to examine historical outcomes.
n Managerial estimates based on judgment and
experience can also be used.

Harry’s Auto Tire
Historical Daily Demand for Radial Tires at Harry’s
Auto Tire and Probability Distribution
FREQUENCY
DEMAND FOR TIRES (DAYS) PROBAILITY OF OCCURRENCE
0 10 10/200 = 0.05
1 20 20/200 = 0.10
2 40 40/200 = 0.20
3 60 60/200 = 0.30
4 40 40/200 = 0.20
5 30 30/200 = 0.15
200 200/200 = 1.00
Table 14.1

Harry’s Auto Tire
Step 2: Building a cumulative probability distribution
for each variable
n Converting from a regular probability to a
cumulative distribution is an easy job.
n A cumulative probability is the probability that a
variable will be less than or equal to a particular
value.
n A cumulative distribution lists all of the possible
values and the probabilities, as shown in Table
14.2.

Harry’s Auto Tire
Cumulative Probabilities for Radial Tires
DAILY DEMAND PROBABILITY CUMULATIVE PROBABILITY

0 0.05 0.05
1 0.10 0.15
2 0.20 0.35
3 0.30 0.65
4 0.20 0.85
5 0.15 1.00
Table 14.2

Harry’s Auto Tire
Step 3: Setting random number intervals
n Assign a set of numbers to represent each
possible value or outcome.
n These are random number intervals.
n A random number is a series of digits that have

been selected by a totally random process.
n Therange of the random number intervals

corresponds exactly to the probability of the
outcomes as shown in Figure 14.2.

Harry’s Auto Tire
Graphical Representation of the Cumulative
Probability Distribution for Radial Tires
1.00
1.00 – – 00
0.85
– 86
0.80 – 85
Represents 4
Tires Demanded
Cumulative Probability
0.65 66
– 65
0.60 –
Random
Number
0.40 – 0.35
s
– 36
35
0.20 – 0.15
– 16
15 Represents 1
0.05 06 Tire Demanded
– 05
Figure 14.2 0.00 – –
01
0 1 2 3 4 5
Daily Demand for Radials
Harry’s Auto Tire
Assignment of Random Number Intervals for Harry’s
Auto Tire
DAILY DEMAND PROBABILITY CUMULATIVE INTERVAL OF
PROBABILITY RANDOM NUMBERS
0 0.05 0.05 01 to 05
1 0.10 0.15 06 to 15
2 0.20 0.35 16 to 35
3 0.30 0.65 36 to 65
4 0.20 0.85 66 to 85
5 0.15 1.00 86 to 00
Table 14.3

Harry’s Auto Tire
Step 4: Generating random numbers
n Random numbers can be generated in several
ways.
n Large problems will use computer program to
generate the needed random numbers.
n For small problems, random processes like
roulette wheels or pulling chips from a hat may
be used.
n The most common manual method is to use a
random number table.
n Because everything is random in a random
number table, we can select numbers from
anywhere in the table to use in the simulation.
Harry’s Auto Tire
Table of random numbers (partial)
52 06 50 88 53 30 10 47 99 37
37 63 28 02 74 35 24 03 29 60
82 57 68 28 05 94 03 11 27 79
69 02 36 49 71 99 32 10 75 21
98 94 90 36 06 78 23 67 89 85
96 52 62 87 49 56 59 23 78 71
33 69 27 21 11 60 95 89 68 48
50 33 50 95 13 44 34 62 64 39
88 32 18 50 62 57 34 56 62 31
90 30 36 24 69 82 51 74 30 35
Table 14.4

Harry’s Auto Tire
Step 5: Simulating the experiment
n We select random numbers from Table 14.4.
n The number we select will have a corresponding
range in Table 14.3.
n We use the daily demand that corresponds to the
probability range aligned with the random
number.

Harry’s Auto Tire
Ten-day Simulation of Demand for Radial Tires
DAY RANDOM NUMBER SIMULATED DAILY DEMAND
1 52 3
2 37 3
3 82 4
4 69 4
5 98 5
6 96 5
7 33 2
8 50 3
9 88 5
10 90 5
39 = total 10-day demand
3.9 = average daily demand for tires
Table 14.5
Harry’s Auto Tire
Note that the average demand from this simulation
(3.9 tires) is different from the expected daily
demand.
Expected 5
daily = å(Probability of i tires)(Demand of i tires)
demand i =0
= (0.05)(0) + (0.10)(1) + (0.20)(2) + (0.30)(3)

