Decision Analysis:

Decision Making under

Uncertainty & Risk
Decision Making

Ata ul Musawir

This topic is based on:

Anderson, D. R., Sweeney, D. J., Williams, T. A., Camm, J. D., Cochran, J. J., Fry, M. J. &
Ohlmann, J. W. (2013). Quantitative methods for business. 12th ed. Chapter 4: Decision
analysis. South-Western, Cengage Learning.
 What is Decision Analysis?
• Decision analysis is a systematic, quantitative, and visual approach to
addressing and evaluating important choices confronted by businesses. The
term decision analysis originated in 1964 by Ronald A. Howard, professor of
Management Science and Engineering at Stanford University.

• Decision analysis utilizes a variety of tools to evaluate all relevant

information to aid in the decision making process and incorporates aspects
of psychology, management techniques and training, and economics.

• Decision analysis can be used to develop an optimal strategy when a

decision maker is faced with several decision alternatives and an uncertain
or risk-filled pattern of future events.

• It can be used by individuals or groups attempting to make a variety of

decisions relating to, for example, risk management, capital investments,
and business strategy.
 Good Decisions vs. Good Outcomes

• The goal of decision analysis is to help individuals make good decisions.

But good decisions do not always result in good outcomes.

• For example, suppose that after carefully considering all the factors
involved in the two job offers, you decide to accept the position with
company B. After working for this company for a few months, it suddenly
announces that, in an effort to cut costs, it is closing the office in which
you work and eliminating your job.

• Did you make a bad decision? Probably not. Unforeseeable

circumstances beyond your control caused you to experience a bad
outcome, but it would be unfair to say that you made a bad decision.
Good decisions sometimes result in bad outcomes.
 Good Decisions vs. Good Outcomes
• The techniques for decision analysis presented in this topic can help you
make good decisions within the constraints of bounded rationality, but
cannot guarantee that good outcomes always will occur as a result of
those decisions.

• Even when a good decision is made, probabilities often play a role in

determining whether a good or bad outcome occurs.

• However, using a structured approach to make decisions should provide

us with better insight and intuition about the decision problems we face.

• Therefore, we can expect good outcomes to occur more frequently

when using a structured approach to decision making than if we make
decisions in a more haphazard manner (assuming good data are used
with reasonable assumptions).
 Outline

1. Problem Formulation

2. Decision Making without Probabilities

3. Introduction to Probability Theory

4. Decision Making with Probabilities

5. Risk Analysis and Sensitivity Analysis

 Decision Analysis vs.
Other DM Tools & Techniques
Qualitative Quantitative

Qualitative T&Ts AHP Decision Analysis

(i) Payoffs of decision alternatives
Data Minimal (varies based on Pairwise comparisons of under all possible states of nature
Required technique) alternatives and/or criteria (ii) Probabilities of states of nature
(if possible)
Systematic conversion of Calculating optimal payoffs and/or
Analytical Subjective
pairwise comparisons into expected values of decision
Method assessments
preference vectors alternatives
Type of Subjective (qualitative Subjective (quantified Objective (expected values of
Output factors) preferences) decision alternatives)
No. of
Single/Multiple Multiple Single (e.g. profit or cost)
Criteria
Making sense of available Choosing a decision
Choosing a decision alternative
Suitable information, generating and alternative that best
that provides the highest (or
for discussing ideas, identifying satisfies the preferences of
lowest) expected value
decision alternatives the decision maker(s)
 1. Problem Formulation

• The first step in the decision analysis process is problem formulation.

• We begin with a statement of the problem.

• We then identify:
• The decision alternatives
• The uncertain future events, referred to as chance events (or states of nature)
• The consequences (or payoffs) associated with each combination of decision
alternative and chance event outcome

• We can further summarize the information of the problem using various

decision analysis tools, including:
• Influence diagrams
• Payoff tables
• Decision trees
 1. Problem Formulation
PDC Example
• Let us begin by considering the example of a construction project of the
Pittsburgh Development Corporation (PDC). The company recently
purchased land that will be the site of a new luxury condominium
complex. Once the complex is completed, PDC plans sell the
condominium units for profit.

• PDC commissioned preliminary architectural drawings for three different

potential projects:
• One with 30 condominium units
• One with 60 condominium units
• One with 90 condominium units

• The financial success of the project depends upon the size of the
condominium complex and the chance event concerning the demand for
the condominiums.
 1. Problem Formulation
PDC Example: Decision Alternatives
• The statement of the PDC decision problem is to select the size of the
new luxury condominium project that will lead to the largest profit given
the uncertainty concerning the demand for the condominiums.

• Given the statement of the problem, it is clear that the decision is to

select the best size for the condominium complex. PDC has the following
three decision alternatives:

1. d1 = a small complex with 30 condominiums

2. d2 = a medium complex with 60 condominiums

3. d3 = a large complex with 90 condominiums

 1. Problem Formulation:
PDC Example: States of Nature
• A factor in selecting the best decision alternative is the uncertainty
associated with the chance event concerning the demand for the
condominiums. In this case, there are two possible chance event
outcomes: strong demand or weak demand. In decision analysis, the
possible chance event outcomes are referred to as the states of nature.

