4 Decision Theory

Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT 1

 Outline
◆ Decision Theory Elements
◆ Decision Theory Environment
◆ Decision Making under Certainty
◆ Decision Making under Uncertainty
a) Maximin,
b) Maximax,
c) Laplace, and
d) Minimax regret.
◆ Decision Making Under Risk

2
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Learning Objectives
When you complete this lecture you should
be able to:
1. Describe the different environments under which
operations decisions are made.
2. Describe and use techniques that apply to decision
making under uncertainty.
3. Describe and use the expected value approach.

3
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Decision Theory
An Analytic and Systematic Approach to the
Study of Decision Making
Good decisions: Bad decisions:
◆ based on logic ◆ not based on logic

◆ consider all available data and ◆ do not consider all available

possible alternatives data and possible alternatives

◆ employ a quantitative ◆ do not employ a quantitative

approach approach

A good decision may A bad decision may

occasionally result in an occasionally result in a
unexpected outcome; it is good outcome if you are
still a good decision if lucky; BUT it is still a bad
made properly decision 4
 Decision Theory Elements
A list of
alternatives for
the manager to
choose from

A known A set of
payoff for possible
each future
alternative conditions

5
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Decision Theory Environment
◆Operations management decision environments are
classified according to the degree of certainty
present. There are three basic categories:
1. Certainty means that relevant parameters such
as costs, capacity, and demand have known
values.
2. Uncertainty means that it is impossible to assess
the likelihood of various possible future events.
3. Risk means that certain parameters have
probabilistic outcomes.

6
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 1. Decision Making under
Certainty
◆The decision is usually relatively straight
forward.
◆ Simply choose the alternative that has the
best payoff under that state of nature.

7
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Example
◆Determine the best alternative in the preceding
payoff table for each of the cases: It is known with
certainty that demand will be: (a) low, (b)
moderate, (c) high.
$ Low demand $ Med demand $ High demand
Small facility 10 10 10
Medium facility 7 12 12
Large facility -4 2 16

$ Low demand $ Med demand $ High demand

Small facility 10 10 10
Medium facility 7 12 12
Large facility -4 2 16

8
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Example
◆Determine the best alternative in the preceding
payoff table for each of the cases: It is known with
certainty that demand will be: (a) low, (b)
moderate, (c) high.
$ Low demand $ Med demand $ High demand
Small facility 10 10 10
Medium facility 7 12 12
Large facility -4 2 16

$ Low demand $ Med demand $ High demand

Small facility 10 10 10
Medium facility 7 12 12
Large facility -4 2 16
9
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 2. Decision Making under
Uncertainty
◆At the opposite extreme is complete uncertainty:
no information is available on how likely the
various states of nature are.
◆Under those conditions, four possible decision
criteria are:
a) Maximin,
b) Maximax,
c) Laplace, and
d) Minimax regret.

10
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 a) Maximin
Pessimistic Decision
◆The maximin approach is a pessimistic one,
because it takes into account only the worst
possible outcome for each alternative.
◆First, determines the worst possible payoff for
each alternative.
◆Then, select the maximum of these minimum
payoffs “best worst”.
◆The actual outcome may not be as bad as that,
but this approach establishes a “guaranteed
minimum”.
11
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Maximin Example
$ Low $ Med $ High MaxiMin
demand demand demand

Small facility 10 10 10 10
Medium facility 7 12 12 7
Large facility -4 2 16 -4

Since $ 10 is the maximum,

Then choose to build the small facility

12
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 b) Maximax
Optimistic Decision
◆The maximax approach is optimistic decision, “go for
it ” strategy. It selects the decision that results in the
maximum of the maximum payoffs.
◆Decision maker assumes that the most favorable
state of nature for each decision alternative will
occur.
◆If the payoff table was in terms of profits, the
decision with the highest profit would be chosen.
◆If the payoff table was in terms of costs, the
decision with the lowest cost would be chosen.
13
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Maximax Example
$ Low $ Med $ High
Maximax
demand demand demand

Small facility 10 10 10 10

Medium facility 7 12 12 12

Large facility -4 2 16 16
Since $ 16 is the maximum,
then choose to build the large facility

14
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 c) Laplace
◆Assumes that all states of nature are equally likely
to occur.
◆Calculate the average payoff for each alternative
and select the alternative with the maximum
number.
◆Average payoff: the sum of all payoffs divided by
the number of states of nature.
◆Select the decision that gives the highest average
payoff.

