Decision Analysis

Chapter 12

Copyright © 2016 Pearson Education, Inc. 12-1

Chapter Topics

■ Components of Decision Making

■ Decision Making without Known Probabilities
■ Decision Making with Known Probabilities

Copyright © 2016 Pearson Education, Inc. 12-2

Decision Analysis
Overview

 Making decisions is an important part of life and doing

business.
 Good decision-making requires information collection and
scientific theory to support.
 Uncertainties often exist which complicates decision making
 Two types of situations regarding uncertainties:
 Probabilities are known about occurrences of future events
 Probabilities are unknown about occurrences of future
events
Copyright © 2016 Pearson Education, Inc. 12-3
Components of Decision Making:
State of Nature
■ A state of nature:
An event that may occur in the future, which would directly
impact the outcome of your decision.
■ At the time a decision is made, the decision maker is uncertain
which state of nature will occur in the future and also has no
control over them
Example:
State of nature: US economy for the next five years (good or bad?)
Decision: invest in US stock market or bond (now)?

Copyright © 2016 Pearson Education, Inc. 12-4

Components of Decision Making:
Payoff Table
■ A payoff table: is a means of organizing a decision situation,
presenting the payoffs from different decisions given the various
states of nature.

State of Nature
Decision Good economy Bad economy
Stock $10,000 -$10,000
Bond $1,000 -$1,000

Example of a payoff table

Copyright © 2016 Pearson Education, Inc. 12-5

Decision Analysis
Decision Making without Known Probabilities

Decision to make:
which real estate
investment option
should she choose?

State of nature:
economic condition
in the future (good
or poor), which will
determine profit for
each option

Figure 12.1 Decision

situation with real estate
investment alternatives

Copyright © 2016 Pearson Education, Inc. 12-6

Decision Analysis
Decision Making without Known Probabilities

Table 12.2 Payoff table for the real estate investments

Decision-Making Criteria
maximax maximin
minimax regret Hurwicz equal likelihood
Copyright © 2016 Pearson Education, Inc. 12-7
Decision Making without Probabilities
Maximax Criterion
In the maximax criterion, a decision maker selects a decision
that results in the maximum of maximum payoffs; an optimistic
criterion.

Table 12.3 Payoff table illustrating a maximax decision

Under maximax criterion, the decision maker will select office building
as his/her investment option.
Copyright © 2016 Pearson Education, Inc. 12-8
Decision Making without Probabilities
Maximin Criterion
In the maximin criterion, the decision maker selects a decision
that results in the maximum of the minimum payoffs; a
pessimistic criterion.

Table 12.4 Payoff table illustrating a maximin decision

Under maximin criterion, the decision maker will choose apartment

building as his/her investment option.
Copyright © 2016 Pearson Education, Inc. 12-9
Decision Making without Probabilities
Hurwicz Criterion
 The Hurwicz criterion is a balanced one between the maximax
and maximin criteria.
 A coefficient of optimism, , is a measure of the decision
maker’s optimism.
 For each decision option, multiplies its best payoff by  and the
worst payoff by 1- , and then the highest total payoff option is
selected.

Copyright © 2016 Pearson Education, Inc. 12-10

Decision Making without Probabilities
Hurwicz Criterion

Assume
α=.4

Decision Payoff Values

Apartment building $50,000(.4) + 30,000(.6) = 38,000

Office building $100,000(.4) - 40,000(.6) = 16,000

Warehouse $30,000(.4) + 10,000(.6) = 18,000
Under Hurwicz criterion with α=.4, choose apartment building
Copyright © 2016 Pearson Education, Inc. 12-11
Decision Making without Probabilities
Minimax Regret Criterion
 Regret: is the difference in payoff when compare one decision
against the best decision under the same state of nature.
If the economic
condition is good, the
possible maximum
Table 12.2 payoff is $100,000; if
Payoff table for the economic
the real estate condition is poor, the
investments possible maximum
payoff is $30,000

Table 12.5
100, 000  50, 000  50, 000
Regret table

30, 000  (40, 000)  70, 000

30, 000  10, 000  20, 000
100, 000  30, 000  70, 000
Copyright © 2016 Pearson Education, Inc. 12-12
Decision Making without Probabilities
Minimax Regret Criterion
 The decision maker first computes regrets for all decision options
under each state of nature to build a regret table.
 Then, he/she identifies the maximum regret for each decision option.
 Finally, select the decision option that minimizes the maximum
regret.