• The states of nature are defined so they are mutually exclusive (no more
than one can occur) and collectively exhaustive (at least one must
occur); thus one and only one of the possible states of nature will occur.

• For the PDC problem, the chance event concerning the demand for the
condominiums has two possible states of nature:

1. s1 = strong demand for the condominiums

2. s2 = weak demand for the condominiums

 1. Problem Formulation
PDC Example: Influence Diagram
• An influence diagram is a graphical device that shows the relationships
among the decisions, the chance events, and the consequences for a
decision problem. Rectangles or squares depict decision nodes, circles
or ovals depict chance nodes, and diamonds depict consequence
nodes. The lines connecting the nodes, referred to as arcs, show the
direction of influence that the nodes have on one another.
 1. Problem Formulation
PDC Example: Payoff Table
• Given the three decision alternatives and the two states of nature, which
complex size should PDC choose? To answer this question, PDC will need
to know the consequence associated with each decision alternative and
each state of nature. In decision analysis, we refer to the consequence
resulting from a specific combination of a decision alternative and a
state of nature as a payoff. A table showing payoffs for all combinations
of decision alternatives and states of nature is a payoff table.

• The figure below shows the payoff table for the PDC problem, which was
developed based on data collected and analyzed by the management.
 1. Problem Formulation
PDC Example: Decision Tree
• A decision tree provides a graphical representation of the decision-making
process.

• Figure 4.2 presents a decision tree for the PDC problem. It shows the natural
or logical progression that will occur over time. First, PDC must make a
decision regarding the size of the condominium complex (d1, d2, or d3). Then,
after the decision is implemented, either state of nature s1 or s2 will occur.
The number at each endpoint of the tree indicates the payoff associated
with a particular sequence.

• The decision tree in Figure 4.2 shows four nodes: 1 to 4. Squares are used to
depict decision nodes and circles are used to depict chance nodes. Thus,
node 1 is a decision node, and nodes 2, 3, and 4 are chance nodes.

• The branches connecting the decision node (1) to the chance nodes (2, 3, 4)
correspond to the three decision alternatives. The branches leaving each
chance node correspond to the states of nature. The payoffs are shown at
the end of the states of nature branches.
 1. Problem Formulation
PDC Example: Decision Tree
 2. Decision Making without Probabilities
• In this section we consider approaches to decision making that do not
require knowledge of the probabilities of the states of nature.

• Three general approaches may be used in such situations:

1. The Optimistic Approach
2. The Conservative Approach
3. The Minimax Regret Approach

• These approaches are appropriate in situations in which the decision

maker has little confidence in his or her ability to assess the
probabilities, or in which a simple best-case and worst-case analysis is
desirable.

• Because different approaches sometimes lead to different decision

recommendations, the decision maker must understand the approaches
available and then select the specific approach that, according to the
judgment of the decision maker, is the most appropriate.
 2. Decision Making without Probabilities
PDC Example: Optimistic Approach
• The optimistic approach evaluates each decision alternative in terms of
the best payoff that can occur. The decision alternative that is
recommended is the one that provides the best possible payoff.

• For a problem in which maximum profit is desired, as in the PDC

problem, the optimistic approach would lead the decision maker to
choose the alternative corresponding to the largest profit.

• For problems involving minimization (e.g. costs), this approach leads to

choosing the alternative with the smallest payoff.
2. Decision Making without Probabilities
PDC Example: Optimistic Approach
 2. Decision Making without Probabilities
PDC Example: Conservative Approach
• The conservative approach evaluates each decision alternative in terms
of the worst payoff that can occur. The decision alternative
recommended is the one that provides the best of the worst possible
payoffs.

• For a problem in which the output measure is profit, as in the PDC

problem, the conservative approach would lead the decision maker to
choose the alternative that maximizes the minimum possible profit that
could be obtained.

• For problems involving minimization (e.g. costs), this approach identifies

the alternative that will minimize the maximum payoff.
2. Decision Making without Probabilities
PDC Example: Conservative Approach
 2. Decision Making without Probabilities
PDC Example: Minimax Regret Approach
• In decision analysis, regret is the difference between the payoff
associated with a particular decision alternative and the payoff
associated with the decision that would yield the most desirable payoff
for a given state of nature.

• Thus, regret represents how much potential payoff one would forgo by
selecting a particular decision alternative given that a specific state of
nature will occur. This is why regret is often referred to as opportunity
loss.

• Under the minimax regret approach to decision making one would

choose the decision alternative that minimizes the maximum state of
regret that could occur over all possible states of nature. This approach
is neither purely optimistic nor purely conservative.
2. Decision Making without Probabilities
PDC Example: Minimax Regret Approach
2. Decision Making without Probabilities
PDC Example: Minimax Regret Approach

|20 - 8| = |7 - 7| =
|20 - 14| = |7 - 5| =
|20 - 20| = |7 - (-9)| =
2. Decision Making without Probabilities
PDC Example: Minimax Regret Approach
 2. Decision Making without Probabilities
Investments Example
• Khushkismat Khan has found a Rs. 1,000 note in his old shalwar. He has to
decide how to invest the money for one year.