15
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Laplace Example
(10+10+10)/3
$ Low $ Med $ High
Average
demand demand demand
Small facility 10 10 10 10

Medium facility 7 12 12 10.33

Large facility -4 2 16 4.67

Since $ 10.33 is the highest average,

then choose to build the medium facility

16
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 d) Minimax Regret
◆ This approach seeks to made the decision on the least
regret for making that choice.
◆ Regret is the amount of loss due to making an
incorrect decision (opportunity cost)
1. Choose the alternative with the best payoff.
2. Establish the difference between the payoff that is
given and the best payoff for each state of nature.
3. Choose the alternative that minimizes the maximum
regret associated with each alternative by determining
the maximum regret for each alternative.
4. Pick the alternative with the minimum number
(lowest regret).
17
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Minimax Regret Example
Maximum outcome 10 12 16 Step 1
$ Low $ Med $ High
demand demand demand

Small facility 10 10 10

Medium facility 7 12 12 Step 3

Large facility -4 2 16

$ Low $ Med $ High Minimax

demand demand demand Regret
10-10= 12-10= 16-10=
Small facility
0 2 6 6
Medium
facility
10-7=
3
12-12=
0
16-12=
4 4 lowest
regret
10-(-4)= 12-2= 16-16=
Large facility
14 2 0 14
18
Step 2 Step 4
 3. Decision Making
Under Risk
◆ Between the two extremes of certainty and uncertainty
lies the case of risk.
◆ The probability of occurrence for each state of nature is
known (these probabilities must add to 1.00).
◆ Decision makers should know the probability of
occurrence for each possible outcome in an attempt to
maximize the expected payoff.
◆ A widely used approach under such situations is the
expected monetary value (EMV).

20
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Expected Monetary Value (EMV)
◆ Expected monetary value is the best expected value
among the alternatives.
◆ The expected value is the sum of the payoffs for an
alternative where each payoff is weighted by the
probability for the relevant state of nature.
EMV (alternative i) =
(payoff of 1st state of nature)X(probability of 1st state of
nature) + (payoff of 2nd state of nature) X (probability of
2nd state of nature) + . . . + (payoff of last state of nature)
X (probability of last state of nature)
◆ The expected value is computed for each alternative, and
the one with the highest expected value is selected.
21
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 EMV Example
$ Low demand $ Med demand $ High demand
0.30 0.50 0.20
Small facility 10 10 10
Medium facility 7 12 12
Large facility -4 2 16

$ Low $ Med $ High

EMV
demand demand demand
3+5+2=
Small facility 10(0.30) = 3 10(0.50) = 5 10(0.20)= 2
10

Medium facility 7(0.30) =2.1 12(0.50)=6 12(0.20)=2.4 10.5

Large facility (-4)(0.30)=-1.2 2(0.50)=1 16(0.20)=3.2 3

highest expected value is selected (10.5)

So, select Medium Facility
22
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 EMV Example
◆ An investor is going to purchase one of three types of
real estate: an apartment building, an office building, or
a warehouse. The two future states of nature that will
determine how much profit the investor will make are
either good economic conditions or bad economic
conditions. The profits that will result from each decision
given these two states of nature are summarized below:

States of Nature
Purchase Good Economy Poor Economy
Decision p = 0.6 p = 0.4
Apt Bldg $ 50,000 $ 30,000
Office Bldg 100,000 -40,000
Warehouse 30,000 10,000
23
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 EMV Example
50,000 * 0.6 = 30,000 + 12,000 =
30,000 42,000