+ (0.20)(4) + (0.15)(5)
= 2.95 tires
If this simulation were repeated hundreds or
thousands of times it is much more likely the
average simulated demand would be nearly the
same as the expected demand.
QM for Windows Output Screen for
Simulation of Harry’s Auto Tire Example
Program 14.1

Simulation with Excel Spreadsheets
Using Excel 2010 to Simulate Tire Demand for Harry’s
Auto Tire Shop
Program 14.2

Using Excel 2010 to Simulate Tire Demand for Harry’s
Auto Tire Shop
Program 14.2

Generating Normal Random Numbers in Excel
Program 14.3
Excel QM Simulation of Harry’s Auto Tire Example
Program 14.4

Simulation and Inventory Analysis
n We have seen deterministic inventory models.

n In many real-world inventory situations, demand
and lead time are variables.
n Accurate analysis is difficult without simulation.
n We will look at an inventory problem with two
decision variables and two probabilistic
components.
n The owner of a hardware store wants to establish
order quantity and reorder point decisions for a
product that has probabilistic daily demand and
reorder lead time.

Simkin’s Hardware Store
n The owner of a hardware store wants to find a
good, low cost inventory policy for an electric
drill.
n Simkin identifies two types of variables,
controllable and uncontrollable inputs.
n The controllable inputs are the order quantity and
reorder points.
n The uncontrollable inputs are daily demand and
variable lead time.
n The demand data for the drill is shown in Table
14.6.

Probabilities and Random Number Intervals for
Daily Ace Drill Demand
(1) (2) (4) (5)

DEMAND FOR FREQUENCY (3) CUMULATIVE INTERVAL OF
ACE DRILL (DAYS) PROBABILITY PROBABILITY RANDOM NUMBERS
0 15 0.05 0.05 01 to 05
1 30 0.10 0.15 06 to 15
2 60 0.20 0.35 16 to 35
3 120 0.40 0.75 36 to 75
4 45 0.15 0.90 76 to 90
5 30 0.10 1.00 91 to 00
300 1.00
Table 14.6

Probabilities and Random Number Intervals for
Reorder Lead Time
(1) (2) (4) (5)

LEAD TIME FREQUENCY (3) CUMULATIVE RANDOM NUMBER
(DAYS) (ORDERS) PROBABILITY PROBABILITY INTERVAL
1 10 0.20 0.20 01 to 20
2 25 0.50 0.70 21 to 70
3 15 0.30 1.00 71 to 00
50 1.00
Table 14.7

n The third step is to develop a simulation model.
n A flow diagram, or flowchart, is helpful in this
process.
n The fourth step in the process is to specify the
values of the variables that we wish to test.
n The first policy that Simkin wants to test is an
order quantity of 10 with a reorder point of 5.
n The fifth step is to actually conduct the
simulation.
n The process is simulated for a 10 day period.

Flow
Diagram for
Simkin’s
Inventory
Example
Figure 14.3

Using the table of random numbers, the simulation is
conducted using a four-step process:
1. Begin each day by checking whether an ordered
inventory has arrived. If it has, increase the current
inventory by the quantity ordered.
2. Generate a daily demand from the demand probability by
selecting a random number.
3. Compute the ending inventory every day. If on-hand
inventory is insufficient to meet the day’s demand, satisfy
as much as possible and note the number of lost sales.
4. Determine whether the day’s ending inventory has
reached the reorder point. If necessary place an order.

Simkin Hardware’s First Inventory Simulation
ORDER QUANTITY = 10 UNITS REORDER POINT = 5 UNITS
(2) (3) (4) (6) (7) (9) (10)
(1) UNITS BEGINNING RANDOM (5) ENDING LOST (8) RANDOM LEAD
DAY RECEIVED INVENTORY NUMBER DEMAND INVENTORY SALES ORDER NUMBER TIME
1 … 10 06 1 9 0 No
2 0 9 63 3 6 0 No
3 0 6 57 3 3 0 Yes 02 1
4 0 3 94 5 0 2 No
5 10 10 52 3 7 0 No
6 0 7 69 3 4 0 Yes 33 2
7 0 4 32 2 2 0 No
8 0 2 30 2 0 0 No
9 10 10 48 3 7 0 No
10 0 7 88 4 3 0 Yes 14 1
Total 41 2
Table 14.8