• His wise mamu suggests five potential investments (decision alternatives):

1. Gold
2. Junk Bond
3. Growth Stock
4. Certificate of Deposit (fixed deposit)
5. Stock Option

• The payoff from each of these options will be affected by the performance of
the stock market over the coming year. The stock market’s performance can
be classified into five possibilities (states of nature):
1. Large Rise
2. Small Rise
3. No Change
4. Small Fall
5. Large Fall
 2. Decision Making without Probabilities
Investments Example
• The table below shows the expected payoffs for the five potential
investment options under the five potential states of nature.

Decision States of Nature

Alternatives Large Rise Small Rise No Change Small Fall Large Fall
Gold -100 100 200 300 0
Bond 250 200 150 -100 -150
Stock 500 250 100 -200 -600
C/D account 60 60 60 60 60
Stock option 200 150 150 -200 -150
 2. Decision Making without Probabilities
Investments Example
• Note that the stock option alternative is dominated by the bond alternative,
i.e. for every possible state of nature, the bond alternative leads to an equal
or greater payoff than the stock option alternative.

• Therefore, we can omit the stock option alternative from our calculations.

Decision States of Nature

Alternatives Large Rise Small Rise No Change Small Fall Large Fall
Gold -100 100 200 300 0
Bond 250 200 150 -100 -150
Stock 500 250 100 -200 -600
C/D account 60 60 60 60 60
Stock option 200 150 150 -200 -150
 2. Decision Making without Probabilities
Investments Example
• For the remaining options, the optimistic approach suggests that stock is
the best alternative as it has the highest maximum payoff of 500.

The Maximax Criterion Maximum

Decision Large rise Small rise No change Small fall Large fall Payoff
Gold -100 100 200 300 0 300
Bond 250 200 150 -100 -150 200
250
Stock 500 250 100 -200 -600 500
C/D 60 60 60 60 60 60
 2. Decision Making without Probabilities
Investments Example
• The conservative approach suggests that the C/D account is the best
alternative as it has the highest minimum payoff of 60.

The
TheMaximin
MaximinCriterion
Criterion Minimum
Minimum
Decisions
Decisions Large
LargeRise
Rise Small
Smallrise
rise NoNoChange
Change Small
SmallFall
Fall Large
LargeFall
Fall Payoff
Payoff
Gold
Gold -100
-100 100
100 200
200 300
300 00 -100
-100
Bond
Bond 250
250 200
200 150
150 -100
-100 -150
-150 -150
-150
Stock
Stock 500
500 250
250 100
100 -200
-200 -600
-600 -600
-600
C/D
C/Daccount
account 60
60 60
60 60
60 60
60 60
60 60
60
 2. Decision Making without Probabilities
Investments Example
• The minimax regret approach suggests that bond is the best alternative
as it has the lowest maximum regret of 400.

The
The Regret
RegretTable
Table Maximum
Maximum
Decision
Decision Large
Large rise
rise Small
Smallrise
riseNo
Nochange
change Small
Smallfall
fall Large
Large fall
fall Regret
Regret
Gold
Gold 600
600 150
150 00 00 60
60 600
600
Bond
Bond 250
250 50
50 50
50 400
400 210
210 400
400
Stock
Stock 00 00 100
100 500
500 660
660 660
660
C/D
C/D 440
440 190
190 140
140 240
240 00 440
440
 2. Decision Making without Probabilities

• Note that the three approaches discussed in this section provide

different recommendations, which in itself is not a bad thing. It simply
reflects the difference in decision-making philosophies that underlie the
various approaches.

• Ultimately, the decision maker will have to choose the most appropriate
approach and then make the final decision accordingly.

• The main criticism of the approaches discussed in this section is that

they do not consider any information about the probabilities of the
various states of nature.

• In the next section we discuss an approach that utilizes probability

information in selecting a decision alternative. But first, we will take a
brief overview of probability theory.
Exercises
Exercises
Exercises
Exercises
 3. Introduction to Probability Theory
• Probability is a numeric measure of the likelihood that an event will
occur. Probability values are always assigned on a scale from 0 to 1.
• A probability of 0 indicates that an event will not occur (impossible event)
• A probability of 1 indicates that an event is certain to occur (certain event)
• Other probabilities between 0 and 1 represent varying degrees of likelihood
that an event will occur
 3. Introduction to Probability Theory
Defining an Experiment
• In discussing probability, we define an experiment to be any process that
generates well-defined outcomes.

