# Operations Management

**Decision-Making Tools
**

A-1

Outline

♦ Decision Making & Models. ♦ Decision Tables.

♦

Decision making under uncertainty. ♦ Decision making under risk. ♦ Expected value of perfect information (EVPI).

♦ Decision Trees.

A-2

**The Decision-Making Process
**

Quantitative Analysis Logic Historical Data Marketing Research Scientific Analysis Modeling Qualitative Analysis Emotions Intuition Personal Experience and Motivation Rumors

A-3

Problem

Decision

**Models and Scientific Management
**

♦ Can Help Managers to: to

♦ Gain

**deeper insights into the business. better decisions!
**

♦ Better

♦ Make

assess alternative plans and actions.

♦ Quantify,

reduce and understand the uncertainty surrounding A-4 business plans and actions.

Steps to Good Decisions
.
A-5 ♦ Implement decision. ♦ Select decision-making tool (model). ♦ Establish decision criteria. ♦ Identify and evaluate alternatives using decisionmaking tool (model). ♦ Select best alternative.♦ Define problem and influencing factors.

♦ Less expensive and disruptive than experimenting with the real world system.Benefits of Models
♦ Allow better and faster decisions. ♦ Force a consistent and systematic approach to the analysis of problems. ♦ Allow managers to ask “What if…?” questions. A-6
♦
.

♦ May use assumptions that oversimplify the real world. ♦ May be unused.
♦
Due to mathematical and logical complexity.
♦ May downplay the value of qualitative information. A-7
. misused or misunderstood (and feared!).Limitations of Models
♦ May be expensive and timeconsuming to develop and test.

State of nature: An occurrence over which the decision maker has no control. A-8
.Decision Theory
Terms:
Alternative: Course of action or choice. Decision-maker chooses among alternatives.

Decision Table
States of Nature State 1 State 2
Alternativ Outcome 1 Outcome 2 e1 Alternativ Outcome 3 Outcome 4 e2
A-9
.

If a large plant is constructed and the market is favorable.A firm has two options for expanding production of a product: (1) construct a large plant. then the result is a loss of $20.000. If a small plant is constructed and the market is unfavorable. then the result is a loss of $180.000. Of A-10
Example . then the result is a profit of $200.000. or (2) construct a small plant. then the result is a profit of $100. If a small plant is constructed and the market is favorable. If a large plant is constructed and the market is unfavorable. the future market for the product will be either favorable or unfavorable. Whether or not the firm expands.Decision Making Under Uncertainty
.000.

000 -$20.000 -$180.000
$0
$0
A-11
.Decision Making Under Uncertainty
States of Nature
Alternatives avorable F Unfavorable
large plant Construct $100.Example .000 small plant Do nothing
Market Market Construct $200.

♦ Expected Value .Decision Making Under Uncertainty .Choose alternative that maximizes the minimum outcome for every alternative (Pessimistic criterion). ♦ Maximin .Criteria
♦ Maximax .Choose A-12 alternative with the highest
.Choose alternative that maximizes the maximum outcome for every alternative (Optimistic criterion).

000 -$180.Example .000
$0
$0
Maximax decision is to construct large plant.000 small plant Do nothing
Market Market Construct $200.000 -$20.Maximax
States of Nature
Alternatives avorable F Unfavorable
large plant Construct $100.
A-13
.

Example . (Maximum of minimums for each A-14 alternative)
.000
$0
$0
$0
Maximin decision is to do nothing.000-$180.Maximin
States of Nature
Alternatives avorable F Unfavorable Minimum
large plant Construct $100.000 -$180.000 -$20.000 small plant Do nothing
in Row Market Market Construct $200.000 -$20.

Decision Making Under Risk
♦ States of nature have probabilities of occurrence.♦ Probabilistic decision situation.
♦
EV = Average return for alternative if decision were repeated many times.
A-15
. ♦ Select alternative with largest expected value (EV).

Expected Value Equation
Number of states of nature Value of Payoff Probability of payoff
EV (Ai )= Σ Vi *P (Vi )
i=1
N
=V1 *P (V1)+V2 * P (V2 )+...+ N * P (VN ) V
Alternative i
A-16
.