Maximum regret
Maximum regret

Maximum regret Under minimax regret

criterion, choose apartment
Table 12.5 Regret table
building
Copyright © 2016 Pearson Education, Inc. 12-13
Decision Making without Probabilities
Equal Likelihood Criterion

The equal likelihood criterion multiplies the decision payoff

under each state of nature by an equal weight or probability
(like .5 for two states of nature), thus assuming that each state
of nature is equally likely to occur. Select the one that has the
highest total payoff.

Copyright © 2016 Pearson Education, Inc. 12-14

Decision Making without Probabilities
Equal Likelihood Criterion

Decision Values
Apartment building $50,000(.5) + 30,000(.5) = 40,000

Office building $100,000(.5) - 40,000(.5) = 30,000

Warehouse $30,000(.5) + 10,000(.5) = 20,000

Under equal likelihood criterion, choose apartment building

Copyright © 2016 Pearson Education, Inc. 12-15
Decision Making without Probabilities
Summary of Criteria Results
■ The appropriate criterion to use is dependent on the “risk”
personality and philosophy of the decision maker.
Criterion Decision (Purchase)
Maximax Office building
Maximin Apartment building
Minimax regret Apartment building
Hurwicz Apartment building
Equal likelihood Apartment building

Warehouse is never a chosen option for the decision maker, why?

Copyright © 2016 Pearson Education, Inc. 12-16

Decision Making without Probabilities
Dominant Decision and Dominated Decision

A dominant decision is one that has a better payoff than another

decision under ALL states of nature.
A dominated decision is one that has a less payoff than another
decision under ALL states of nature.
Between apartment and warehouse, choosing warehouse is a dominated decision and
choosing apartment is dominant decision.
Copyright © 2016 Pearson Education, Inc. 12-17
Decision Making with Known Probabilities
Expected Value

Table 12.7 Payoff table with probabilities for states of nature

 Expected value (payoff) : multiply each decision payoff under each

state of nature by the probability of its occurrence and then totals.
EV(Apartment) = $50,000(.6) + 30,000(.4) = $42,000
choose office
EV(Office) = $100,000(.6) - 40,000(.4) = $44,000
building (highest
EV(Warehouse) = $30,000(.6) + 10,000(.4) = $22,000
EV)
Copyright © 2016 Pearson Education, Inc. 12-18
Decision Making with Known Probabilities
Expected Opportunity Loss

Table 12.8 Regret table with probabilities for states of nature

■ The expected opportunity loss (EOL) is the expected value of the
regret for one decision.
EOL(Apartment) = $50,000(.6) + 0(.4) = $30,000
EOL(Office) = $0(.6) + 70,000(.4) = $28,000 choose office building
EOL(Warehouse) = $70,000(.6) + 20,000(.4) = $50,000 (lowest EOL)
■ Maximizing the expected value and minimizing the expected
opportunity loss, these two criteria will result in the same decision
(choose office building)
Copyright © 2016 Pearson Education, Inc. 12-19
Decision Making with Known Probabilities
Expected Value of Perfect Information

■ Information has values for decision makers. Perfect information

would significantly improve decision making.
■ The expected value of perfect information (EVPI) is the
maximum amount a decision maker would pay for additional
information.
EVPI = EV (with the perfect information) ‒
EV (without perfect information)

■ EVPI also equals the expected opportunity loss (EOL) of

choosing the best decision (office building), $28,000.

