IIMT 2641 Introduction to Business Analytics

Module 2: Decision Analysis

Developing a New Product, Summer Job Decision
Xiao Lei
Today’s Objectives
1. Use probabilities to determine optimal decisions under uncertainty (based on
expected value criterion) .

2. Set up and solve decision trees - Summer Job Decision

3. Developing a new product case

4. Sensitivity analysis – revisit Summer Job Decision

2
What is Analytics?
• The science of using data to build models that lead to
better decisions that add value to individuals, to companies, to
institutions
• Descriptive analytics: identify patterns in the data
• Summary statistics
• Hypothesis Test
• Visualizations
• Clustering
• Predictive analytics: predict different outcomes
• Linear Regression
• Logistic Regression, CART, Random Forests
• Prescriptive analytics: give advice on actions to take
3
The Science of Decision Making
▪ Making a decision is basically making a choice.
– Whether to pursue a graduate study
– Whether to buy a stock and how much money to invest
– Whether to expand the product line and how to expand

▪ Six Steps in Decision Making

– Clearly define the problem (Goal to achieve)
– List the possible alternatives
– Identify the possible outcomes or states of nature
– List the payoff of each alternative in each state of nature
– Select one of the decision theory models
– Apply the model and make your decision

4
Thompson Lumber Company

▪ John Thompson, the founder and president, needs to

make the decision on whether to expand his product
line by manufacturing and marketing a new product.
▪ Define the problem

STATE OF NATURE

FAVORABLE MARKET UNFAVORABLE MARKET

ALTERNATIVE (profit in $) (profit in $)

Construct a large plant 200,000 –180,000

Construct a small plant 100,000 –20,000
Do nothing 0 0

5
Thompson Lumber Company
▪ List all possible alternatives
▪ Identify possible outcomes/states of nature
▪ List the payoff

STATE OF NATURE

FAVORABLE MARKET UNFAVORABLE MARKET

ALTERNATIVE (profit in $) (profit in $)

Construct a large plant 200,000 –180,000

Construct a small plant 100,000 –20,000
Do nothing 0 0

6
Types of Decision-Making Environments
▪ Decision making under certainty
– The decision maker knows with certainty the consequences of every alternative
or decision choice.

▪ Decision making under risk

– There are several possible outcomes for each alternative, and decision maker
knows the probabilities of each outcome.

▪ Decision making under uncertainty

– There are several possible outcomes for each alternative, and decision maker
does not know the probabilities of the various outcomes.

7
Decision Making under Risk

STATE OF NATURE Expected

Monetary
FAVORABLE MARKET UNFAVORABLE Value
ALTERNATIVE (profit in $) MARKET (profit in $) (EMV, $)

Construct a large plant 200,000 –180,000 –9,000

Construct a small plant 100,000 –20,000 34,000
Do nothing 0 0 0
Probability 0.45 0.55

Expected Monetary Value (EMV) – EMV is the weighted sum

of possible payoffs for each alternative
If John maximize the EMV, then he should choose to construct
a small plant.

8
Decision Making under Risk
▪ Scientific Marketing, Inc. offers analysis that will provide certainty
about market conditions. S.M. would charge $65,000 for the
information. Should John buy the information?

▪ To make this decision, John has to evaluate the expected value of

perfect information (EVPI) by computing the expected value with
perfect information (EVwPI) and the best EMV under risk.

9
Decision Making under Risk
▪ EVwPI = σ best payoff in state 𝑖 ∙ probability of state 𝑖

STATE OF NATURE Expected

Monetary
FAVORABLE MARKET UNFAVORABLE Value
ALTERNATIVE (profit in $) MARKET (profit in $) (EMV, $)

Construct a large plant 200,000 –180,000 –9,000

Construct a small plant 100,000 –20,000 34,000
Do nothing 0 0 0
Best payoff 200,000 0 90,000
Probability 0.45 0.55 EVwPI

EVPI = EVwPI – Best EMV

= $90,000 – $34,000 = $56,000 < $65,000. Don’t buy! 10
Decision Making under Risk

▪ To minimize expected opportunity loss (EOL)

STATE OF NATURE Expected

Monetary
FAVORABLE MktMARKET(Opp. UNFAVORABLE
UNFAVORABLE Mkt EOL
Value
ALTERNATIVE Loss inin$)$)
(profit (Opp. Loss
MARKET in $)
(profit in $) (in $)
(EMV,

Construct a large plant 200,0000 180,000

–180,000 99,000
–9,000
Construct a small plant 100,000 20,000
–20,000 56,000
34,000
Do nothing 200,0000 0 90,0000
Probability 0.45 0.55

▪ Minimizing EOL always results in the same decision as maximizing EMV.

