CHAPTER 5
SUPPLEMENT:
Decision
Analysis
Slides prepared by
Romulus Cismaru
University of Regina
Introduction
Decision Theory Elements
A set of possible future conditions
exists that will have a bearing on the
results of the decision
A list of alternatives for the manager to
choose from
A known payoff for each alternative
under each possible future condition
5S-1
Introduction
5S-2
Fundamentals of
Decision Theory - continued
Decision Theory Process
1. Identify possible future conditions called
states of nature
2. Develop a list of possible alternatives, one
of which may be to do nothing
3. Determine the payoff associated with each
alternative for every future condition
4. Estimate the likelihood for each possible
future condition
5. Evaluate alternatives according to some
decision criterion
Terms:
Alternative: course of action or choice
State of nature: an occurrence over
which the decision maker has no control
5S-3
5S-4
Introduction
Fundamentals of
Decision Theory
Payoff table: table showing the payoffs for each alternative of a
decision in every possible state of nature, of a random variable
The three types of decision models:
Alternative
Decision making under uncertainty
Small
Facility
Medium
Facility
Large
Facility
Decision making under risk
Decision making under certainty
Possible Future Demand
Low
Moderate
High
(p=0.3)
(p=0.5)
(p=0.2)
10*
10
10
7
12
12
(4)
16
*Present value in $ millions
5S-5
5S-6
We calculate expected value
Alternative
Small
Facility
Alternative
Possible Future Demand
Low
Moderate
High
(p=0.3)
(p=0.5)
(p=0.2)
10*
10
10
Large
Facility
Medium
Facility
Low
(p=0.3)
7
Moderate
(p=0.5)
12
High
(p=0.2)
12
High
(p=0.2)
16
=3
Previously, we have
EVSMALL = 0.3 (10 ) + 0.5 (10 ) + 0.2 (10 ) = 10
EVMEDIUM = 0 .3 (7 ) + 0.5 (12 ) + 0.2 (12 ) = 10 .5
EVMEDIUM = 0.3 (7 ) + 0.5 (12 ) + 0.2 (12)
= 10.5
Moderate
(p=0.5)
2
EVLARGE = 0.3 ( 4 ) + 0.5 (2 ) + 0.2 (16 )
EVSMALL = 0.3 (10 ) + 0.5 (10 ) + 0.2 (10 ) = 10
Alternative
Low
(p=0.3)
(4)
We will select the medium facility
5S-7
5S-8
Decision Trees
Decision Trees
Symbols used in decision tree:
Decision Tree:
A schematic representation of the
decision variables, random variables,
and payoffs
Composed of a number of nodes that
have branches
A decision point from which one of several alternatives
may be selected
A chance event node out of which various state of nature
will occur
5S-9
Example S2--A manager must decide on the size of a video
arcade to construct. The manager has narrowed the choices to
two: large or small. Information has been collected on payoffs,
and the following decision tree has been constructed. Analyze
the decision tree and determine which initial alternative (build
small or build large) should be chosen in order to maximize
expected monetary value.
Format of a Decision Tree
Decision Point
Chance Event
5S-10
Payoff 1
Payoff 2
2
Payoff 3
1
Possible second decision
Payoff 4
Payoff 5
Payoff 6
5S-11
5S-12
$55*0.6+$40*0.4=49
$55
$50
$50*0.4+$70*0.6=62
5S-13
Expected Value With Perfect Information
(EV|PI)
Expected Value of Perfect Information
Expected value of perfect information: the
difference between the expected payoff under
certainty and the expected payoff under risk
Expected value of
perfect information
Expected payoff
under certainty
5S-14
Expected value under certainty
Expected payoff
i =n
under risk
= Best outcome for the ith state of nature *P( S i )
i=1
EVPI places an upper bound on what one would pay for
additional information
where P(Si ) = Probability of the ith state of nature
and i = 1 to n, the number of states of nature
5S-15
5S-16
Expected Value of Perfect Information
Alternative
Small
Facility
Medium
Facility
Large
Facility
EVPI = Expected value under Certainty maximum Expected Monetary Value (EMV)
Possible Future Demand
Low
Moderate
High
(p=0.3)
(p=0.5)
(p=0.2)
10
10
10
Expected
monetary
value
10
12
12
10.5
(4)
16
With perfect info, our expected return will be:
10*0.3+12*0.5+16*0.2=12.2
Maximum EMV
EVPI=12.2-10.5=1.7
5S-17
5S-18
State of Nature
Sensitivity Analysis
Sensitivity Analysis: determining the range of
probability for which an alternative has the best
expected payoff
Alternative
Alternative
#2
(unfavourable
market)
12
16
12
We do not know the exact probability that #1 and #2 will occur
In the following discussion, we use P(1) represent probability of
#1, P(2) represent probability of #2.
P(1)+P(2)=1
5S-19
#1 Payoff
#2 Payoff
16
14
12
10
8
6
4
2
0
16
14
12
10
8
6
4
2
0
#2 Payoff
If P(2)=0, P(1)=1
Total payoff= 4*1+12*0=4
If P(2)=1, P(1)=0
Total payoff= 4*0+12*1=12
Probability of #2
#1 Payoff
#2 Payoff
16
14
12
10
8
6
4
2
0
5S-21
5S-22
State of Nature
#1(favourable
market)
Alternative
#1 Payoff
#2 Payoff
16
14
12
10
8
6
4
2
0
16
14
12
10
8
6
4
2
0
If P(2)=1, P(1)=0
EV= 4*0+12*1=12
#1(favourable market) #2 (unfavourable
market)
16
14
12
10
8
6
4
2
0
1
5S-20
State of Nature
16
16
14
12
10
8
6
4
2
0
0
Probability of #2
If P(2)=0, P(1)=1
EV= 4*1+12*0=4
Expected value (EV) =4*P(1)+12*P(2) ; P(1)+P(2)=1
P(1)=1-P(2);
EV=4*(1-P(2))+12*P(2)
=4-4*P(2)+12*P(2)
=4+8*P(2)
12
Probability of #2
Alternative
#1 Payoff
#1(favourable
market)
16
14
12
10
8
6
4
2
0
#2 (unfavourable market)
Expected value =4*P(1)+12*P(2) ; P(1)+P(2)=1
State of Nature
#1(favourable market)
Probability of #2
1
5S-23
#2
(unfavourable
market)
12
16
12
#1 Payoff
#2 Payoff
16
14
12
10
8
6
4
2
0
A
C
Probability of #2
16
14
12
10
8
6
4
2
0
1
5S-24
#1 Payoff
#1 Payoff
#2 Payoff
16
14
12
10
8
6
4
2
0
A
C
B best
C best
A best
?
?
Probability of #2
16
14
12
10
8
6
4
2
0
A
C
B best
#2 Payoff
16
14
12
10
8
6
4
2
0
C best
0.4
Line B: EV=16-14*P(2)
Line C: EV=12-4*P(2)
We have: 16-14*P(2)=12-4*P(2)
4=10*P(2)
P(2)=0.4
5S-25
#1 Payoff
A
C
B best
5S-26
#2 Payoff
16
14
12
10
8
6
4
2
0
A best
16
14
12
10
8
6
4
2
0
1
C best
0.4
Line A: EV=4+8*P(2)
Line C: EV=12-4*P(2)
A best
0.67
16
14
12
10
8
6
4
2
0
1
Homework
Problem 1, 2, 4, 5, 7, 9
We have: 4+8*P(2)=12-4*P(2)
12*P(2)=8
P(2)=0.67
5S-27
5S-28
