Decision Analysis:

Choice of the Best Alternative

Alamanda
MBTI

1
 Course Objectives

 Understanding how to set “what is

problem”;
 Ability to choose the best alternative among
available alternatives by:
 Identifying their consequences;
 Asking experts, especially if there are multi
criteria/attributes.
 Ability to analyze confrontation;
 Enjoy and survive during this course!
Problems of Choice

C1 O1
D
C2 O2

 D : a decision maker
 C : possible courses of action
(alternatives)
 O1 : desirable outcome; O2 : undesirable
outcome
 Model of Decision Analysis

E1,1 p1,1
O1
C1 p1,2
E1,2 O2
D O1
E2,1 p2,1
C2
E2,2 O2
p2,2
 D : a decision maker
 C : possible courses of action (alternatives)
 O1 and O2 : possible outcomes/consequences/payoffs
 Ei,j : Events (State of Natures/SON)
 pi,j : probabilities
Structure of Decision Tree
 Decision Node
 Alternatives available for decision maker to
choose;
 Situation controllable by decision maker.

Alternatives of actions
Structure of Decision Tree
 Event Node
 Events may happen after every action made
by decision maker;
 Uncontrollable by decision maker;
 Decision maker only has information about
probability of each event  no complete
information.
p1
E1

E2 p2
An action
E3
p3

Events
Building Decision Tree

1. Identify what decisions should be made by

DM;
 What are the first decision, and next
decisions to be made?
2. Identify what SON happen after each
decision;
3. Draw decision node and event node (SON);
4. Complete information about probabilities;
5. Complete information about payoff.
Goferbroke Company Case
 Goferbroke Company(1)
 Max Flyer is the founder of and sole owner of the Goferbroke
Company, which develops oil wells in unproven territory. Max’s
friends refer to him affectionately as a wildcatter. However he
prefers to think himself as an entrepreneur. He has poured his life
saving’s into the company in the hope of making it big with a large
strike of oil.
 Now his chance possibly has come. His company has purchased
various tracts of land that larger oil companies have spurned as
unpromising even though they are near some large oil fields. Now
Max has received an exciting report about one of these tracts. A
consulting geologist has just informed Max that he believes there
is one chance in four of oil there.
 Max has learned from bitter experience to be skeptical about the
chances of oil reported by consulting geologist. Drilling for oil on
this tract would require an investment of about $100,000. If the
lands turns out to be dry (no oil), the entire investment would be
lost. Since his company doesn’t have much capital left, this lost
would be quite serious.
Goferbroke Company(2)
 On the other hand, if the tract does contain oil, the consulting
geologist estimates that there would be enough there to
generate a net revenue of approximately $800,000, leaving an
approximate profit of:
 Profit if find oil = Revenue if find oil – Drilling cost
= $800,000 - $100,000
= $700,000

 There is another option that another oil company has gotten

wind of consulting geologist’s report and so has offered to
purchase the tract of land from Max for $90,000. This is very
tempting. This too would provide a welcome infusion of capital
into the company, but without incurring the large risk of a very
substansial loss of $100,000.
Decision Trees of Goferbroke

 Decision tree is decision making

help tools that could describe entire
alternatives with whole events that
may happen (SoN).
 Showing : Alternatives, SoN, Prior
Probability, and Payoff.
 Using Bayes’ Decision Rule to choose
the best action.
DT of Goferbroke’s Case
 Decision:
 Drill or Sell the Land

 SON
 Oil or Dry

Decision nodes Drill

-100

Sell

90
DT of Goferbroke’s Case
 Event nodes
Oil 0.25
800
Drill
-100 Dry 0,75
0
Sell

90
DT of Goferbroke’s Case
 Payoff
Oil 0.25
700
800
Drill

-100 Dry 0,75

-100
0
Sell
90
90
DT of Goferbroke’s Case
 Payoff
Oil 0.25
700
800
Drill

-100 100 Dry 0,75

-100
0
Sell
100
90
90 90

 Expected Value (EV) per event node;