▪ The minimum EOL always equal the EVPI.

11
Decision Making under Risk

Favorable Market VwPI Unfavorable Market

OL2
OL3 EVwPI
VwPI EOL1 EOL2 EOL3
MV1
OL2
MV2 MV3 OL1 MV3
MV2 EMV1 EMV2 EMV3
MV1
Option Option Option Option Option Option Option 1 Option 2 Option 3
1 2 3 1 2 3

0.45 prob. 0.55 prob.

12
In-class Exercise: Café du Donut
▪ The Café buys donuts each day for $40 per carton of 20 dozen donuts. Any cartons
not sold are thrown away at the end of the day. If a carton is sold, the total revenue
is $60.
DAILY DEMAND PROBABILITY CUMULATIVE
(CARTONS) PROBABILITY
• The original plan is 4 0.05 0.05
to order 6 cartons 5 0.15 0.2
6 0.15 0.35
per day. Should the
7 0.20 0.55
Café increase the 8 0.25 0.8
order size to 7? 9 0.10 0.9
10 0.10 1.0
Total 1.00

13
In-class Exercise: Café du Donut
▪ Monetary Payoff (Profit) Table

• Should we reduce the order size from 6 to 5? What is the EMV

of Q=5?

• If we can only choose between 6 and 7, what is the EVPI?

14
Decision Making under Risk
▪ If the estimation of probability is changed, how will John change his decision?
▪ A sensitivity analysis is needed!

STATE OF NATURE

FAVORABLE UNFAVORABLE EMV

ALTERNATIVE MARKET (profit in $) MARKET (profit in $) (in $1,000)

Construct a large plant 200,000 –180,000 380p – 180

Construct a small plant 100,000 –20,000 120p – 20
Do nothing 0 0 0
Probability p 1–p

15
Decision Making under Risk

EMV

$300,000

$200,000 EMV (large plant)

Point 2 380,000p-180,000

$100,000 EMV (small plant)

Point 1 120,000p-20,000

0 EMV (do nothing)

0
.167 .615 1
–$100,000 Values of p

–$200,000

16
Results of Sensitivity Analysis

BEST ALTERNATIVE RANGE OF P VALUES

Construct a large plant >0.615

Construct a small plant 0.167 - 0.615

Do nothing <0.167

17
Sensitivity Analysis in Excel
▪ Step 1: setup the payoff model (data and formulas)
▪ Step 2: setup the data table
– Set a list of values in the first column for the parameter to be explored, except
for the first row.
– Reference the output values in the first row, starting from the second column.
▪ Step 3: generate the one-way data table
– Select the entire data table.
– Click: DATA -> What-If Analysis -> Data Table.
– Set the Column input cell to the parameter cell in the original model.

18
In-class Exercises
▪ The Monty Hall Problem Revisited
– Suppose you're on a game show, and you're given the choice of three doors: Behind one door is
a car; behind the others, goats. You pick a door, say No. 1, and the host, who knows what's
behind the doors, opens another door, say No. 3, which has a goat. He then says to you, "Do you
want to pick door No. 2?" Is it to your advantage to switch your choice?

Given door 1 is Car behind Car behind Car behind Expected

chosen door 1 door 2 door 3 Value
Switch 0 1 1 2/3

Do not switch 1 0 0 1/3

Probability 1/3 1/3 1/3

19
In-Class Exercise
▪ Assume directly measuring monetary payoff is infeasible
▪ Opportunity Loss Table

State I State II
Option A 5 1
Option B 0 3
Option C 6 0
Probability 0.3 0.7

▪ Which is the option that maximizes EMV?

20
In-Class Exercise
▪ Given Opportunity Loss Table
State I State II
Option A 5 1
Option B 0 3
Option C 6 0
Probability 0.3 0.7

▪ How to restore the payoff table?

State I State II
Option A 1 3
EVwPI
Option B =6 6(0.3) + 4(0.7)
1 = 4.6
Option C 0 4

▪ What is the EVwPI?