• On any single repetition of an experiment, one and only one of the

possible experimental outcomes will occur. Several examples of
experiments and their associated outcomes follow.
 3. Introduction to Probability Theory
Defining the Sample Space
• The first step in analyzing a particular experiment is to carefully define
the experimental outcomes. When we define all possible experimental
outcomes, we identify the sample space for the experiment; that is, the
set of all possible experimental outcomes. For example:

• The sample space for flipping a coin is:

• The sample space for selecting a part for inspection is:

• The sample space for rolling a die is:

 3. Introduction to Probability Theory
Two Principles for Assigning Probabilities
• The probability of an experimental outcome is a numeric measure of the
likelihood that the experimental outcome will occur on a single
repetition of the experiment. In assigning probabilities to experimental
outcomes, two basic principles of probability must be satisfied:
 3. Introduction to Probability Theory
Assigning Probabilities: Classical Method
• To illustrate the classical method of assigning probabilities, let us again
consider the experiment of flipping a coin. On any one flip, we will
observe one of two experimental outcomes: head or tail. It is reasonable
to assume the two possible outcomes are equally likely. Therefore, there
is a 0.50 probability of getting either head or tail.

• Another example:
 3. Introduction to Probability Theory:
Assigning Probabilities: Relative Frequency Method
• To illustrate the relative frequency method, suppose that in a test
market evaluation of a product, 400 potential customers were
contacted, where 100 purchased the product, but 300 did not.

• In effect, we have repeated the experiment of contacting a customer 400

times and have found that the product was purchased 100 times. Thus,
we might decide to use the relative frequency of the number of
customers that purchased the product as an estimate of the probability
of a customer making a purchase:
• We could assign a probability of 100/400 = 0.25 to the experimental outcome of
purchasing the product.
• Similarly, 300/400 = 0.75 could be assigned to the experimental outcome of not
purchasing the product.

• It is important to note that if we repeated this experiment several times,

we may not obtain exactly the same probability estimate that we
obtained from the first 400 trials of the experiment.
 3. Introduction to Probability Theory
Assigning Probabilities: Subjective Method
• The subjective method of assigning probabilities is most appropriate
when we cannot realistically assume that the experimental outcomes
are equally likely and when little relevant data are available.

• When the subjective method is used to assign probabilities to the

experimental outcomes, we may use any information available, such as
our experience or intuition.

• After considering all available information, a probability value that

expresses our degree of belief (on a scale from 0 to 1) that the
experimental outcome will occur is specified.

• Because subjective probability expresses a person’s degree of belief, it is

personal. Using the subjective method, different people may assign
different probabilities to the same experimental outcome.
 3. Introduction to Probability Theory
Events and Their Probabilities
• An event is a collection of experimental outcomes. For example, in the
experiment of rolling a standard die, the sample space has six sample
points and is denoted S = {1, 2, 3, 4, 5, 6}. Now consider the event that
the number of dots shown on the upward face of the die is an even
number. The three sample points in this event are 2, 4, and 6. Using the
letter A to denote this event, we write A as a collection of sample points:

• If the probabilities of the sample points are defined, the probability

of an event is equal to the sum of the probabilities of the sample
points in the event. Hence, the probability of event A (an even
number of dots on the upward face of the die) is:
 3. Introduction to Probability Theory
Computing Probabilities: Complement of an Event
• For an event A, the complement of event A is the event consisting of all
sample points in sample space S that are not in A. The complement of A
is denoted by Ac. Figure 2.2 provides a diagram, known as a Venn
diagram, that illustrates the concept of a complement.
 3. Introduction to Probability Theory
Computing Probabilities: Complement of an Event
• In the diagram, the rectangular area represents the sample space for the
experiment and as such contains all possible sample points. The circle
represents event A and contains only the sample points that belong to A.
The remainder of the rectangle contains all sample points not in event A,
which by definition is the complement of A.

• Hence, the probability of the complement for an event A is given as:

• For example, suppose a vendor states a 0.90 probability that a shipment

will have no defective parts. Using the complement, we can conclude
that there is a 1 - 0.90 = 0.10 probability that the shipment will contain
some defective parts.
 3. Introduction to Probability Theory
Computing Probabilities: Union of Events
• For two events A and B, the union of events A and B is the event
containing all sample points belonging to A or B or both. The union is
denoted by A ∪ B. The Venn diagram shown in Figure 2.3 depicts the
union of events A and B; the shaded region contains all the sample
points in event A, as well as all the sample points in event B. The fact
that the circles overlap (or intersect) indicates that some sample points
are contained in both A and B.
 3. Introduction to Probability Theory
Computing Probabilities: Intersection of Events
• For two events A and B, the intersection of events A and B is the event
containing the sample points belonging to both A and B. The intersection
is denoted by A ∩ B. The Venn diagram depicting the intersection of the
two events is shown in Figure 2.4. The area where the two circles
overlap is the intersection; it contains the sample points that are in both
A and B.
 3. Introduction to Probability Theory
Computing Probabilities: Addition Law
• The addition law provides a way to compute the probability of event A
or B or both occurring. In other words, the addition law is used to
compute the probability of the union of two events, A ∪ B. The addition
law is formally stated as follows:

• To obtain an intuitive understanding of the addition law, note that the

first two terms in the addition law, P(A) + P(B), account for all the sample
points in A ∪ B.