Purchase Good Economy Poor Economy

Decision p = 0.6 p = 0.4 EMV
Apt Bldg $ 50,000 $ 30,000 ?
Office Bldg 100,000 -40,000 ?
Warehouse 30,000 10,000 ?
States of Nature
Purchase Good Economy Poor Economy
Decision p = 0.6 p = 0.4 EMV
Apt Bldg $ 30,000 $ 12,000 $ 42,000
Office Bldg 60,000 -16,000 $ 44,000
Warehouse 18,000 4,000 $ 22,000
Select highest expected value = $44000
So, select Office Building 24
 Expected value of
perfect information (EVPI)
◆In certain situations, it is possible to determine which
state of nature will actually occur in the future.
◆For instance, the choice of location for a restaurant
may weigh greatly on whether a new highway will be
constructed or not.
◆Or information about consumer preferences might
come from market research, or any additional
information about product that came from product
testing.

25
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Expected value of
perfect information (EVPI)
◆ A decision maker may have probabilities for these states of
nature; however, it may be possible to delay a decision
until it is clear which state of nature will occur.

◆ The question to consider is whether the cost of taking

decision now will be less than the expected gain due to
delaying the decision.

◆ The expected gain is the expected value of perfect

information, or EVPI.
26
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Expected value of
perfect information (EVPI)
◆EVPI is the difference between the expected payoff with
perfect information (under certainty) and the expected
payoff under risk.

Expected value of perfect information (EVPI )=

Expected payoff under certainty - Expected payoff under risk

EVPI = EPC - EMV

27
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
Expected payoff under certainty
(EPC)
◆ To determine the EVPI. First, compute the expected payoff
under certainty (EPC).

◆ To do this, identify the best payoff under each state of

nature.

◆ Then combine these by weighting each payoff by the

probability of that state of nature and adding the amounts.

28
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 EVPI and EPC Example
Best Payoff from each
alternative (column) for the
3 facilities

$ Low demand $ Med demand $ High demand

0.30 0.50 0.20
Small facility 10 * 0.30 = 3 10 10

Medium facility 7 12 *0.50 = 6 12

Large facility -4 2 16 *0.20 = 3.2

Expected payoff under certainty (EPC)

= 0.30($10) + 0.50($12) + 0.20($16)

EPC = $12.2
29
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 EVPI and EPC Example
◆EPC = $12.2
◆If the expected payoff under risk (EMV) is $10.5 (from
previous example).
Expected value of perfect information (EVPI )=
Expected payoff under certainty - Expected payoff under risk
EVPI = EPC – EMV
EVPI = $12.2 - $10.5 = $1.7

Expected value of perfect information (EVPI )= $1.7

You would be willing to spend up to $1.7 million
to obtain perfect information

30
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Solve
Alternative State of nature
#1 #2 #3 #4
A 1 0 1 6
B 1 5 4 2
C 3 2 2 3

31
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 Solve
1. If somehow you find out for certain that state of nature #4 is going to occur,
which alternative will you select?
2. If you are uncertain which state of nature will occur, and use the maximin
criterion, which alternative will you select?
3. If you are uncertain which state of nature will occur, and use the maximax
criterion, which alternative will you select?
4. If you are uncertain which state of nature will occur, and use the Laplace
criterion, which alternative will you select?
5. If you are uncertain which state of nature will occur, and use the minimax regret
criterion, which alternative will you select?
6. If you feel that P(#1) = .4, P(#2) = .3, P(#3) = .2, and P(#4) = .1, what is the
expected monetary value for each alternative and which alternative will you
select?
7. If you feel that P(#1) = .4, P(#2) = .3, P(#3) = .2, and P(#4) = .1, what is your
expected payoff under certainty?
8. If you feel that P(#1) = .4, P(#2) = .3, P(#3) = .2, and P(#4) = .1, what is your
expected value of perfect information?
32
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 1. If somehow you find out for certain that state of nature
#4 is going to occur, which alternative will you select?

Alternative State of nature

#1 #2 #3 #4
A 1 0 1 6 A=6
B 1 5 4 2
C 3 2 2 3

Choose the alternative that has the best payoff

under that state of nature.