21
Summary
▪ Six steps for decision making
▪ Decision Making under Risk
– Maximize EMV
– Minimize EOL gives the same choices
– EVPI=EVwPI-best EMV=best EOL
– Sensitivity analysis

22
Decision Making under Uncertainty (Optional)
What if we do not know the probabilities?

▪ Maximax (optimistic)
▪ Maximin (pessimistic)
▪ Minimax regret

23
Decision Making under Uncertainty (Optional)
▪ Maximax (optimistic)

STATE OF NATURE

FAVORABLE UNFAVORABLE MAXIMUM IN A

ALTERNATIVE MARKET ($) MARKET ($) ROW ($)
Construct a large plant 200,000 –180,000 200,000

Construct a small plant 100,000 –20,000 100,000

Do nothing 0 0 0

24
Decision Making under Uncertainty (Optional)
▪ Maximin (pessimistic)

STATE OF NATURE

FAVORABLE UNFAVORABLE MINIMUM IN A

ALTERNATIVE MARKET ($) MARKET ($) ROW ($)
Construct a large plant 200,000 –180,000 –180,000

Construct a small plant 100,000 –20,000 –20,000

Do nothing 0 0 0

25
Decision Making under Uncertainty (Optional)
▪ Minimax regret
– Based on opportunity loss or regret.

STATE OF NATURE

FAVORABLE UNFAVORABLE MAXIMUM IN A

ALTERNATIVE MARKET ($) MARKET ($) ROW ($)
Construct a large plant 0 180,000 180,000

Construct a small plant 100,000 20,000 100,000

Do nothing 200,000 0 200,000

26
Today’s Objectives
1. Use probabilities to determine optimal decisions under uncertainty (based on
expected value criterion) .

2. Set up and solve decision trees - Summer Job Decision

3. Developing a new product case

4. Sensitivity analysis – revisit Summer Job Decision

27
Decision Tree
▪ Any problem that can be presented in a decision table can also be
graphically illustrated in a decision tree.

▪ Advantages of decision tree:

– Incorporate sequential decision making
– Incorporate different states and probabilities for different options

▪ Any decision tree has two types of node:

– Decision node, from which one of several alternatives may be chosen
– State-of-nature node, out of which one state of nature will occur

28
Decision Making under Risk

STATE OF NATURE Expected

Monetary
FAVORABLE MARKET UNFAVORABLE Value
ALTERNATIVE (profit in $) MARKET (profit in $) (EMV, $)

Construct a large plant 200,000 –180,000 –9,000

Construct a small plant 100,000 –20,000 34,000
Do nothing 0 0 0
Probability 0.45 0.55

Expected Monetary Value (EMV) – EMV is the weighted sum

of possible payoffs for each alternative
If John maximize the EMV, then he should choose to construct
a small plant.

29
Thompson’s Decision Tree

A State-of-Nature Node Payoffs

Favorable Market (0.45)
$200,000
A Decision Node EMV1 = -9k 1
Unfavorable Market (0.55)
–$180,000

(0.45)
EMV2 = 34k Favorable Market $100,000
Construct a
2
Small Plant Unfavorable Market (0.55)
–$20,000

$0

30
Summer job decision

▪ Bill is a first-year MBA student at the HKU

▪ He met Vanessa, an investment banker, on a flight in August before the school began
▪ Impressed by his background, Vanessa asked Bill to contact her in November for a potential job
opportunity for the next summer
▪ When Bill quitted his job before studying for his MBA, his former boss, John, had promised him a job
for the next summer
▪ The summer salary would be $12,000, but the offer would expire by the end of October
▪ Questions:
– Should Bill accept John's summer job offer before he knows any details about Vanessa's potential
job offer?
– If Bill were to turn down John's offer and if Vanessa's potential job offer indeed were to
materialize, should he accept the job offer, or decline it and search for a different summer job by
participating in the corporate summer recruiting program that HKU offers early next year?

31
Bill’s decision criterion

▪ Assume Bill's only criterion on which to differentiate

between summer jobs is the summer salary
– He feels all summer job opportunities would offer him
similar learning, networking, and resumé-building
experiences

32
Decision tree

▪ A decision tree is a systematic way of organizing and

representing the various decisions and uncertainties
that a decision-maker faces

▪ There are two decisions that Bill needs to make

– Whether or not to accept John's summer job offer
– If he were to decline John's offer, and Vanessa's firm were
to offer him a job in November, he must then decide
whether to accept Vanessa's offer or to instead participate
in the school's corporate summer recruiting program

