Decision Theory

Decision Theory represents a general

approach to decision making which is suitable for a
wide range of operations management decisions,
including:
product and
service design
equipment
selection
location
planning
Decision Theory
Capacity
planning
A set of possible future conditions exists that will
have a bearing on the results of the decision
A list of alternatives for the manager to choose
from
A known payoff for each alternative under each
possible future condition
Decision Theory Elements
Identify possible future conditions called states of
nature
Develop a list of possible alternatives, one of
which may be to do nothing
Determine the payoff associated with each
alternative for every future condition
Decision Theory Process
If possible, determine the likelihood of
each possible future condition
Evaluate alternatives according to some
decision criterion and select the best
alternative
Decision Theory Process (Contd)
Bounded Rationality
The limitations on decision
making caused by costs,
human abilities, time, technology, and
availability of information
Causes of Poor Decisions
Suboptimization
The result of different
departments each
attempting to reach a
solution that is
optimum for that
department
Causes of Poor Decisions (Contd)
Certainty - Environment in which
relevant parameters have known values
Risk - Environment in which certain
future events have probable outcomes
Uncertainty - Environment in which it is
impossible to assess the likelihood of
various future events
Decision Environments
ONE STAGE DECISIONMAKING PROBLEMS
Maximin or Minimax- Choose the alternative with
the best of the worst possible payoffs
Maximax or Minimin- Choose the alternative with
the best possible payoff
Laplace - Choose the alternative with the best
average payoff of any of the alternatives
Minimax Regret or Savage Principle- Choose the
alternative that has the least of the worst regrets
Hurwicz Principle- Stipulates the decision makers
view.

Decision Making under Uncertainty
Problem 1
A book store sells a particular monthly for
Rs.100.He purchases the magazine for Rs.
80.Unsold copies can be disposed of for Rs.30.
Weekly demand varies as per the follows six
events:
E1 : 18 copies E2 :19 copies
E3 :20 copies E4 :21 copies
E5 :22 copies E6 :23 copies
How many copies of the magazine should be
purchased for
next week?
Developing a Pay-off Table

Acts
Events
A1:18 A2:19 A3:20 A4:21 A5:22 A6:23
E1:18 360 310 260 210 160 110
E2:19 360 380 330 280 230 180
E3:20 360 380 400 350 300 250
E4:21 360 380 400 420 370 320
E5:22 360 380 400 420 440 390
E6:23 360 380 400 420 440 460
Developing Regret Table(OL
table)

Acts
Events
A1:18 A2:19 A3:20 A4:21 A5:22 A6:23
E1:18 0 50 100 150 200 250
E2:19 20 0 50 100 150 200
E3:20 40 20 0 50 100 150
E4:21 60 40 20 0 50 100
E5:22 80 60 40 20 0 50
E6:23 100 80 60 40 20 0
Laplace or Equally Likely
Principle
Laplace principle is based on the principle that if
we are uncertain about various events than we
treat them equally probable.
Under his assumption the expected mean value
of pay-off of each strategy is determined and
strategy with highest mean value is adopted.
If the pay-offs are cost ,strategy with lowest mean
is adopted.
Problem 1
Act Mean Pay-Of
A1 Rs.360.00
A2 Rs.368.30
A3 Rs.365.00
A4 Rs.350.00
A5 Rs.323.30
A6 Rs.285.00

A2 is maximum, it would be adopted.
Maximin or Minimax Principle
Using this approach ,the minimum pay-off
resulting from each strategy is considered and
the maximum among the values is selected.
Choosing the best(maximum) profit amongst the
set of worst (minimum)profit.
It is a pessimistic approach.
In case of lose best(minimum) loss amongst the
set of worst(maximum) losses, Minimax Principle.
Problem 1
Act Minimum Pay-Off
A1 Rs.360.00
A2 Rs.310.00
A3 Rs.260.00
A4 Rs.210.00
A5 Rs.160.00
A6 Rs.110.00

A1 is maximum, it would be adopted.
Maximax or Minimin Principle
I t is a optimistic principle. It suggests that for
each strategy maximum profit should be
considered and strategy with highest of these
values is associated should be chosen.
For each strategy minimum loss should be
considered and strategy with minimum of these
values is associated should be chosen
Problem 1
Act Maximum Pay-Off
A1 Rs.360.00
A2 Rs.380.00
A3 Rs.400.00
A4 Rs.420.00
A5 Rs.440.00
A6 Rs.460.00

A6 is maximum, it would be adopted.