• However, as the sample points in the intersection A ∩ B are in both A

and B, when we compute P(A) + P(B), we in effect are counting each of
the sample points in A ∩ B twice. We correct for this double counting by
subtracting P(A ∩ B).
 3. Introduction to Probability Theory
Computing Probabilities: Addition Law
• To apply the addition law, let us consider the following situations in a
college course in quantitative methods for decision making. Of 200
students taking the course, 160 passed the midterm examination and
140 passed the final examination; 124 students passed both exams. Let:

• This relative frequency information leads to the following probabilities:

• Assume that to obtain a passing grade, a student must pass at least one
of the two exams. Based on the data provided, what is the probability
that a student taking this course will pass at least one of the exams?
 3. Introduction to Probability Theory
Computing Probabilities: Addition Law
• Using the addition law for the events A and B, we have:

• Knowing the three probabilities on the right-hand side of this equation,

we obtain:

• This result indicates an 88% chance of a student passing the course (not
considering any other factors) because of the 0.88 probability of passing
at least one of the exams.
 3. Introduction to Probability Theory
Computing Probabilities: Addition Law
• Before proceeding, let us consider how the addition law is applied to
mutually exclusive events. Two or more events are said to be mutually
exclusive if the events do not have any sample points in common – that
is, there are no sample points in the intersection of the events.

• For two mutually exclusive events A and B, there is no intersection

between the events. Hence, P(A ∩ B) = 0. Therefore, for mutually
exclusive events, P(A ∪ B) = P(A) + P(B).
 3. Introduction to Probability Theory
Computing Probabilities: Multiplication Law
• The multiplication law can be used to find the probability of an
intersection of two events.

• For two events, A and B, that are independent (i.e. the occurrence of
one does not influence the occurrence of the other), the multiplication
law is given as follows:

• For two events, A and B, that are dependent (i.e. the occurrence of one
influences the occurrence of the other), the multiplication law is given as
follows:
 3. Introduction to Probability Theory
Computing Probabilities: Multiplication Law
• Thus, to compute the probability of the intersection of two independent
events, we simply multiply the corresponding probabilities.

• For example, a retail store manager knows from past experience that
40% of her customers use a credit card when making purchases. What is
the probability that the next two customers will both use a credit card?

• Let:

• The event of interest is A ∩ B. With no other information, it is reasonable

to assume A and B are independent events. Thus:
 3. Introduction to Probability Theory
Computing Probabilities: Multiplication Law
• On the other hand, to compute the probability of the intersection of two
dependent events, we multiply the probability that an event (e.g. A) occurs
given that another event (e.g. B) has occurred with the probability of the
other event (B) occurring.

• For example, suppose that a newspaper circulation department knows that

84% of its customers subscribe to the daily edition of the paper. Let D denote
the event that a customer subscribes to the daily edition; hence, P(D) = 0.84.

• In addition, the department knows that the conditional probability that a

customer who already holds a daily subscription also subscribes to the
Sunday edition (event S) is 0.75; that is, P(S | D) = 0.75. What is the
probability that a customer subscribes to both the daily and Sunday editions
of the newspaper? Using equation (2.9), we compute P(D ∩ S):

• This result tells us that 63% of the newspaper’s customers subscribe to both
the daily and Sunday editions.
 3. Introduction to Probability Theory
Computing Probabilities: Multiplication Law
• Note: mutually exclusive events are always dependent. We know this is
true because if one event occurs, it means the other event(s) will not
occur. Thus the occurrence of each event affects the occurrence of all
other events.

• However, since mutually exclusive events have no intersection points,

the multiplication law formula will always give a result of 0.

• For example, assume there are two mutually exclusive events A = strong
market demand and B = weak market demand. Also, P(A) = 0.40 and P(B)
= 0.60. We know that if A occurs, then B cannot occur (i.e. P(A|B) = 0)
and vice versa (i.e. P(B|A) = 0). Therefore, P(A ∩ B) must be equal to 0.

For mutually exclusive events, these terms = 0. Hence, the entire expression becomes 0.
Exercises
Exercises
Exercises
Exercises
Exercises
 4. Decision Making with Probabilities
The Expected Value Approach
• In many decision-making situations, we can obtain probability
assessments for the states of nature. When such probabilities are
available, we can use the expected value approach to identify the best
decision alternative.

• Let:

• Because one and only one of the N states of nature can occur, the
probabilities must satisfy two conditions:
 4. Decision Making with Probabilities
The Expected Value Approach
• The expected value (EV) of decision alternative di is defined as follows:

where:
• di = decision alternative i
• P(sj) = probability of the state of nature j occurring
• Vij = payoff of the decision alternative i for the state of nature j

• Hence, the expected value of a decision alternative (di) is the sum of

payoffs for the decision alternative for each state of nature (Vij) weighted
by the probability of each state of nature (sj) occurring.
 4. Decision Making with Probabilities
PDC Example
• Let us return to the PDC problem to see how the expected value
approach can be applied.

• PDC is optimistic about the potential for the luxury high-rise

condominium complex.

• Suppose that this optimism leads to an initial subjective probability

assessment of 0.8 that demand will be strong (s1) and a corresponding
probability of 0.2 that demand will be weak (s2).

• Thus, P(s1) = 0.8 and P(s2) = 0.2.