100=(0.25*700) + (0.75*(-100))
 Using Treeplan software

0.25
Event node Oil
700
Drill 800 700 100=(0.25*700) + (0.75*(-100))

Decision node -100 100 0.75 Expected payoff

Dry = MAX [100,90]
-100 = 100
1 0 -100
100
Action: Drill
Sell
90
90 90
Goferbroke’s Case Continued
 Survey by geologist will provide more
accurate information about P(oil);
 How if Max has to decide two alternatives:
1. Do survey before drill/sell
2. Drill/sell without Survey
 Events:
 Do Survey
 FSS : Favorable Seismic Sounding : Oil is fairly likely
 USS : Unfavorable seismic sounding: Oil is quite unlikely.
 Drill or Sell
 Oil
 Dry
Max`s Experience
•P(state) prior; which is P(Oil)=0.25 & P(Dry)=0.75;
•P (finding|state) being known based on Max’s
experiences; which is
•P(FSS|Oil)=0.6,
•P(USS|Oil)=0.4,
•P(FSS|Dry)=0.2, and
•P(USS|Dry)=0.8

Which:
•State : Oil and Dry;
•Finding : FSS and USS;
•FSS : favorable seismic sounding; oil is fairly likely;
•USS : unfavorable seismic sounding; oil is quite unlikely.
Leveled Decision Analysis
Decision T ree for Goferbroke Co. Problem (With Survey)
0,143
Oil
P(Oil|USS)
P(USS) Drill 800 670
670

0,7
-100 -15,714 0,857
Dry P(Dry|USS)
Unfavorable -130
2 0 -130
0 60

Sell
60
90 60
Do Survey
0,5 P(Oil|FSS)
-30 123 Oil

P(FSS) Drill 800 670

670

-100 270 0,5 P(Dry|FSS)

0,3 Dry
Favorable -130
1 0 -130
0 270

1 Sell
123 60
90 60

0,25
Oil P(Oil)
700
Drill 800 700

-100 100 0,75

Dry P(Dry)
No Survey -100
1 0 -100
0 100

Sell
90
90 90
Decision Tree:
What-If Analysis
Sensitivity Analysis

The Study of how different

assumptions about future
(parameters) would affect the
recommended decision.
 Current Solution
Decision T ree for Goferbroke Co. Problem (With Survey)
0,143
Oil
670
Drill 800 670

-100 -15,714 0,857

0,7 Dry
Unfavorable -130
2 0 -130
0 60

Sell
60
90 60
Do Survey
0,5
-30 123 Oil
670
Drill 800 670

-100 270 0,5

0,3 Dry
Favorable -130
1 0 -130
0 270

1 Sell
123 60
90 60

0,25
Oil
700
Drill 800 700

-100 100 0,75

Dry
No Survey -100
1 0 -100
0 100

Sell
90
90 90

Do survey, if USS then Sell, else Drill

 Sensitivity Analysis Graph

 There are three Sensitivity Analysis tools that being

depicted in graph to help analyze Decision Tree
solution:
 Plot graph
 Spider graph
 Tornado graph
 All of these graphs are built by Sensit-Sensitivity
Analysis software.
Sensitivity Analysis Graph Tips

 Plot graph were sensitivity analysis tools that being

used to investigating the effect of prior probability of
finding oil on the expected payoff;
 Spider graph were sensitivity analysis tool that being
used to investigating how the expected payoff would
change if one of parameters changes (the cost or
revenues) by up to plus or minus 10%;
 Tornado diagram are being used to analyze different
parameters that had different degrees of variability on
the expected payoff.
 Sensitivity Analysis(1)
Decision T ree for Goferbroke Co. Problem (With Survey)
0,143
Oil
670
Drill 800 670

-100 -15,714 0,857

0,7 Dry
Unfavorable -130
2 0 -130
0 60

Sell
60
90 60
Do Survey
0,5
-30 123 Oil
670
Drill 800 670

-100 270 0,5

0,3 Dry
Favorable -130
1 0 -130
0 270

1 Sell
123 60
90 60

0,25
Oil
700
Drill 800 700

-100 100 0,75

Dry
No Survey -100
1 0 -100
0 100

Sell
90
90 90

Effect of P(oil) and P(dry) on expected payoff

 Sensitivity Analysis(2)
Decision T ree for Goferbroke Co. Problem (With Survey)
0,143
Oil
670
Drill 800 670