33
Decision tree
Accept John’s Offer

Reject John’s Offer

A

▪ A decision is represented with a box that is called a decision node

▪ Each possible choice is represented as a line called a branch that emanates from the
decision node
▪ If he were to accept John's job offer, then there are no other decisions or uncertainties
Bill would need to consider
▪ If he were to reject John's job offer, then Bill would face the uncertainty of whether or
not Vanessa's firm would subsequently offer Bill a summer job
34
Decision tree
Accept John’s Offer

Reject John’s Offer Offer from Vanessa

A B

No Offer from Vanessa

▪ An uncertain event is represented with a circle called an event node

▪ Each possible outcome of the event is also represented as a branch
▪ The outcome branches must represent a mutually exclusive and collectively
exhaustive set of possible events
▪ At an event node, the decision-maker cannot select which branch to opt for

35
Decision tree
Accept John’s Offer Accept Vanessa’s Offer

Reject John’s Offer Offer from Vanessa Reject Vanessa’s Offer

A B C

No Offer from Vanessa

▪ If Vanessa's firm were to make Bill a job offer, then Bill would subsequently
have to decide to accept or to reject the firm's job offer
▪ In this case, and if Bill were to accept the firm's job offer, then his summer
job problem would be resolved
▪ If Bill were to instead reject their offer, then Bill would then have to search
for summer employment through the school's corporate summer recruiting
program
36
Assigning probabilities
▪ Another aspect of constructing a decision tree is the assignment of
the probability that each of the various uncertain outcomes will occur
▪ Bill has visited the career service center at HKU and learned that on
average, Vanessa's firm would make summer job offers of $14,000 for
MBA students
▪ Bill has also gathered some data on the salary range for all summer
jobs that went to MBA students last year
Total Summer Percentage of Students Who Received This
Pay Salary
$21600 5%
$16800 25%
$12000 40%
$6000 25%
$0 5% 37
Decision tree
Accept John’s Offer Accept Vanessa’s Offer 5%
$21600
25%
Offer from $16800
Reject John’s Reject Vanessa’s
Vanessa 40%
A Offer B C Offer E $12000
5% 25%
$6000
No Offer from $21600
Vanessa 25% 5%
$16800 $0
40%
D $12000
25%
$6000
5% $0

▪ Bill believes the likelihood that he would receive the indicated salaries if he were to
participate in the school's corporate summer recruiting is the similar to other MBA
students last year
▪ An end node (a triangle) indicates that the no further decisions or uncertain events
would emanate from there 38
Decision tree
Accept John’s Offer Accept Vanessa’s Offer 5%
$21600
25%
Reject Offer from Reject Vanessa’s $16800
John’s Offer Vanessa Offer 40%
A B C E $12000
60%
5% 25%
$21600 $6000
No Offer from
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
25%
$6000
5% $0

▪ Bill needs to estimate the likelihood that Vanessa's firm will offer him a job
▪ On reflection, he believes Vanessa was impressed with him, and she sounded certain that she wanted to hire him
▪ However, the competition for investment banking jobs is very intense, and many of Bill's classmates are equally
talented
▪ Bill assigns the probability of receives a job offer from Vanessa's firm to be 60%
39
Decision tree
Accept John’s Offer Accept Vanessa’s Offer
5%
$12000 $14000 $21600
25%
Reject Offer from Reject Vanessa’s $16800
John’s Offer Vanessa Offer 40%
A B C E $12000
60% 25%
5% $6000
No Offer from $21600
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
25%
$6000
5% $0

The next step in the decision analysis modelling methodology is to assign numerical
values to the outcomes associated with the end nodes, based on the decision
criterion that has been adopted

Bill's decision criterion is his summer salary 40

Key characteristics of a decision tree
▪ Time in a decision tree flows from left to right, and the placement of the
decision nodes and the event nodes is logically consistent with the way events
will play out in reality
– Any event or decision that must logically precede certain other events and decisions is
appropriately placed in the tree to reflect this logical dependence
▪ The branches emanating from each decision node represent all of the possible
decisions under consideration at that point in time under the appropriate
circumstances
– The branches emanating from each event node represent a set of mutually exclusive and
collectively exhaustive outcomes of the event node
– The sum of the probabilities of each outcome branch emanating from a given event node
must sum to one
▪ Each and every "final" branch of the decision tree has a numerical value
associated with it. This numerical value usually represents some measure of
monetary value, such as salary, revenue, cost, etc.

