Decision Analysis

(Theory)

To accompany Quantitative Analysis 3-1 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Introduction
• Decision theory is an analytical
and systematic way to tackle
problems

• A good decision is based on

logic.

To accompany Quantitative Analysis 3-2 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 The Six Steps in
Decision Theory
1) Clearly define the problem at
hand
2) List the possible alternatives
3) Identify the possible outcomes
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
To accompany Quantitative Analysis 3-3 © 2003 by Prentice Hall, Inc.
for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision Table
for Thompson Lumber

State of Nature
Alternative Favorable Unfavorable
Market Market
Construct a $200,000 -$180,000
large plant
Construct a $100,000 -$20,000
small plant
Do nothing $0 0

To accompany Quantitative Analysis 3-4 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Types of Decision-
Making Environments
• Type 1: Decision-making under certainty
• decision-maker knows with certainty the
consequences of every alternative or decision choice

• Type 2: Decision-making under risk

• The decision-maker does know the probabilities of
the various outcomes

• Type 3: Decision-making under uncertainty

• The decision-maker does not know the probabilities
of the various outcomes

To accompany Quantitative Analysis 3-5 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision-Making
Certainty

Knows with certainty the result

of every alternative

To accompany Quantitative Analysis 3-6 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision-Making
Under Risk

Expected Monetary Value

n
EMV(Alternative)   Payoff S j * P ( S j )
j 1

where n  number of stages of nature.

To accompany Quantitative Analysis 3-7 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision Table
(Matrix)
for Thompson Lumber

Favorable Unfavorable
Market Market
Alternative State of Nature
Construct a $200,000 -$180,000 $10,000
large plant
Construct a $100,000 -$20,000 $40,000
small plant
Do nothing $0 0

0.50 0.50

To accompany Quantitative Analysis 3-8 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Expected Value of
Perfect Information
(EVPI)
• EVPI places an upper bound on
what one would pay for
additional information

• EVPI is the expected value with

perfect information minus the
maximum EMV

To accompany Quantitative Analysis 3-9 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Expected Value With
Perfect Information
(EV | PI)

n
EV | PI   (Best outcome for state of nature) * P(S j )
j1

n  number of states of nature.

To accompany Quantitative Analysis 3-10 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Expected Value of
Perfect Information

• EVPI = EV|PI - maximum EMV

To accompany Quantitative Analysis 3-11 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Expected Value of
Perfect Information
State of Nature
Alternative Favorable Unfavorable EMV
Market Market
Construct a $200,000
large plant

Construct a $40,000
small plant

Do Nothing $0

0.50 0.50

To accompany Quantitative Analysis 3-12 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Expected Value of
Perfect Information

EVPI = expected value with perfect

information - max(EMV)

= $200,000*0.50 + $0*0.50 - $40,000

= $60,000

To accompany Quantitative Analysis 3-13 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Expected Opportunity
Loss

• EOL is the cost of not picking

the best solution
• EOL = Expected Regret

To accompany Quantitative Analysis 3-14 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Computing EOL - The
Opportunity Loss Table

State of Nature

Alternative Favorable Market Unfavorable

($) Market ($)
Large Plant 200,000 - 200,000 0 - (-180,000)

Small Plant 200,000 - 100,000 0 -(-20,000)

Do Nothing 200,000 - 0 0-0

Probability 0.50 0.50

To accompany Quantitative Analysis 3-15 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 The Opportunity Loss
Table continued

State of Nature

Alternative Favorable Unfavorable

Market Market
Large Plant 0 $180,000

Small Plant $100,000 $20,000

Do Nothing $200,000 0

Probability 0.50 0.50

To accompany Quantitative Analysis 3-16 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 The Opportunity Loss
Table - continued

Alternative EOL

Large Plant (0.50)*$0 + $90,000

(0.50)*($180,000)
Small Plant (0.50)*($100,000) $60,000
+ (0.50)(*$20,000)
Do Nothing (0.50)*($200,000) $100,000
+ (0.50)*($0)

To accompany Quantitative Analysis 3-17 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Sensitivity Analysis

EMV(Large Plant) = $200,000P - (1-

P)$180,000

EMV(Small Plant) = $100,000P -

$20,000(1-P)

EMV(Do Nothing) = $0P + 0(1-P)

To accompany Quantitative Analysis 3-18 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Sensitivity Analysis -
continued