-100 -15,714 0,857

0,7 Dry
Unfavorable -130
2 0 -130
0 60

Sell
60
90 60
Do Survey
0,5
-30 123 Oil
670
Drill 800 670

-100 270 0,5

0,3 Dry
Favorable -130
1 0 -130
0 270

1 Sell
123 60
90 60

0,25
Oil
700
Drill 800 700

-100 100 0,75

Dry
No Survey -100
1 0 -100
0 100

Sell
90
90 90

Effect of revenue and costs on expected payoff

 Sensit-Sensitivity Analysis-Plot
SensIt - Sensitivity Analysis - Plot

700
0.168 0.308
Expected Payoff

600
500
400
300
200
100
0
0 0,2 0,4 0,6 0,8 1
Prior Probability Of Oil

Let p = prior probability of oil;

• If p ≤ 0.168, then sell the land (no seismic survey);
• If 0.169 ≤ p ≤ 0.308, then do the survey,
•drill if favorable and sell if not;
• If p ≥ 0.309, then drill for oil (no seismic survey)
 Sensit-Sensitivity Analysis-Spider
Sensit - Sensitivity Analysis - Spider
Most Sensitive
136
134
132 Current Situation
130
Expected Payoff Value

128
126 Cost of Survey
124 Cost of Drilling
122 Revenue if Oil
120 Revenue if Sell

118
116
114
112
110
90% 92% 94% 96% 98% 100% 102% 104% 106% 108% 110%
% Change in Input Value

Conclusion : Payoff more influenced by Revenue if Oil variable than others.

Sensit-Sensitivity Analysis-Tornado
Sensit - Sensitivity Analysis - Tornado

Revenue if Oil 600 800 1000

Cost of Drilling 140 100 75

Revenue if Sell 85 90 95

Cost of Survey 32 30 28

90 100 110 120123 130 140 150 160

Expected Payoff

Conclusion : Although revenue if oil 25% reduced, but expected payoff still > $ 90,000
(Drill is robust)
Thanks
Posterior Probability

 Given:
 P(state)=prior probability: P(oil) and P(dry)
 P(finding|state) = Max’s experience on probabilities of finding (FSS or
USS) could occur if some SoN (oil or dry) has been already happened.

P(FSS|oil) 0.25*0.6=0.15 0.15/0.3=0.5 P(oil|FSS)

Oil and FSS Oil, diket FSS
0.6 et Oil
, di
k P(Oil and FSS) P(state|finding)
F SS Prob. posterior
US 0
P(oil) S, .4 P(Oil and USS)
dik
et O
25 il 0.25*0.4=0.1 0.1/0.7=0.14
P(oil|USS)
0. il P(USS|oil) Oil and USS Oil, diket USS
O

P(finding and state)

0.
Dr 75 P(FSS|dry) 0.75*0.2=0.15 0.15/0.3=0.5
P(dry|FSS)
y Dry and FSS Dry, diket FSS
0.2 et Dry P(Dry and FSS)
P(dry) dik
S,
FS
US 0.
S , di 8 P(Dry and USS)
ket
Dry 0.75*0.8=0.6 0.6/0.7=0.86
P(USS|dry) Dry and USS Dry, diket USS
P(dry|USS)
 0,143
Oil
670
Drill 800 670

-100 -15,6 0,857

0,7 Dry
Unfavorable -130
2 0 -130
0 60

Sell
60
90 60
Do Survey
0,5
-30 123 Oil
670
Drill 800 670

-100 270 0,5

0,3 Dry
Favorable -130
1 0 -130
0 270

1 Sell
123 60
90 60

0,25
Oil
700
Drill 800 700

-100 100 0,75

Dry
No Survey -100
1 0 -100
0 100

Sell
90
90 90
 Posterior Probability Formula
Given :
• P(state) prior probability: P(Oil)=0.25 & P(Dry)=0.75;
• P(finding|state) P(FSS|Oil)=0.6, P(USS|Oil)=0.4,
P(FSS|Dry)=0.2, and P(USS|Dry)=0.8.
Then :
• P(State and Finding)=P(State)P(Finding|State)
• such that : P(Oil and FSS)=P(Oil)P(FSS|Oil)
P( state and finding )
P( state | finding ) 
P( finding )