33
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 2. If you are uncertain which state of nature will occur,
which alternative will you select using the maximin
criterion?
Alter. State of nature
maximin
#1 #2 #3 #4
A 1 0 1 6 0
B 1 5 4 2 1
C 3 2 2 3 C=2

Select the maximum of the minimum

34
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 3. If you are uncertain which state of nature will occur,
which alternative will you select using the maximax
criterion?

Alter. State of nature

maximax
#1 #2 #3 #4
A 1 0 1 6 A=6
B 1 5 4 2 5
C 3 2 2 3 3

Select the maximum of the maximum

35
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 4. If you are uncertain which state of nature will occur, which
alternative will you select using the Laplace criterion?
Alter. State of nature
Laplace
#1 #2 #3 #4
A 1 0 1 6 1+0+1+6/4= 2

B 1 5 4 2 12/4=3 B=3
C 3 2 2 3 10/4=2.5

Select the decision that gives the

highest average payoff

36
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 5. If you are uncertain which state of nature will occur,
which alternative will you select using the minimax regret
criterion?

Alter. State of nature

#1 #2 #3 #4 minimax
regret
Maximum 3 5 4 6
outcome
A 3-1= 5-0= 4-1= 6-6= 5
2 5 3 0
B 3-1= 5-5= 4-4= 6-2= 4
2 0 0 4
C 3-3= 5-2= 4-2= 6-3= 3 C=3
0 3 2 3

Pick the alternative with the max. value then

select the minimum number
37
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 6. If you feel that P(#1) = 0.4, P(#2) = 0.3, P(#3) = 0.2, and
P(#4) = 0.1, what is the expected monetary value for each
alternative and which alternative will you select?
Alter. State of nature

#1 #2 #3 #4
EMV
P 0.4 0.3 0.2 0.1
1*0.4= 0*0.3= 1*0.2= 6 *0.1= 0.4+0.2+0.6=
A
0.4 0 0.2 0.6 1.2
1*0.4= 5*0.3= 4*0.2= 2*0.1= 0.4+1.5+0.8+0.2= B
B
0.4 1.5 0.8 0.2 2.9
3*0.4= 2*0.3= 2*0.2= 3*0.1= 1.2+0.6+0.4+0.3=
C 2.5
1.2 0.6 0.4 0.3
The expected value is the sum of the payoffs for an alternative where each
payoff is weighted by the probability for the relevant state of nature.
The expected value is computed for each alternative, and the one with the
highest
Production expected
and Operation value is selected.
Management 38
 7. If you feel that P(#1)= 0.4, P(#2)= 0.3, P(#3)= 0.2, and
P(#4)= 0.1, what is your expected payoff under certainty?
Alter. State of nature
#1 #2 #3 #4
P 0.4 0.3 0.2 0.1
A 1 0 1 6
B 1 5 4 2
C 3 2 2 3

Expected payoff under certainty (EPC) =

3*0.4 + 5*0.3 + + 4*0.2 + 6*0.1 = 1.2 + 1.5 + 0.8 + 0.6
= 4.1

39
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
 8. If you feel that P(#1)= 0.4, P(#2)= 0.3, P(#3)= 0.2, and
P(#4)= 0.1, what is your expected value of perfect
information?

Expected value of perfect information (EVPI) = EPC - EMV

= 4.1 - 2.9
= 1.2

40
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT
Decision Theory Environment
• An environment where the set of alternative
managerial actions is known, and also the
Certainty outcome of each action is known (with
certainty).

• An environment where the set of alternative

managerial actions is known, but absolutely
Uncertainty no information is available regarding the
possible outcomes of each action.

• An environment where the set of alternative

managerial actions is known, but where the
Risk outcome of each action can be estimated
only in terms of a probability distribution.
41
 Four criteria used for Decision
Making under Uncertainty
Maximin
• Choose the alternative with the best of the
(Pessimistic worst possible payoffs.
Decision)
Maximax
• Choose the alternative with the best
(Optimistic possible payoff.
Decision)
• Choose the alternative with the best
Laplace average payoff of any of the alternatives.

• Choose the alternative that has the least of

Minimax regret the worst regrets.

42
Production Operations Management – Dr. Nevien Farouk Khourshed - AASTMT