250000
200000
Point 1 Point 2
150000 Small Plant
100000
EMV Values

50000
0
-50000 0 0.2 0.4 0.6 0.8 1
-100000
-150000 Large Plant EMV

-200000
Values of P

To accompany Quantitative Analysis 3-19 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision Making
Under Uncertainty

• Maximax
• Maximin
• Equally likely (Laplace)
• Criterion of Realism
• Minimax

To accompany Quantitative Analysis 3-20 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision Making
Under Uncertainty
Maximax - Choose the alternative
with the maximum output
State of Nature
Alternative Favorable Unfavorable
Market Market
Construct a 200,000 -180,000
large plant
Construct a 100,000 -20,000
small plant
Do nothing 0 0
Probability 0.50 0.50

To accompany Quantitative Analysis 3-21 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision Making
Under Uncertainty
Maximin - Choose the alternative with
the maximum minimum output

State of Nature
Alternative Favorable Unfavorable
Market Market
Construct a 200,000 -180,000
large plant
Construct a 100,000 -20,000
small plant
Do nothing 0 0
Probabilities

To accompany Quantitative Analysis 3-22 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision Making
Under Uncertainty
Equally likely (Laplace) - Assume all states of
nature to be equally likely, choose maximum
Average

States of Nature
Alternative Favorable Unfavorable Avg.
Market Market
Construct $200,000 -$180,000 10,000
Large Plant
Construct 100,000 -20,000 40,000
small plant
Do nothing 0 0 0

To accompany Quantitative Analysis 3-23 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision Making
Under Uncertainty
Criterion of Realism (Hurwicz):
CR = *(row max) + (1-)*(row min)

State of Nature
Alternative Favorable Unfavorable CR
Market Market
Construct $200,000 -180,000 124,000
large plant
Construct $100,000 -20,000 76,000
small plant
Do nothing 0 0 0
0.80 0.20

To accompany Quantitative Analysis 3-24 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Decision Making
Under Uncertainty
Minimax - choose the alternative with
the minimum maximum Opportunity
Loss
Regret Matrix
States of Nature
Alternative Favorable Unfavorable Max
Market Market
Construct a 0$ $180,000 $180,000
large plant
Construct a $100,000 20,000 100,000
small plant
Do nothing 200,000 0 200,000

To accompany Quantitative Analysis 3-25 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Marginal Analysis -
Discrete Distributions
• Steps using Normal Distributions:
• Determine the value for P.

ML
P
MLMP
• Locate P on the normal distribution. For a
given area under the curve, we find Z from
the standard Normal table.
X 
*
• Using Z we can now

solve for X*

To accompany Quantitative Analysis 3-26 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Joe’s Newsstand
Example A

• ML = 4

• MP = 6

  = Average demand = 50
papers per day

  = Standard deviation of
demand = 10

To accompany Quantitative Analysis 3-27 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Joe’s Newsstand
Example A - continued
ML 4
Step 1: p    0.40
ML  MP 4  6

Step 2: Look in the Normal table

for P = 0.6 (i.e., 1 – 0.4)
X *  50
Z  0.25 
10
or.
X *  10 * 0.25  50
 52.5 or 53 newspapers

To accompany Quantitative Analysis 3-28 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Joe’s Newsstand
Example A continued

To accompany Quantitative Analysis 3-29 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Joe’s Newsstand
Example B
• ML = 8
• MP = 2
  = Average demand = 100
papers per day
  = Standard deviation of
demand = 10

To accompany Quantitative Analysis 3-30 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Joe’s Newsstand
Example B - continued
ML 8
• Step 1: p    0.80
ML  MP 8  2
• Step 2:
Z = -0.84 for an area of 0.80

and
X  1000 *
 0.84 
10
or:
X *  0  0.84(10)  100
 91.6 or 92 newspapers

To accompany Quantitative Analysis 3-31 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
 Joe’s Newsstand
Example B continued

To accompany Quantitative Analysis 3-32 © 2003 by Prentice Hall, Inc.

for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
To accompany Quantitative Analysis 3-33 © 2003 by Prentice Hall, Inc.
for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
To accompany Quantitative Analysis 3-34 © 2003 by Prentice Hall, Inc.
for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
To accompany Quantitative Analysis 3-35 © 2003 by Prentice Hall, Inc.
for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna
To accompany Quantitative Analysis 3-36 © 2003 by Prentice Hall, Inc.
for Management, 8e Upper Saddle River, NJ 07458
by Render/Stair/Hanna