Analyzing Simkin’s Inventory Cost
n The objective is to find a low-cost solution so
Simkin must determine the costs.
n Equations for average daily ending inventory,
average lost sales, and average number of orders
placed.
Average
41 totalunits
ending = = 4.1 units per day
inventory 10 days
Average = 2 sales lost = 0.2 unit per day

lost sales 10 days
Average 3 orders
number of = = 0.3 order per day
orders placed
10 days
n Simkin’s store is open 200 days a year.
n Estimated ordering cost is $10 per order.
n Holding cost is $6 per drill per year.
n Lost sales cost $8.
Daily order cost = (Cost of placing one order)

x (Number of orders placed per day)
= $10 per order x 0.3 order per day = $3
Daily holding cost = (Cost of holding one unit for one day) x
(Average ending inventory)
= $0.03 per unit per day x 4.1 units per day
= $0.12

n Simkin’s store is open 200 days a year.
n Estimated ordering cost is $10 per order.
n Holding cost is $6 per drill per year.
n Lost sales cost $8.
Daily stockout cost = (Cost per lost sale)

x (Average number of lost sales per
day)
= $8 per lost sale x 0.2 lost sales per day
= $1.60
Total daily
inventory cost = Daily order cost + Daily holding cost
+ Daily stockout cost
= $4.72
n For the year, this policy would cost approximately
$944.
n This simulation should really be extended for
many more days, perhaps 100 or 1,000 days.
n Even after a larger simulation, the model must be
verified and validated to make sure it truly
represents the situation on which it is based.
n If we are satisfied with the model, additional
simulations can be conducted using other values
for the variables.
n After simulating all reasonable combinations,
Simkin would select the policy that results in the
lowest total cost.
Simulation of a Queuing Problem
n Modeling waiting lines is an important application
of simulation.
n The assumptions of queuing models are quite
restrictive.
n Sometimes simulation is the only approach that
fits.
n In this example, arrivals do not follow a Poisson
distribution and unloading rates are not
exponential or constant.

Port of New Orleans
n Fully loaded barges arrive at night for unloading.
n The number of barges each night varies from 0 – 5,
and the number of barges vary from day to day.
n The supervisor has information which can be used
to create a probability distribution for the daily
unloading rate.
n Barges are unloaded first-in, first-out.
n Barges must wait for unloading which is expensive.
n The dock superintendent wants to do a simulation
study to enable him to make better staffing
decisions.

Port of New Orleans
Overnight Barge Arrival Rates and Random Number
Intervals
NUMBER OF CUMULATIVE RANDOM

ARRIVALS PROBABILITY PROBABILITY NUMBER INTERVAL
0 0.13 0.13 01 to 13
1 0.17 0.30 14 to 30
2 0.15 0.45 31 to 45
3 0.25 0.70 46 to 70
4 0.20 0.90 71 to 90
5 0.10 1.00 91 to 00
Table 14.9

Port of New Orleans
Unloading Rates and Random Number Intervals
DAILY UNLOADING CUMULATIVE RANDOM

RATE PROBABILITY PROBABILITY NUMBER INTERVAL
1 0.05 0.05 01 to 05
2 0.15 0.20 06 to 20
3 0.50 0.70 21 to 70
4 0.20 0.90 71 to 90
5 0.10 1.00 91 to 00
1.00
Table 14.10

Queuing Simulation of Port of
New Orleans Barge Unloadings
(1) (2) (3) (4) (5) (6) (7)
DAY NUMBER DELAYED RANDOM NUMBER OF TOTAL TO BE RANDOM NUMBER
FROM PREVIOUS DAY NUMBER NIGHTLY ARRIVALS UNLOADED NUMBER UNLOADED
1 — 52 3 3 37 3
2 0 06 0 0 63 0
3 0 50 3 3 28 3
4 0 88 4 4 02 1
5 3 53 3 6 74 4
6 2 30 1 3 35 3
7 0 10 0 0 24 0
8 0 47 3 3 03 1
9 2 99 5 7 29 3
10 4 37 2 6 60 3
11 3 66 3 6 74 4
12 2 91 5 7 85 4
13 3 35 2 5 90 4
14 1 32 2 3 73 3
15 0 00 5 5 59 3
20 41 39
Total delays Total arrivals Total unloadings
Table 14.11
Port of New Orleans
Three important pieces of information:
Average number of barges = 20 delays

delayed to the next day 15 days
= 1.33 barges delayed per day
Average number of 41 arrivals

= = 2.73 arrivals
nightly arrivals 15 days
Average number of barges 39 unloadings

unloaded each day = = 2.60 unloadings
15 days

Excel Model for the Port of New
Orleans Queuing Simulation
Program 14.5
Excel Model for the Port of New
Orleans Queuing Simulation
Program 14.5

Simulation Model for a
Maintenance Policy
n Simulation can be used to analyze different
maintenance policies before actually
implementing them.
n Many options regarding staffing levels, parts
replacement schedules, downtime, and labor
costs can be compared.
n This can include completely shutting down
factories for maintenance.