 4. Decision Making with Probabilities
PDC Example: EV
• Using the payoff values in Table 4.1 and equation (4.4), we compute the
expected value for each of the three decision alternatives (small,
medium, or large complex) as follows:

• Thus, using the expected value approach, we find that the large
condominium complex, with an expected value of $14.2 million, is the
recommended decision.
 4. Decision Making with Probabilities
PDC Example: EV
• The calculations required to identify the decision alternative with the
best expected value can be conveniently carried out on a decision tree.

• Recall that in a decision tree, squares are used to depict decision nodes
and circles are used to depict chance nodes.

• Figure 4.3 shows the decision tree for the PDC problem with state-of-
nature branch probabilities. Working backward through the decision
tree, we first compute the expected value at each chance node. That is,
at each chance node, we weight each possible payoff by its probability of
occurrence.

• By doing so, we obtain the expected values for nodes 2, 3, and 4, as

shown in Figure 4.4.
 4. Decision Making with Probabilities
PDC Example: Decision Tree

Work backwards from the branch levels to determine the EV of each decision alternative.
 4. Decision Making with Probabilities
PDC Example: Decision Tree Solution

When the final level is reached, we have the expected value of each decision alternative.
This is the solution of the decision tree using the expected value approach, which
suggests that alternative d3 has the highest expect value of $14.2 million.
 4. Decision Making with Probabilities
Expected Value of Perfect Information (EVPI)
• The expected value of perfect information (EVPI) is computed as follows:

• Note the role of the absolute value in equation (4.5). For minimization
problems, the expected value with perfect information is always less
than or equal to the expected value without perfect information.

• Hence, EVPI is the magnitude of the difference between EVwPI and

EVwoPI, or the absolute value of the difference.
 4. Decision Making with Probabilities
PDC Example: EVPI
• Suppose that PDC has the opportunity to conduct a market research
study that would help evaluate buyer interest in the condominium
project and provide information that management could use to improve
the probability assessments for the states of nature.

• To determine the potential value of this information, we begin by

supposing that the study could provide perfect information regarding the
states of nature, i.e. we assume for the moment that PDC could
determine with certainty, prior to making a decision, which state of
nature is going to occur.

• To make use of this perfect information, we will develop a decision

strategy that PDC should follow once it knows which state of nature will
occur. A decision strategy is simply a decision rule that specifies the
decision alternative to be selected after new information becomes
available.
 4. Decision Making with Probabilities
PDC Example: EVPI

• Note that, if PDC knew for sure that state of nature s1 would occur, the
best decision alternative would be d3, with a payoff of $20 million.
Similarly, if PDC knew for sure that state of nature s2 would occur, the
best decision alternative would be d1, with a payoff of $7 million. Thus,
we can state PDC’s optimal decision strategy when the perfect
information becomes available as follows:
 4. Decision Making with Probabilities
PDC Example: EVPI
• What is the expected value for this decision strategy using perfect
information?

• To compute the expected value with perfect information, we return to

the original probabilities for the states of nature: P(s1) = 0.8 and P(s2) =
0.2. Thus:
• There is a 0.8 probability that the perfect information will indicate state of nature
s1, and the resulting decision alternative d3 will provide a $20 million profit.
• Similarly, with a 0.2 probability for state of nature s2, the optimal decision
alternative d1 will provide a $7 million profit.

• Thus, from equation (4.4) the expected value of the decision strategy
that uses perfect information is 0.8(20) + 0.2(7) = 17.4. We refer to the
expected value of $17.4 million as the expected value with perfect
information (EVwPI).
 4. Decision Making with Probabilities
PDC Example: EVPI
• Earlier in this section we showed that the recommended decision using
the expected value approach is decision alternative d3, with an expected
value of $14.2 million. Because this decision recommendation and
expected value computation were made without the benefit of perfect
information, $14.2 million is referred to as the expected value without
perfect information (EVwoPI).

• The expected value with perfect information is $17.4 million, and the
expected value without perfect information is $14.2; therefore, the
expected value of the perfect information (EVPI) is $17.4 - $14.2 = $3.2
million.

• In other words, $3.2 million represents the additional expected value

that can be obtained if perfect information were available about the
states of nature.
 4. Decision Making with Probabilities
PDC Example: Calculating EVPI using the Regret Table
 Decision Tree Software

SilverDecisions: http://silverdecisions.pl/
Exercises
Exercises
Exercises
Exercises
Exercises
 5. Risk Analysis and Sensitivity Analysis
Risk Analysis
• Risk analysis helps the decision maker recognize the difference between
the expected value of a decision alternative and the payoff that may
actually occur.

• Risk analysis helps decision makers to look beyond the expected value of
a decision alternative and also take into consideration the level of risk
involved, in terms of variations in the potential payoffs.

• Risk analysis may be conducted for:

• The payoffs for each decision alternative by developing risk profile graphs for the
decision alternatives of interest
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Risk Analysis
• Let us demonstrate risk analysis and the construction of a risk profile by
returning to the PDC condominium construction project.

• Using the expected value approach, we identified the large

condominium complex (d3) as the best decision alternative. The
expected value of $14.2 million for d3 is based on a 0.8 probability of
obtaining a $20 million profit and a 0.2 probability of obtaining a $9
million loss.