000 -$180.5 Probability of unfavorable market = 0.000
$0
$0
$0
Decision is to “Construct small plant”.Expected Value
large plant Construct $100.000 -$20. A-17
.000 small plant Do nothing
Expect ed Market Market Value Construct $200.5 States of Nature
Alternatives avorable F Unfavorable
Example .000 $40.000 $10.Suppose: Probability of favorable market = 0.

3 States of Nature
Alternatives avorable F Unfavorable
Example .7 Probability of unfavorable market = 0.000 $86. decision is to “Construct large A-18 plant”.Expected Value
large plant Construct $100.000 -$20.000
$0
$0
$0
Now.Suppose: Probability of favorable market = 0.
.000 small plant Do nothing
Expect ed Market Market Value Construct $200.000 -$180.000 $64.

Over what range of values for probability of favorable market is “Construct large plant” preferred?
Example .000
$0
$0
Solve for x: 380000x-180000 > 120000x-20000 A-19
.000x .000x .000 120.180.000 -$20.000 380.Expected Value
States of Nature
Alternatives avorable F Unfavorable
large plant Construct $100.000 small plant Do nothing
Expected Market Market Value Construct $200.20.000 -$180.

then “Construct large plant”.6154 So.Expected Value
. as long as probability of a favorable market exceeds 0.
A-20
Example .180000 > 120000x .Over what range of values for probability of favorable market is “Construct large plant” preferred? Solve for x: 380000x .6154.20000 x > 0.

♦ EVPI is the expected value under certainty (EVUC) minus the maximum EV.Expected Value of Perfect Information (EVPI) bound on ♦ EVPI places an upper
what one would pay for additional information.maximum EV
A-21
.
♦
EVPI is the maximum you should pay to learn the future. EVPI = EVUC .

n
A-22
.Expected Value Under Certainty (EVUC)
EVU = ∑ (Best outcome for the state of * P(Sj ) j= 1 nature j) C where: P(Sj ) = The probability of state of nature j. n = Number of states of nature.

Example .000 small plant Do nothing
Market Market Construct $200.000
$0
$0
Best outcome for Favorable Market = $200.000 Best outcome for Unfavorable A-23
.000 -$20.EVUC
States of Nature
Alternatives avorable F Unfavorable
large plant Construct $100.000-$180.

5 EV
= ($200.Expected Value of Perfect Information
Suppose: Probability of favorable market = 0.000 = $60.5 Probability of unfavorable market = EVUC .000 to learn whether A-24 the market will be favorable or
.max(EV) EVPI = 0.50 + 0*0.000*0.000
Thus. you should be willing to pay up to $60.50) .$40.

30) .Expected Value of Perfect Information
Now suppose: Probability of favorable market = 0.000
Now.000 to learn whether A-25 the market will be favorable or
.3 EV
= ($200. you should be willing to pay up to $54.70 + 0*0.7 Probability of unfavorable market = EVUC .max(EV) EVPI = 0.$86.000*0.000 = $54.

A-26
. ♦ Used for solving problems with several sets of alternatives and states of nature (sequential decisions).
♦
Decision tables can not be used for more than one decision.Decision Trees
♦ Graphical display of decision process.
♦ Expected Value criterion is used.

♦ Assign probabilities to all states of nature. ♦ Estimate payoffs for each combination of alternatives and states of nature. ♦ Draw the decision tree. ♦ Solve the problem:
♦
Compute expected values for each state-of-nature node moving right A-27
.Using Decision Trees
♦ Define the problem.

A-28 A state of nature node out of which
.Decision Theory
Terms:
♦ Alternative: Course of action or
choice.
Symbols used in decision tree:
A decision node from which one of several alternatives may be selected. ♦ State of nature: An occurrence over which the decision maker has no control.

Decision Tree
State 1
e1 tiv a rn lte A lt ern ati ve 2
1
State 2 State 1
Outcome 1 Outcome 2
2
Decision Node
Outcome 3 State 2 Outcome 4 State of Nature Node
A-29
.

3) La -$180.7) ma $100.000 th in g
$200.3) no -$20.000 d uil B Bui ld S Favorable Mkt (0.000
$0
A-30
.000 ll Do Unfavorable Mkt (0.Decision Tree for Example
Favorable Mkt (0.7)
rge Unfavorable Mkt (0.