41
Expected monetary value (EMV)
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
25%
Reject Offer from Reject Vanessa’s $16800
John’s Offer Vanessa Offer 40%
A B C E $12000
60% 25%
5% $6000
No Offer from $21600
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
25%
$6000
5% $0

▪ At node C of the decision tree, how would he decide between obtaining a summer
salary of $14,000 with certainty, and the distribution of possible salaries he might
obtain from participating in the school's corporate summer recruiting?
▪ A common criterion in this setting is to convert the distribution of possible salaries
to a single numerical value using the EMV of the possible outcomes

42
Expected monetary value (EMV)
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
25%
Reject Offer from Reject Vanessa’s $16800
John’s Offer Vanessa Offer
$11580 40%
A B C E $12000
60% 25%
5% $6000
No Offer from $21600
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
25%
$6000
5% $0

▪ The EMV of an uncertain event is the weighted average of all possible numerical
outcomes, with the probabilities of each of the possible outcomes used as the weights
– The EMV of participating in corporate summer recruiting is:
EMV=0.05×21.6+0.25×16.8+0.40×12+0.25×6+0.05×0=11.58 (thousand $)

43
Expected monetary value (EMV)
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
25%
Reject Offer from Reject Vanessa’s $16800
John’s Offer Vanessa Offer
$11580 40%
A B C E $12000
60% 25%
5% $6000
No Offer from $21600
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
25%
$6000
5% $0

▪ The EMV of a certain event is defined to be the

monetary value of the event
– Suppose that Bill were to receive a job offer from Vanessa's
firm, and that he were to accept the job offer
– The EMV of this choice would simply be $14,000 44
Folding back the decision tree
Accept Vanessa’s Offer 5%
$12000 $14000 $21600
25%
Reject Offer from Reject Vanessa’s $16800
John’s Offer Vanessa Offer
$11580 40%
A B C E $12000
60% 25%
5% $6000
No Offer from $21600
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
25%
$6000
5% $0

▪ Begin with the end nodes of the tree, and then work "backwards" to the
starting node
– Evaluate each event node using the EMV of the event node
– Evaluate each decision node by choosing that decision which has the best EMV
45
Folding back the decision tree
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
25%
Reject Offer from $16800
$13032 $14000 Reject Vanessa’s
John’s Offer Vanessa Offer
$11580 40%
A B C E $12000
60% 25%
$13032 5% $6000
No Offer from $21600
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
$11580 25%
$6000
5% $0

▪ The EMV of node D is the same as node E

▪ Node C is a decision node, having two choices
– Choice 1 has an EMV of $14,000 and choice 2 has an EMV of $11,580
– The EMV of node C is $14,000
▪ Node B is an event node, so its EMV is 0.6×$14,000+0.4×$11,580=$13,032
46
Folding back the decision tree
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
25%
Reject Offer from $16800
$13032 $14000 Reject Vanessa’s
John’s Offer Vanessa Offer
$11580 40%
A B C E $12000
60% 25%
$13032 5% $6000
No Offer from $21600
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
$11580 25%
$6000
5% $0

▪ Remove the inferior choice at each decision node by crossing off the branch

▪ The solution of the decision tree is a decision strategy, which states what decisions should be made
under each possible uncertain outcome that might prevail

47
Bill's optimal decision strategy
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
25%
Reject Offer from $16800
$13032 $14000 Reject Vanessa’s
John’s Offer Vanessa Offer
$11580 40%
A B C E $12000
60% 25%
$13032 5% $6000
No Offer from $21600
25% 5%
Vanessa $16800 $0
40%
40%
D $12000
$11580 25%
$6000
5% $0

▪ Bill should reject John's offer in October

▪ If Vanessa's firm offers him a job, he should accept it
▪ If Vanessa's firm does not offer him a summer job, he should participate in the school's corporate
summer recruiting
▪ The EMV of this strategy is $13,032 (Node A)
48
Procedure for solving a decision tree
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
Reject John’s 25%
Offer from Reject Vanessa’s $16800
Offer $13032 Vanessa $14000 Offer $11580 40%
A B C E $12000
60% 25%
$13032 5% $6000
No Offer from $21600
Vanessa 25% 5%
$16800 $0
40%
40%
D $12000
$11580 25%
$6000
5% $0

▪ Start with the end nodes of the decision tree, and evaluate each event node and each decision node
– For an event node, compute the EMV of the node by computing the weighted
average of the EMV of each branch weighted by its probability
– For a decision node, compute the EMV of the node by choosing that branch
emanating from the node with the best EMV value
▪ The decision tree is solved when all nodes have been evaluated
▪ The EMV of the optimal decision strategy is the EMV computed for the starting branch of the tree 49
Today’s Objectives
1. Use probabilities to determine optimal decisions under uncertainty (based on
expected value criterion) .