And P(finding):
•P(FSS)=P(oil and FSS)+P(dry and FSS)
•P(USS)=P(oil and USS)+P(dry and USS)
GRADE MALE FEMALE TOTAL

A 30 40 70

B 10 20 30

TOTAL 40 60 100

P B   P  A 
30 70
100 100
GRADE MALE FEMALE TOTAL

A 30 40 70

B 10 20 30

TOTAL 40 60 100

PFemale  
60
PMale  
40
100 100
GRADE MALE FEMALE TOTAL

A 30 40 70

B 10 20 30

TOTAL 40 60 100

P A and Male   PMale and A 

30
100
PB and Male   PMale and B  
10
100
GRADE MALE FEMALE TOTAL

A 30 40 70

B 10 20 30

TOTAL 40 60 100

PB and Male 

10
PB Male  
1
 100 
PMale  40 4
100
GRADE MALE FEMALE TOTAL

A 30 40 70

B 10 20 30

TOTAL 40 60 100

PMale and B 
10
PMale B  
1
 100 
P B  30 3
100
 GRADE MALE FEMALE TOTAL

A 30 40 70

B 10 20 30

TOTAL 40 60 100

PB   PB and Male   PB and Female 

10 20 30
  
100 100 100
 GRADE MALE FEMALE TOTAL

A 30 40 70

B 10 20 30

TOTAL 40 60 100

PMale   P A and Male   PB and Male 

30 10 40
  
100 100 100
 Max`s Preference on Risk

 According to Max, loss as much as 130

thousands dollar in current difficult time gives
deadly effect to his company;
 But, in normal condition, loss as much as 130
thousands dollar can be handled easily.

1. Pay off in unit money can not accommodate DM’s preference on risk
2. It is needed to consider DM`s preference on risk into payoff
3. Utility function accommodates DM’s on risk by transforming payoff in unit money
onto utility value
 Utility Function for Money

U(M) U(M)

U(M)
m M M
(a) Risk averse (c) Risk neutral

m M
(b) Risk seeking
 Dealing with the Goferbroke Co.
Problem
 Let according to Max U(0)=0, and U(-130)=-150, then calculate
U(700);
 Give two choices to Max:
 A1: Obtain a payoff of 700 with probability p.
Obtain a payoff of –130 with probability (1-p);
 A2 : Definitely obtain payoff of 0.
 Ask Max to choose value of p such that he will be
indifferent over A1 and A2;
 If Max has chosen p=0.2 then
 E(A1)=E(A2);
 0.2U(700)+(1-0.2)U(-130)=0
 U(700)=600
 Transform other payoffs: U(-100)=-105, and U(90)=90, etc.
Max’s Utility Function for Money

Money M Max’s Utility U(M)

-130 -150
-100 -105
0 0
60 60
90 90
670 580
700 600
Revision of DT using Max’s Utility
Function for Money
Decision T ree for Goferbroke Co. Problem (With Max's Utility Function)
0,143
Oil
580
Drill 580 580

0 -45,61 0,857
0,7 Dry
Unfavorable -150
2 -150 -150
0 60

Sell
60
60 60
Do Survey
0,5
0 106,5 Oil
580
Drill 580 580

0 215 0,5
0,3 Dry
Favorable -150
1 -150 -150
0 215

1 Sell
107 60
60 60

0,25
Oil
600
Drill 600 600

0 71,25 0,75
Dry
No Survey -105
2 -105 -105
0 90

Sell
90
90 90
 Calculation of Utility Function for
money

 Fundamental property : if a decision maker is

indifferent over two alternatives then the
alternatives have the same expected utility.