7)
rge Unfavorable Mkt (0.Decision Tree for Example .7) ma $100.000 th in g $0
$200.000
Favorable Mkt (0.3) La -$180.00 Favorable Mkt (0.Solution
$86.3) no -$20.000 ll 0 Do Unfavorable Mkt (0.000 d uil B Bui ld S $64.000
$0
A-31
.

they can do nothing and payoff is $45 million. they can do nothing and payoff is -
. If they build “small” and demand is “high”.) If they build “small” and demand is “low”. or they can expand. the payoff is $40 million. there is a 30% chance the demand drops off and the payoff will be $35 million. If they build “large” and A-32 demand is “low”. If they build “large” and demand is “high”.6. If they expand. Demand for the new product will be high or low initially.4. The probability of high demand is 0. the payoff is $60 million. and a 70% chance the demand grows and the payoff is $48 million. (The probability of low demand is 0.Decision Tree Example
A firm can build a large plant or small plant initially (for a new product).

Build “Large” or “Small” plant initially. 2.Decision Tree Example
Three decisions:
1. then demand “Grows” A-33 (0. 2. then “Expand” or “Do nothing”. demand is “High”. Demand is “High” (0. 3.
.4) initially. If build “Small”.
Two states of nature:
1.6) or “Low” (0. If build “Large” and demand is “Low”. then decide to “Reduce prices” or “Do nothing”. and decision is “Expand”. If build “Small” and demand is “High”.7) or demand “Drops” (0.3).

Decision Tree
Demand grows (0. 4)
$20 -$10
3
Do nothing
A-34
.6 2 (0
$48 $35 $45
d an p Ex
Demand drops (0. 4)
$40 $60
Reduce prices
1
Bu ild lar ge
.6) igh (0 H
Lo w
(0.7)
) .3)
d uil B
all m s
h ig H
Do nothing
Low (0.

Start with Decision 3: “Reduce prices” or “Do nothing”. Choose “Reduce prices” (20 > -10).
A-35
.Decision Tree Solution
Work right to left (from end back to beginning).

7)
) . 4)
$20
3
$20 -$10
A-36
.Decision Tree
Demand grows (0.6) igh (0 H
Lo w
(0.6 2 (0
$48 $35 $45
d an p Ex
Demand drops (0.3)
d uil B
all m s
h ig H
Do nothing
Low (0. 4)
$40 $60
Reduce prices Do nothing
1
Bu ild lar ge
.

Decision Tree Solution
Consider Decision 2: “Expand” or “Do nothing”.
A-37
.3) = 44.1).1 Choose “Do nothing” (45 > 44. To compare outcomes we need expected value if we “Expand”: (48*0.7) + (35*0.

3)
$48 $35 $45
$45
Do nothing
d uil B
all m s
Low (0.6) igh (0 H
Lo w
(0. 4)
$20
3
$20 -$10
A-38
. d an 1 $45xp
Demand grows (0.Decision Tree
h ig H ) E .7) Demand drops (0.6 2 (0
$44. 4)
$40 $60
Reduce prices Do nothing
1
Bu ild lar ge
.

6 2 (0
$44. 4)
$20
3
$20 -$10
A-39
.3)
$48 $35 $45
$45
Do nothing
d uil B
all m s
Low (0.Decision Tree
$43 ig H
h ) E .7) Demand drops (0.6) H
Lo w (0. d an 1 $45xp
Demand grows (0. 4)
$40 $60
Reduce prices Do nothing
1
Bu ild lar ge
$44 igh (0.

Expected payoff = $44 million.
A-40
.Decision Tree Final Solution
Decisions: 1. then “Reduce prices”. 2. If demand is “Low”. Build “Large”.

3 0.4 0.6 3 0.Larger Decision Tree
$10
2 0.3 0.2 0.3 A-41 0.5 1 0.4 0.4 0.6
$8 $12 $9 $11 $6 $12 $8 $9 $8
.

Larger Decision Tree Solution $10
$10.6 0.28
0.5
$11
0.6 $9.4
2 0.6
$8 $12 $9
$10.4
0.2
$10.3 0.4 0.3 A-42
1
$9.4 0.3
.3 0.4 0.6
3
$6 $12 $8 $9 $8
$8.