2. Set up and solve decision trees - Summer Job Decision

3. Developing a new product case

4. Sensitivity analysis – revisit Summer Job Decision

50
New Product "Suds-Away"

▪ Caroline is the marketing manager for a company that is

considering whether to produce a new automatic dishwashing
detergent called "Suds-Away"

▪ The market for Suds-Away will either be weak or strong

– If the market is strong, the company will make $18 million on Suds-Away
– If the market is weak, the company will lose $8 million

▪ Caroline has estimated that there is a 30% chance that the market
will be strong

51
Market survey test
▪ Prior to deciding whether or not to produce Suds-
Away, Caroline can conduct a nationwide market
survey test of Suds-Away with cost $2.4 million

▪ Such market survey tests cannot predict the market

for new products with certainty
– If the market is weak, there is a 10% chance that the test
will be positive
– If the market is strong, there is a 20% chance that the test
will be negative

52
Caroline's decisions
▪ Should the company not produce Suds-Away?

▪ Or should it go ahead with production without

conducting such a market survey test?

▪ Or should it conduct the market survey test first

(with additional cost) and then decide whether or
not to produce?

53
Decision Tree

54
Decision contingent on the survey results
▪ After she receives the market survey results, Caroline will need to decide
whether or not to produce Suds-Away: node D and node F (decision node and
branches)
▪ The accuracy of the survey and the potential revenue implications of the state
of the market for Suds-Away makes this seemingly obvious decision non-trivial

▪ Suppose that the market survey test outcome is positive and that Caroline
decides to produce Suds-Away
▪ Even though the market survey test is positive in this case, there is still the
possibility that the test will misread the market and that the market for Suds-
Away might be weak rather than strong
▪ Hence, we must place an event node (chances branch), node E, after Caroline's
decision to produce Suds-Away
55
Assigning probabilities

▪ For node B, without taking the market survey, Caroline knows from her
experience that there is a 30% chance that the market for Suds-Away
will be strong and a 70% chance that the market will be weak

▪ However, it is a bit of difficult to determine the probabilities for the

other branches
– E.g., What values are 𝑃1 and 𝑃2 for the branches emanating from node C?

▪ We will derive the unknown probabilities based on probability theory

we learned

56
Probability table
▪ S: "the market for Suds-Away is strong”
▪ W: "the market for Suds-Away is weak”
▪ Q: "the market survey test results are positive"
▪ N: "the market survey test results are negative”
▪ The information we have:
– There is a 30% chance that the market will be strong
– If the market is weak, there is a 10% chance that the test will be positive
– If the market is strong, there is a 20% chance that the test will be negative
𝑃 𝑆 = 0.3, 𝑃 𝑄 𝑊 = 0.1, 𝑎𝑛𝑑 𝑃 𝑁 𝑆 = 0.2

Market is Market is Weak

Total
Strong (S) (W)
Market Test is
? ? ?
Positive
Market Test is ? ?
?
Negative

Total 0.3 0.7 1.00

57
Making decisions
▪ Given the probabilities, we can now solve the decision tree using the folding-back
procedure
– Start with the end nodes of the decision tree, and evaluate
each event node and each decision node
❑ For an event node, compute the EMV of the node by computing
the weighted average of the EMV of each branch weighted by its
probability
❑ For a decision node, compute the EMV of the node by choosing
that branch emanating from the node with the best EMV value
– The decision tree is solved when all nodes have been
evaluated
– The EMV of the optimal decision strategy is the EMV
computed for the starting branch of the tree

58
Caroline’s optimal decision strategy

▪ Caroline should conduct the market survey test?

▪ If the survey results are positive, should she decide to

produce the new product?

▪ If the survey results are negative, should she decide

not to produce the new product?

▪ The EMV of the optimal strategy is ?

59
Value of Information
▪ If the survey is conducted by a consulting company, what is the price
they would offer to Caroline?

60
Today’s Objectives
1. Use probabilities to determine optimal decisions under uncertainty (based on
expected value criterion) .