Three Hills Power Company
n Three Hills provides power to a large city through
a series of almost 200 electric generators.
n The company is concerned about generator
failures because a breakdown costs about $75
per generator per hour.
n Their four repair people earn $30 per hour and
work rotating 8 hour shifts.
n Management wants to evaluate the:
1. Service maintenance cost.
2. Simulated machine breakdown cost.
3. Total cost.

n There are two important maintenance system
components:
n Time between successive generator breakdowns which
varies from 30 minutes to three hours.
n The time it takes to repair the generators which ranges
from one to three hours in one hour blocks
n A next event simulation is constructed to study
this problem.

Three Hills
Flow
Diagram
Figure 14.4

Time between generator breakdowns at Three Hills Power
TIME BETWEEN
RECORDED NUMBER RANDOM
MACHINE OF TIMES CUMULATIVE NUMBER
FAILURES (HRS) OBSERVED PROBABILITY PROBABILITY INTERVAL
0.5 5 0.05 0.05 01 to 05
1.0 6 0.06 0.11 06 to 11
1.5 16 0.16 0.27 12 to 27
2.0 33 0.33 0.60 28 to 60
2.5 21 0.21 0.81 61 to 81
3.0 19 0.19 1.00 82 to 00
Total 100 1.00
Table 14.12

Generator repair times required
NUMBER RANDOM
REPAIR TIME OF TIMES CUMULATIVE NUMBER
REQUIRED (HRS) OBSERVED PROBABILITY PROBABILITY INTERVAL
1 28 0.28 0.28 01 to 28
2 52 0.52 0.80 29 to 80
3 20 0.20 1.00 81 to 00
Total 100 1.00
Table 14.13

Simulation of generator breakdowns and repairs
(5) (6) (9)
TIME REPAIR- RANDOM NUMBER
(2) (3) PERSON IS NUMBER (7) (8) OF
(1) RANDOM TIME (4) FREE TO FOR REPAIR TIME HOURS
BREAKDOWN NUMBER FOR BETWEEN TIME OF BEGIN THIS REPAIR TIME REPAIR MACHINE
NUMBER BREAKDOWNS BREAKDOWNS BREAKDOWN REPAIR TIME REQUIRED ENDS DOWN
1 57 2 02:00 02:00 07 1 03:00 1
2 17 1.5 03:30 03:30 60 2 05:30 2
3 36 2 05:30 05:30 77 2 07:30 2
4 72 2.5 08:00 08:00 49 2 10:00 2
5 85 3 11:00 11:00 76 2 13:00 2
6 31 2 13:00 13:00 95 3 16:00 3
7 44 2 15:00 16:00 51 2 18:00 3
8 30 2 17:00 18:00 16 1 19:00 2
9 26 1.5 18:30 19:00 14 1 20:00 1.5
10 09 1 19:30 20:00 85 3 23:00 3.5
11 49 2 21:30 23:00 59 2 01:00 3.5
12 13 1.5 23:00 01:00 85 3 04:00 5
13 33 2 01:00 04:00 40 2 06:00 5
14 89 3 04:00 06:00 42 2 08:00 4
15 13 1.5 05:30 08:00 52 2 10:00 4.5
Total 44
14-54
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall
Table 14.14
Cost Analysis of the
Simulation
n The simulation of 15 generator breakdowns
covers 34 hours of operation.
n The analysis of this simulation is:
Service
maintenance = 34 hours of worker service time
cost x $30 per hour
= $1,020
Simulated machine
breakdown cost = 44 total hours of breakdown
x $75 lost per hour of downtime
= $3,300
Total simulated
maintenance cost of = Service cost + Breakdown cost
the current system = $1,020 + $3,300
= $4,320
Cost Analysis of the
Simulation
n The cost of $4,320 should be compared with
other alternative plans to see if this is a “good”
value.
n The company might explore options like adding
another repairperson.
n Strategies such as preventive maintenance might
also be simulated for comparison.

Excel Spreadsheet Model for Three Hills
Power Company Maintenance Problem
Program 14.6

Excel Spreadsheet Model for Three Hills
Power Company Maintenance Problem
Program 14.6

Other Simulation Models
n Simulation models are often broken into

three categories:
n The Monte Carlo method.
n Operational gaming.
n Systems simulation.
n Though theoretically different,
computerized simulation has tended to
blur the differences.

Operational Gaming
n Operational gaming refers to simulation involving
two or more competing players.
n The best examples of this are military games and
business games.
n These types of simulation allow the testing of
skills and decision-making in a competitive
environment.