Minimax Regret or Savage
Principle
This principle is based on the
concept of regret and calls for
the selecting the course of
action that minimize the
maximum regret.
Problem 1
Act Maximum Regret
A1 Rs.100.00
A2 Rs. 80.00
A3 Rs.100.00
A4 Rs.150.00
A5 Rs.200.00
A6 Rs.250.00

A2 has least regret value, it would be adopted.


Hurwicz or Principle of Realism
Decision makers view may fall
somewhere between extreme
pessimism (Maximin) to extreme
Optimism( Maximax).This principle
provides a mechanism by which
different level of optimism and
pessimism may be shown. For this an
index of optimism , is defined on scale
ranging from 0 to 1.An =0 means
Maximin Principle( pessimism) =1
means Maximax(Optimism).
Problem 1
ACT Max Min ( max vale)+(1-)(min value) Value
A1 360 360 360X0.6+360X0.4 360
A2 380 310 380X0.6+310X0.4 352
A3 400 260 400X0.6+260X0.4 344
A4 420 210 420X0.6+210X0.4 336
A5 440 160 440X0.6+160X0.4 328
A6 460 110 460X0.6+110X0.4 320
Suppose that Index of optimism =0.6
A1 is maximum, it would be adopted.
Decision Making Under Risk
The decision situation wherein the decision
maker chooses to consider several possible
outcome and probabilities of their occurrence can
be stated. Probabilities of various outcomes may
be determined from past record

Acts
Events
Probability

P
A1:18 A2:19 A3:20 A4:21 A5:22 A6:23
E1:18 0.05 360 310 260 210 160 110
E2:19 0.10 360 380 330 280 230 180
E3:20 0.30 360 380 400 350 300 250
E4:21 0.40 360 380 400 420 370 320
E5:22 0.10 360 380 400 420 440 390
E6:23 0.05 360 380 400 420 440 460
Maximum Likelihood Principle
Under this principle the event that most likely to
occur is considered. Then the act is decided for
the maximum conditional pay off.
A1: Rs.360
A2: Rs.380
A3: Rs.400
A4: Rs.420
A5: Rs.370
A5: Rs.320
A4 is maximum, it would be adopted
Expectation Principle
Expected Monetary Value(EMV):EMV for an act is
computed as the weighed average of all possible payoffs
of that act.

i.e. EMV=Pay-off of first outcome X Probability of first
outcome
+Pay-off of 2
nd
outcome X Probability of 2
nd

outcome
+Pay-off of 3
rd
outcome X Probability of 3
rd

outcome
+. .+
+Pay-off of last outcome X Probability of last
outcome.

Acts
Events
Probability

P
A1:18 A2:19 A3:20 A4:21 A5:22 A6:23
E1:18 0.05 360 310 260 210 160 110
E2:19 0.10 360 380 330 280 230 180
E3:20 0.30 360 380 400 350 300 250
E4:21 0.40 360 380 400 420 370 320
E5:22 0.10 360 380 400 420 440 390
E6:23 0.05 360 380 400 420 440 460
Expecte
d
Pay-off
360 376.5 386 374.5 335 288.5
A3 is maximum, it would be adopted
Expected Opportunity Loss
(EOL):
EOL=Regret of first outcome X Probability of first
outcome
+Regret of 2
nd
outcome X Probability of 2
nd

outcome
+Regret of 3
rd
outcome X Probability of 3
rd

outcome
+. .+
+Regret of last outcome X Probability of last
outcome.


Acts
Events
Probabi
lity

P
A1:18 A2:19 A3:20 A4:21 A5:22 A6:23
E1:18 0.05 0 50 100 150 200 250
E2:19 0.10 20 0 50 100 150 200
E3:20 0.30 40 20 0 50 100 150
E4:21 0.40 60 40 20 0 50 100
E5:22 0.10 80 60 40 20 0 50
E6:23 0.05 100 80 60 40 20 0
Expecte
d Regret
51 34.5 25 36.5 76 122.5
A1 has minimum regret, it would be adopted
Expected Value of Perfect Information(EVPI)
& Expected Value with Perfect
Information(EVwPI)
EVwPI=
Best Pay-off of first outcome X Probability of first
outcome
+Best Pay-off of 2
nd
outcome X Probability of 2
nd

outcome
+ Best Pay-off of 3
rd
outcome X Probability of 3
rd

outcome
+. .+
+Best Pay-off of last outcome X Probability of last
outcome.