• The 0.8 probability for the $20 million payoff and the 0.2 probability for
the -$9 million payoff provide the risk profile for the large complex
decision alternative. This risk profile is shown graphically in Figure 4.5.
5. Risk Analysis and Sensitivity Analysis
PDC Example: Risk Analysis
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Risk Analysis
• Sometimes a review of the risk profile associated with an optimal
decision alternative may cause the decision maker to choose another
decision alternative even though the expected value of the other
decision alternative is not as good.

• For example, the risk profile for the medium complex decision
alternative (d2) shows a 0.8 probability for a $14 million payoff and a 0.2
probability for a $5 million payoff. Because no probability of a loss is
associated with decision alternative d2, the medium complex decision
alternative would be judged less risky than the large complex decision
alternative.

• As a result, a decision maker might prefer the less risky medium complex
decision alternative even though it has an expected value of $2 million
less than the large complex decision alternative.
 5. Risk Analysis and Sensitivity Analysis
Sensitivity Analysis
• Sensitivity analysis also helps the decision maker by describing how changes
in the state-of-nature probabilities and/or changes in the payoffs affect the
recommended decision alternative.

• Sensitivity analysis can be used to determine how changes in the

probabilities for the states of nature or changes in the payoffs affect the
recommended decision alternative. In many cases, the probabilities for the
states of nature and the payoffs are based on subjective assessments.
Sensitivity analysis helps the decision maker understand which of these
inputs are critical to the choice of the best decision alternative.

• If a small change in the value of one of the inputs causes a change in the
recommended decision alternative, the solution to the decision analysis problem is
sensitive to that particular input. Extra effort and care should be taken to make sure
the input value is as accurate as possible.

• On the other hand, if a modest-to-large change in the value of one of the inputs does
not cause a change in the recommended decision alternative, the solution to the
decision analysis problem is not sensitive to that particular input. No extra time or
effort would be needed to refine the estimated input value.
 5. Risk Analysis and Sensitivity Analysis
Sensitivity Analysis
• Sensitivity analysis may be conducted for:

• The probability estimates of the states of nature by plotting expected value

graphs as a function of the probability of a state of nature occurring (for two
states of nature only)

• The payoffs for each decision alternative under each state of nature by
calculating the range of payoffs within which the particular decision alternative
remains optimal (for two states of nature only)

• Sensitivity analysis may be conducted manually or with the help of

decision analysis software, such as SilverDecisions
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis by Changing Values
• One approach to sensitivity analysis is to select different values for the
probabilities of the states of nature and the payoffs and then resolve the
decision analysis problem. If the recommended decision alternative
changes, we know that the solution is sensitive to the changes made.

• For example, suppose that in the PDC problem the probability for a
strong demand is revised to 0.2 and the probability for a weak demand is
revised to 0.8. Would the recommended decision alternative change?

• Using P(s1) = 0.2, P(s2) = 0.8, and equation (4.4), the revised expected
values for the three decision alternatives are
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis by Changing Values
• With these probability assessments, the recommended decision
alternative is to construct a small condominium complex (d1), with an
expected value of $7.2 million. The probability of strong demand is only
0.2, so constructing the large condominium complex (d3) is the least
preferred alternative, with an expected value of -$3.2 million (a loss).

• Thus, when the probability of strong demand is large, PDC should build
the large complex; when the probability of strong demand is small, PDC
should build the small complex. Obviously, we could continue to modify
the probabilities of the states of nature and learn even more about how
changes in the probabilities affect the recommended decision
alternative.

• The drawback of this approach for sensitivity analysis is the numerous

calculations required to evaluate the effect of several possible changes in
the state-of-nature probabilities.
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Probabilities of States of Nature

• For the special case of two states of nature, a graphical procedure can
be used to determine how changes for the probabilities of the states of
nature affect the recommended decision alternative. To demonstrate
this procedure, we let p denote the probability of state of nature s1; that
is, P(s1) = p.

• With only two states of nature in the PDC problem, the probability of
state of nature s2 is:

• Using equation (4.4) and the payoff values in Table 4.1, we determine
the expected value for decision alternative d1 as follows:
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Probabilities of States of Nature

• Repeating the expected value computations for decision alternatives d2

and d3, we obtain expressions for the expected value of each decision
alternative as a function of p:

• Thus, we have developed three equations that show the expected value
of the three decision alternatives as a function of the probability of state
of nature s1.

• We continue by developing a graph with values of p on the horizontal

axis and the associated EVs on the vertical axis. Because equations (4.6),
(4.7), and (4.8) are linear equations, we know that the graph of each
equation is a straight line.
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Probabilities of States of Nature

• For each equation, we can obtain the line by identifying two points that
satisfy the equation and drawing a line through the points.

• For instance, if we let p=0 in equation (4.6), EV(d1) = 7. Then, letting p=

1, EV(d1) = 8. Connecting these two points, (0,7) and (1,8), provides the
line labeled EV(d1) in Figure 4.6. Similarly, we obtain the lines labeled
EV(d2) and EV(d3); these lines are the graphs of equations (4.7) and (4.8),
respectively.