2. Set up and solve decision trees - Summer Job Decision

3. Developing a new product case

4. Sensitivity analysis – revisit Summer Job Decision

5. Utility Theory

61
Summer Job - decision tree

Accept John’s Offer Accept Vanessa’s Offer 5%

$12000 $14000 $21600
Reject John’s 25%
Offer from Reject Vanessa’s Offer $16800
Offer $13032 Vanessa $14000 $11580 40%
A B C E $12000
60% 25%
$13032 5% $6000
No Offer from Vanessa $21600
25% 5%
$16800 $0
40%
40%
D $12000
$11580 25%
$6000
5% $0

62
Sensitivity analysis of the optimal decision
One must be careful about the key data assumptions
1. The probability that Vanessa's firm would offer Bill a summer job was assumed to
be 60%
How would changes in this probability might affect the optimal decision strategy?

2. The cost of Bill's time and effort in participating in the school's corporate summer
recruiting was assumed to be zero
How high would this implicit cost have to be before the optimal decision strategy would
change?

3. The distribution of summer salaries that Bill could expect to receive was assumed
to be the same as the MBA students last year
How might changes in this distribution of salaries affect the optimal decision strategy?

The process of testing and evaluating how the solution to a decision tree behaves in
the presence of changes in the data is referred to as sensitivity analysis

63
Using Excel spreadsheet for sensitivity analysis

▪ In order to evaluate how the optimal decision strategy behaves as a

function of changes in the data assumptions, we will have to solve and re-
solve the decision tree model many times, each time with slightly different
values of certain data
▪ This approach is tedious and can be expedited with the help of Excel

▪ The data for the decision tree is given in the upper part of the spreadsheet
▪ The "solution" is computed in the lower part in the "EMV of Nodes" table
– The computation of the EMV of each node is performed automatically as a function
of the data

64
Using Excel spreadsheet for sensitivity analysis

Data

▪ EMV of node D (and node E) can be computed using Excel function SUMPRODUCT
▪ EMV of node C = MAX(EMV of node E, 14000)
▪ EMV of node B = 0.6 × (EMV of node C) + (1 - 0.6) × (EMV of node D)
▪ EMV of node A = MAX(EMV of node B, 12000)

65
Sensitivity analysis of the optimal decision
One must be careful about the key data assumptions
1. The probability that Vanessa's firm would offer Bill a summer job was
assumed to be 60%
How would changes in this probability might affect the optimal decision
strategy?

66
Summer Job - decision tree
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
Reject John’s 25%
Offer from Reject Vanessa’s Offer $16800
Offer $13032 Vanessa $14000 $11580 40%
A B C E $12000
60% 25%
$13032 5% $6000
No Offer from Vanessa
p $21600
25% 5%
$16800 $0
40%
40%
D $12000
$11580 25%
$6000
5% $0

67
Sensitivity relative to key data issue 1
Data

▪ Denote the probability that Vanessa's firm will offer Bill a summer job by p
▪ Vary the value of p to see when the optimal strategy changes

68
Sensitivity relative to key data issue 1
Data

▪ EMVs of node A and node B remain equal to each other while we reduce p from 60%
to 18%
▪ EMV of node A being equal to EMV of node B means the optimal decision is to reject
John's offer

69
Sensitivity relative to key data issue 1
Data

▪ EMV of node A becomes to be greater than EMV of node B for p is 17%

▪ This means the optimal decision is to accept John's offer
▪ Hence, as long as p ≥ 18%, then the optimal decision strategy will still be to reject
John's offer and to accept a summer job with Vanessa's firm if they offer it to him
– This is reassuring, as Bill is very confident that the probability
of Vanessa's firm offering him a summer job is surely greater
than 18% 70
Sensitivity analysis of the optimal decision
One must be careful about the key data assumptions
1. The probability that Vanessa's firm would offer Bill a summer job was assumed to
be 60%
How would changes in this probability might affect the optimal decision strategy?

2. The cost of Bill's time and effort in participating in the school's corporate summer
recruiting was assumed to be zero
How high would this implicit cost have to be before the optimal decision strategy would
change?