Systems Simulation
n Systems simulation is similar in that allows users
to test various managerial policies and decisions
to evaluate their effect on the operating
environment.
n This models the dynamics of large systems.
n A corporate operating system might model sales,
production levels, marketing policies,
investments, union contracts, utility rates,
financing, and other factors.
n Economic simulations, often called econometric
models, are used by governments, bankers, and
large organizations to predict inflation rates,
domestic and foreign money supplies, and
unemployment levels.
Systems Simulation
Inputs and Outputs of a Typical Economic System
Simulation
Income Tax Gross National

Levels Product
Corporate Tax Inflation Rates
Rates Econometric Model Unemployment
Interest Rates (in Series of Rates
Mathematical
Government Equations) Monetary
Spending Supplies
Foreign Trade Population
Policy Growth Rates
Figure 14.5

Verification and Validation
n It is important that a simulation model be checked
to see that it is working properly and providing
good representation of the real world situation.
n The verification process involves determining
that the computer model is internally consistent
and following the logic of the conceptual model.
n Verification answers the question “Did we build
the model right?”
n Validation is the process of comparing a
simulation model to the real system it represents
to make sure it is accurate.
n Validation answers the question “Did we build the
right model?”
Role of Computers in Simulation
n Computers are critical in simulating complex
tasks.
n General-purpose programming languages can be
used for simulation, but a variety of simulation
software tools have been developed to make the
process easier:
n Arena
n ProModel
n SIMUL8
n ExtendSim
n Proof 5
n Excel and add-ins can also be used for
simulation problems

Copyright
All rights reserved. No part of this publication may be

reproduced, stored in a retrieval system, or transmitted, in
any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise, without the prior
written permission of the publisher. Printed in the United
States of America.

Or Departmental Exam

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Or Departmental Exam

Uploaded by

Copyright:

Available Formats

Decision Analysis

1. List the steps of the decision-

n What is involved in making a good

1. Clearly define the problem at hand

Step 1 – Define the problem

Step 4 – List the payoffs

Construct a small plant 100,000 –20,000

Type 1: Decision making under certainty

Weighted average = a(maximum in row)

EMV (alternative i) = (payoff of first state of nature)

EMV (large plant) = (0.50)($200,000) + (0.50)(–$180,000)

EVwPI = (best payoff for first state of nature)

n Expected opportunity loss (EOL) is the

Do nothing 200,000 0 100,000

P = probability of a favorable market

$200,000 EMV (large plant)

$100,000 EMV (small plant)

0 EMV (do nothing)

$200,000 EMV (large plant)

$100,000 EMV (small plant)

0 EMV (do nothing)

1. Define the problem

Favorable Market (0.5)

EMV for Node = (0.5)($100,000)

Favorable Market (0.50)

1. Given favorable survey results,

3. Compute the expected value of the market survey,

$10,000 Favorable Market (0.50)

$40,000 Favorable Market (0.50)

n Thompson wants to know the actual value

= (EV with sample information + cost)

EVSI = ($49,200 + $10,000) – $40,000 = $19,200

Is the $10,000–market survey worthwhile?

n How sensitive are the decisions to

EMV(node 1) = ($106,400)p +($2,400)(1 – p)

We are indifferent when the EMV of node 1 is the

$104,000p + $2,400 = $40,000

Understand the importance and use of

Expected utility of alternative 2 = Expected utility of alternative 1

Utility for $5,000 = U($5,000) = pU($10,000) + (1 – p)U($0)

Utility for $7,000 = 0.90

n Using the three utilities for different dollar amounts,

0.9 – U ($7,000) = 0.90

0.8 – U ($5,000) = 0.80

0.5 – U ($3,000) = 0.50

n A risk seeker gets more utility from greater risk

n Someone with risk indifference will have a linear

n Once a utility curve has been developed

n Mark Simkin loves to gamble.

Decision Facing Mark Simkin

Figure 3.11 Mark Does Not Play the Game

n Step 1– Define Mark’s utilities.

n Step 2 – Replace monetary values with

Using Expected Utilities in Decision Making

E = 0.162 Tack Lands Utility

5-liter container 8-liter container

It is a practical method for finding solutions to extremely

It has the ability to attack difficult problems that other

Bellman was interested in planning, decision making and thinking

• designed to enhance the computational efficiency of solving

Variable Item, j Value, 𝒗𝒋 Weight, 𝒘𝒋

1. Determine future states or conditions by

n Markov analysis is a technique that deals with

n This discussion will be limited to Markov