• Figure 4.6 shows how the recommended decision changes as p, the

probability of the strong demand state of nature (s1), changes.
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Probabilities of States of Nature
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Probabilities of States of Nature

• Note that for small values of p, decision alternative d1 (small complex)

provides the largest expected value and is thus the recommended
decision. When the value of p increases to a certain point, decision
alternative d2 (medium complex) provides the largest expected value
and is the recommended decision. Finally, for large values of p, decision
alternative d3 (large complex) becomes the recommended decision.

• The value of p for which the expected values of d1 and d2 are equal is the
value of p corresponding to the intersection of the EV(d1) and the EV(d2)
lines. To determine this value, we set EV(d1) = EV(d2) and solve for the
value of p:
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Probabilities of States of Nature

• Hence, when p = 0.25, decision alternatives d1 and d2 provide the same

expected value. Repeating this calculation for the value of p
corresponding to the intersection of the EV(d2) and EV(d3) lines, we
obtain p = 0.70.

• Using Figure 4.6, we can conclude that:

• Decision alternative d1 provides the largest expected value for p < 0.25
• Decision alternative d2 provides the largest expected value for 0.25 < p < 0.70
• Decision alternative d3 provides the largest expected value for p < 0.70

• Because p is the probability of state of nature s1 and (1 - p) is the

probability of state of nature s2, we now have the sensitivity analysis
information that tells us how changes in the state-of-nature probabilities
affect the recommended decision alternative. Note that these results
assume that the payoffs of the decision alternatives remain unchanged.
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Values of Payoffs

• Sensitivity analysis calculations can also be made for the values of the
payoffs. In the original PDC problem, the expected values for the three
decision alternatives were as follows: EV(d1) = 7.8, EV(d2) = 12.2, and
EV(d3) = 14.2.

• Decision alternative d3 (large complex) was recommended. Note that

decision alternative d2 with EV(d2) = 12.2 was the second best decision
alternative. Decision alternative d3 will remain the optimal decision
alternative as long as EV(d3) is greater than or equal to the expected
value of the second best decision alternative. Thus, decision alternative
d3 will remain the optimal decision alternative as long as:
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Values of Payoffs

• Let:

• Using P(s1) = 0.8 and P(s2) = 0.2, the general expression for EV(d3) is:
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Values of Payoffs

• Assuming that the payoff for d3 stays at its original value of -$9 million
when demand is weak, the large complex decision alternative will
remain optimal as long as:

• Solving for S, we have:

• Recall that when demand is strong, decision alternative d3 has an

estimated payoff of $20 million. The preceding calculation shows that
decision alternative d3 will remain optimal as long as the payoff for d3
when demand is strong is at least $17.5 million.
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Values of Payoffs

• Assuming that the payoff for d3 when demand is strong stays at its
original value of $20 million, we can make a similar calculation to learn
how sensitive the optimal solution is with regard to the payoff for d3
when demand is weak. Returning to the EV calculation of equation
(4.10), we know that the large complex decision alternative will remain
optimal as long as:

• Solving for W, we have:

• Recall that when demand is weak, decision alternative d3 has an

estimated payoff of -$9 million. The preceding calculation shows that
decision alternative d3 will remain optimal as long as the payoff for d3
when demand is weak is at least -$19 million.
 5. Risk Analysis and Sensitivity Analysis
PDC Example: Sensitivity Analysis for Values of Payoffs

• Based on this sensitivity analysis, we see that the payoffs for the large
complex decision alternative (d3) could vary considerably, and d3 would
remain the recommended decision alternative. Thus, we conclude that
the optimal solution for the PDC decision problem is not particularly
sensitive to the payoffs for the large complex decision alternative.

• We note, however, that this sensitivity analysis has been conducted

based on only one change at a time. That is, only one payoff was
changed and the probabilities for the states of nature remained
unchanged at P(s1) = 0.8 and P(s2) = 0.2.

• Note that similar sensitivity analysis calculations can be made for the
payoffs associated with the small complex decision alternative d1 and
the medium complex decision alternative d2.
Exercises
Exercises
Exercises
 Decision Analysis vs.
Other DM Tools & Techniques
Qualitative Quantitative

Qualitative T&Ts AHP Decision Analysis

(i) Payoffs of decision alternatives
Data Minimal (varies based on Pairwise comparisons of under all possible states of nature
Required technique) alternatives and/or criteria (ii) Probabilities of states of nature
(if possible)
Systematic conversion of Calculating optimal payoffs and/or
Analytical Subjective
pairwise comparisons into expected values of decision
Method assessments
preference vectors alternatives
Type of Subjective (qualitative Subjective (quantified Objective (expected values of
Output factors) preferences) decision alternatives)
No. of
Single/Multiple Multiple Single (e.g. profit or cost)
Criteria
Making sense of available Choosing a decision
Choosing a decision alternative
Suitable information, generating and alternative that best
that provides the highest (or
for discussing ideas, identifying satisfies the preferences of
lowest) expected value
decision alternatives the decision maker(s)