EVPI=EVwPI-Max EMV


Acts
Events
Probability

P
A1:18 A2:19 A3:20 A4:21 A5:22 A6:23
E1:18 0.05 360 310 260 210 160 110
E2:19 0.10 360 380 330 280 230 180
E3:20 0.30 360 380 400 350 300 250
E4:21 0.40 360 380 400 420 370 320
E5:22 0.10 360 380 400 420 440 390
E6:23 0.05 360 380 400 420 440 460
EMV 360 376.5 386 374.5 335 288.5
EVwPI=360X0.05+380X0.10+400X0.30+420X0.40+440X0.10+460X0.05=411
EVPI=411-386=25
Problem 1
Technico Ltd has installed a m/c costing Rs 4
Lacs.The spare parts cost Rs. 4000 each but are
available only if ordered in advance.In case the m/c
fails and no spare is available cost to the company
would be Rs. 18000.The plant has a estimated life
of 8 years and the probability distribution during this
period is as follows
No of failures 0 1 2 3 4 5
6
Probability 0.1 0.2 0.3 0.2 0.1 0.1 0
Ignoring time value of money
Calculate 1:Optimal no. of the spare on the basis of
a)Maximin,Maximax,Laplace,Hurwicz( 0.7),
2) EMV,EVPI
3)The regret table


Multistage Decision Making (Decision Tree)
B
Payoff 1
Payoff 2
Payoff 3
2
Payoff 6
2
Payoff 4
Payoff 5
1
Decision Point
Chance Event
A graphic representation of the sequences of actions-event combination available to the decision maker.
A firm owner is considering of drilling a farm
well . In the past only 70% of wells drilled
were successful at 200 ft. Moreover on
finding no water at 200 ft some drilled it
further 50 ft but only 20% struck water at
250 ft. The prevailing cost of drilling is Rs.
50 per feet. The farm owner has estimated
that in case he does not get his own wells
he will have to pay Rs. 15000 over next 10
years to buy water from neighbor.



Problem 1
Problem 2
An oil company has recently acquired rights in a certain area to conduct
surveys and test drillings to lifting oil if it is found in commercially exploitable
quantities.
On the known conditions, the company estimates that there is 70:30 chances of
tests showing success.
If successful test is carried out expectation of success in drilling is 80:20.e been
If test indicates failure expectation of success drilling is 20:80
If no test have been carried out expectation of success drilling is 20:80
Company has the option of testing,driiling without testing or selling rights to
other company.
Cost and revenues have been estimated for all possible outcomes and the net
present value of each is as follows
Outcome NPV(in millions)
Success With prior test 100
Without prior test 120

Failure With prior test -50
Without prior test -40
Sale of exploitation rights
Prior test shows success 65
Prior test shows failure 15
Without prior test 45



Problem 1
Test
3
Problem 1
Test
-40
120
3
Problem 1
Test
-40
120
2
65
3
Problem 1
Test
-40
120
-50
2
65
3
100
1
Problem 1
Test
-40
120
-50
2
65
3
100
15
1
Problem 1
Test
-40
120
-50
2
65
3
100
15
1
-50
100
Node 1
Outcome Prob. Conditional Value Expected Value
1. Drill Success 0.2 100 20
Failure 0.8 -50 -40
Total -20
2. Sell 1.0 15 15
Node 2
Outcome Prob. Conditional Value Expected Value
1. Drill Success 0.8 100 80
Failure 0.2 -50 -10
Total 70
2. Sell 1.0 65 65
Node 3
Outcome Prob. Conditional Value Expected Value
1. Drill Success 0.55 120 66
Failure 0.45 -40 -18
Total 48

2. Test Positive 0.70 70 49
Negative 0.30 15 4.5
Total 53.5




3. Sell 1.0 45 45
Problem 1
Test
-40
120
-50
2
65
53.5
100
15
1
-50
100
48
53.5
70
15
70