71
Summer Job - decision tree
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
Reject John’s 25%
Offer from Reject Vanessa’s Offer $16800
Offer $13032 Vanessa $14000 $11580 40%
A B C E $12000
60% 25% -c
$13032 5% $6000
No Offer from Vanessa $21600
25% 5%
$16800 $0
40%
40%
D $12000
-c
$11580 25%
$6000
5% $0

72
Sensitivity relative to key data issue 2
Data

▪ Denote implicit cost to Bill of participating in the

school's corporate summer recruiting program by c
▪ Vary the value of c to see when the optimal strategy
changes 73
Sensitivity relative to key data issue 2
Data

▪ The cost c should now be subtracted from the previous EMV of node D
EMV of node D = 0.05×21600 + 0.25×16800 + 0.4×12000 + 0.25×6000 + 0.05×0 − c
▪ EMV of node E is treated in the same way
▪ EMVs of node A and node B are equal for c ≤ 2580

74
Sensitivity relative to key data issue 2
Data

▪ EMV of node A becomes to be greater than EMV of node B for c ≥ 2581

▪ Hence, as long as c ≤ 2580, then the optimal decision strategy will still be to reject
John's offer and to accept a summer job with Vanessa's firm if they offer it to him
– This is reassuring, as the implicit cost to Bill of participating in the school's corporate
summer recruiting program is much less than $2,580

75
Sensitivity analysis of the optimal decision
One must be careful about the key data assumptions
1. The probability that Vanessa's firm would offer Bill a summer job was assumed to be 60%
How would changes in this probability might affect the optimal decision strategy?

2. The cost of Bill's time and effort in participating in the school's corporate summer recruiting was
assumed to be zero
How high would this implicit cost have to be before the optimal decision strategy would change?

3. The distribution of summer salaries that Bill could expect to receive was assumed to be the same as the
MBA students last year
How might changes in this distribution of salaries affect the optimal decision strategy?

The process of testing and evaluating how the solution to a decision tree behaves in the presence of
changes in the data is referred to as sensitivity analysis

76
Summer Job - decision tree
Accept John’s Offer Accept Vanessa’s Offer 5%
$12000 $14000 $21600
Reject John’s 25%
Offer from Reject Vanessa’s Offer $16800
Offer $13032 Vanessa $14000 $11580 40%
A B C E $12000
60% 25% +S
$13032 5% $6000
No Offer from Vanessa $21600
25% 5%
$16800 $0
40%
40%
D $12000
+S
$11580 25%
$6000
5% $0

77
Sensitivity relative to key data issue 3
Data

▪ Suppose we modify all of the possible salary offers by

an amount S
▪ Vary the value of S to see when the optimal strategy
changes 78
Sensitivity relative to key data issue 3
Data

▪ Each possible summer salary is added by S

▪ Recall that EMV of node C = MAX(EMV of node E, 14000)
▪ EMV of node C is equal to 14,000 for S ≤ 2420
▪ EMV of node C being equal to 14000 means that Bill should accept an offer from
Vanessa's firm if it materialized

79
Sensitivity relative to key data issue 3
Data

▪ EMV of node C becomes to be greater than $14,000 for S ≥ 2421

▪ This means Bill should reject an offer from Vanessa's firm if it materialized
▪ Hence, as long as S ≤ 2420, then the optimal decision strategy will still be to accept an
offer from Vanessa's firm if it materialized
– This is reassuring, as summer salaries from corporate summer recruiting in general would
not increase by $2,421 within a year

80
Summary of sensitivity analysis
▪ We have explored three data issues
– The probability p of Vanessa's firm offering Bill a summer job
– The implicit cost c of participating in corporate summer recruiting
– An increase S in all possible salary values from corporate summer recruiting

▪ We have found that the optimal decision strategy does not change
unless the three quantities take on unreasonable values

▪ Hence, it is safe to proceed with confidence in recommending to Bill

that he adopt the optimal decision strategy found in the solution to
the decision tree model
– He should reject John's job offer, and he should accept a job offer from
Vanessa's firm if such an offer is made

81
Principal steps of decision analysis
1. Structure the decision problem
– List all of the decisions that have to be made
– List all of the uncertain events in the problem and all of their possible outcomes
2. Construct the basic decision tree by placing the decision nodes and the event nodes in their chronological and
logically consistent order
3. Determine the probability of each possible outcome of each uncertain event
4. Determine the numerical values of each end node of the decision tree
5. Solve the decision tree using the folding-back procedure
– Start with the end nodes of the decision tree, and evaluate each event node and each
decision node
– The decision tree is solved when all nodes have been evaluated
– The EMV of the optimal decision strategy is the EMV computed for the starting
branch of the tree
6. Perform sensitivity analysis on all key data values
– For each data value for which the decision-maker lacks confidence, test how the
optimal decision strategy will change relative to a change in the data value, one data
value at a time
82