Copyright © 2016 Pearson Education, Inc. 12-20

Decision Making with Known Probabilities
EVPI Example (1 of 2)

Table 12.9 Payoff table for decisions with perfect information

Copyright © 2016 Pearson Education, Inc. 12-21

Decision Making with Known Probabilities
EVPI Example (2 of 2)

■ Expected payoff with perfect information (would choose either

office or apartment depending known economy condition):
EV=$100,000(.60) + 30,000(.40) = $72,000
■ Expected payoff without perfect information (would choose
office building):
EV(office) = $100,000(.60) - 40,000(.40) = $44,000
■ Expected value of perfect information
EVPI = $72,000 - 44,000 = $28,000

EOL(office) = $0(.60) + 70,000(.4) = $28,000

Copyright © 2016 Pearson Education, Inc. 12-22

Decision Making with Known Probabilities
Decision Trees (1 of 3)
 Another useful technique for analyzing a decision situation with
known probabilities is using decision tree.
 A decision tree is a diagram consisting of decision nodes
(represented as squares), probability nodes (circles), and decision
alternatives (branches).

Figure 12.2 Decision tree for real estate investment example

Copyright © 2016 Pearson Education, Inc. 12-23
Decision Making with Known Probabilities
Decision Trees (2 of 3)

Figure 12.2 Decision

tree for real estate
investment example

Compute the expected value (EV) for each decision option node, then select the
one with the largest EV.
EV(node 2) = .60($50,000) + .40(30,000) = $42,000
EV(node 3) = .60($100,000) + .40(-40,000) = $44,000 = EV(node 1)
EV(node 4) = .60($30,000) + .40(10,000) = $22,000
Copyright © 2016 Pearson Education, Inc. 12-24
Decision Making with Known Probabilities
Decision Trees (3 of 3)

$44,000

Figure 12.3 Decision tree with expected value at probability nodes

Copyright © 2016 Pearson Education, Inc. 12-25
Decision Making with Known Probabilities
Sequential Decision Trees (1 of 4)
■ A sequential decision tree is used to illustrate a situation
requiring a series of decisions.

■ Used where a payoff table, limited to a single decision, cannot

be used.

■ The next slide shows the real estate investment example

modified to encompass a ten-year period in which several
decisions must be made.

Copyright © 2016 Pearson Education, Inc. 12-26

Decision Making with Known Probabilities
Sequential Decision Trees (2 of 4)

Figure 12.4 Sequential decision tree

Copyright © 2016 Pearson Education, Inc. 12-27
Decision Making with Known Probabilities
Sequential Decision Trees (3 of 4)
$1,290,000 EV(node 2)
 2,000,000  .60  225,000  .40
 $1, 290,000
$2,540,000
$1,160,000 EV(node 6)
$1,740,000  3,000,000  .80  700,000  .20
 $2,540,000
$1,360,000
EV(node 7)
$1,390,000
 2,300,000  .30  1,000,000  .70
$790,000  $1,390,000

EV(node 4)  Max{EV(node 6)-$800,000, $450,000}  Max{$2,540,000-$800,000, $450,000}=$1,740,000

EV(node 5)  Max{EV(node 7)-$600,000, $210,000}  Max{$1,390,000-$600,000, $210,000}=$790,000
EV(node 3)=1,740,000  .60+790,000  .40=$1,360,000 decision is
EV(node 1)=Max{EV(node 2)-$800,000, EV(node 3)-$200,000}
to buy land
=Max{$1,290,000-$800,000, $1,360,000-$200,000}=Max{$490,000, $1,160,000}=$1,160,000
Copyright © 2016 Pearson Education, Inc. 12-28
Decision Making with Known Probabilities
Sequential Decision Trees (4 of 4)

Figure 12.5 Sequential decision tree with nodal expected values

Copyright © 2016 Pearson Education, Inc. 12-29