CHAPTER FOUR

DECISION THEORY
4.1. INTRODUCTION
Decision theory represents a generalized approach to decision making, which often
serves as the basis for a wide range of managerial decision making. The decision
model includes a list of courses of action that are available and the possible
consequences of each course of action. An important factor in making a decision is
the degree of certainty associated with consequences. This can range anywhere
from complete certainty to complete uncertainty, and it generally affects the way a
decision is reached. Therefore, decision theory provides a rational approach to the
mangers in dealing with problems confronted with partial, imperfect or uncertain
future condition.
Decision theory problems are characterized by the following:
1. A list of alternatives
2. A list of possible future states of nature
3. Payoffs associated with each alternative /state of nature combination
4. An assessment of the degree of certainty of possible future events
5. A decision criterion
a) List of Alternatives
The list of alternatives is a set of mutually exclusive and collectively exhaustive
decisions that are available to the decision maker. (Sometimes, but not always, one
of these alternatives will be to “do nothing”). These alternatives are also termed as
course of actions or simply actions or acts or strategies.
For example, suppose that Tabor Ceramic Manufacturing Company will decide to
increase its production to meet the increasing market demand of ceramic products.
After careful consideration, the company has ruled out “do nothing” and left with
the following list of acceptable alternatives.
 Expand the present plant
 Construct a new plant
 Subcontract production for extra demand
b) States of nature/outcomes
States of nature refer to a set of possible future conditions or events, beyond the
control of the decision maker that will be the primary determinants of the eventual
consequence of the decision. The states of nature, like the list of alternatives, must
be mutually exclusive and collectively exhaustive. Of course, out of this list, only
one of the events will occur.
For example, in the case of Tabor Ceramic Manufacturing Company considered
above the greatest uncertainty will be about product demand. Hence, the future
events/states of nature related to the demand may be:
1. High demand
2. Moderate demand
3. Low demand
4. No demand

1
 c) Payoffs/conditional gain
In order for a decision maker to be able to rationally approach a decision problem, it
is necessary to have some idea of the payoffs that would be associated with each
alternative and the various states of nature. The payoffs might be profits, revenues,
costs, or other measure of value. Usually the measures are financial. They may be
weekly, monthly, or annual amount, or they might represent present values of
future cash flows. Usually, payoffs are estimated values. The more accurate these
estimates, the more useful they will be for decision making purposes and the more
likely it is that the decision maker will choose an appropriate alternative. The
number of payoffs depends on the number of alternative/state of nature
combinations.
d) Degree of Certainty
The approach used by a decision maker often depends on the degree of certainty
that exists. There can be different degrees of certainty. One extreme is complete
certainty and the other is complete uncertainty. Complete uncertainty exists when
the likelihood of the various states of nature is unknown. Between, these two
extremes is risk, a term that implies that probabilities are known for the states of
nature.
Knowledge of the likelihood of each of the states of nature can play an important
role in selecting a course of action. Thus, if a decision maker feels that a particular
state of nature is highly likely, this will mean that the payoffs associated with that
state of nature are also highly likely. This enables the decision maker to focus more
closely on probable results of a decision. Consequently, probability estimates for the
various states of nature can serve an important function if they can be obtained. Of
course, in some situations, accurate estimates of probabilities may not be available,
in such cases the decision maker may have to select a course of action without the
benefit of probabilities.
e) Decision Criterion
The process of selecting one alternative from a list of alternatives is governed by a
decision criterion, which embodies the decision maker’s attitudes toward the
decision as well as the degree of certainty that surrounds a decision. For instance,
some decision makers are more optimistic, whereas others are more pessimistic.
Moreover, some want to maximize gains, whereas others are more concerned with
protecting against losses.
4.2. Steps in Decision Theory Approach
The decision theory approach generally involves five steps. These are:
i) List all the viable alternatives /strategies/ course actions
The decision maker, first, must list all the viable alternatives that can be considered
in the decision.
ii) Identify the states of nature /expected future events.
The decision maker must list all the future events that may occur. Often, it is
possible to identify most of the events that can occur: the difficulty is to identify
which particular event will occur.

2
 iii) Construct a payoff table/decision tree
iv) Select optimum decision criterion
v) Choose an alternative (decision)

4.3. Decision Making Environments

Decisions are made under three types of environments:
1)Decision making under conditions of certainty
In this environment, only one state of nature exists for each alternative, i.e., there is
complete certainty about the future. It is easy to analyze the situation and make
good decision. Since the decision maker has perfect knowledge about the future
outcomes, s/he simply chooses the alternative having optimum payoffs.
2)Decision making under conditions of uncertainty
Here, more than one states of nature exist but the decision maker lacks sufficient
knowledge to allow him/her assign probabilities to the various states of nature.
3)Decision making under conditions of risk
Here also, more than one states of nature exist but the decision maker has
sufficient knowledge to allow him/her to assign probabilities to the various states of
nature.
4.4. The Payoff Table
A payoff table is a device a decision maker can use to summarize and organize
information relevant to a particular decision. It includes a list of alternatives,
possible future states of nature, and payoffs associated with each of the alternative-
states of nature combination. If probabilities for the states of nature are available,
these can also be listed.
General format of a payoff table:
State of nature
S1 S2 S3 Where:
ai = the ith alternative
a1 V11 V12 V13
Sj = the jth state of nature
Alternativ
a2 V21 V22 V23 Vij = the value of payoff that will be
es
realized if alternative ‘i’ is chosen and
a3 V31 V32 V33
event ‘j’ occurs

For Tabor Ceramic Manufacturing Company’s expansion decision stated above, the
payoff table is constructed as shown below:

High Moderate Low Nil

Expand

Br. 50,000 Br. 25,000 Br. -25,000 -45,000

Construct

70,000 30,000 -40,000 -80,000

Subcontract

30,000 15,000 -1,000 -10,000

3
 4.5. Decision Making Under Complete Certainty
Since under this environment only one state of nature exists, the decision maker
simply picks up the best payoff in that state of nature (column) and chooses the
associated alternative. For example, if Tabor Ceramic Manufacturing Company knew
that ceramic product demand would be high, it would choose the alternative
‘construct a new plant’ to get the highest payoff Birr 70,000; if the company knew
that the demand would be low, it would choose alternative ‘subcontract’ to keep the
losses lowest to Birr 1,000. Though, the state of nature is only one (having
probability of 1.00), possible alternatives could be numerous. Few complex
managerial decision making problems, even enjoy the luxury of having complete
information about the future and thus decision making under certainty is of little
consequential interest.
In summary, when a decision is made under conditions of complete certainty, the
attention of the decision maker is focused on the column in the payoff table that
corresponds to the state of nature that will occur. The decision maker then selects
the alternative that will yield the best payoff given that state of nature.
4.6. Decision Making Under Complete Uncertainty
Under complete uncertainty, the decision maker has knowledge about the states of
nature that happen but lacks the knowledge about the probabilities of their
occurrence. The decision maker either is unable to estimate the probabilities for the
occurrence of the different states of nature, or else s/he lacks confidence in
available estimates of probabilities, and for that reason, probabilities are not
included in the analysis.
Under complete uncertainty, a few criteria are available which could be of help to
the decision maker, and a choice among them is determined by the organization’s
policy and attitude of the decision maker. We shall consider five decision criteria to
decision making under complete uncertainty.
They include:
 Maximin (Pessimism)
 Maximax (Optimism)
 Minimax Regret
 Insufficient Reason , and
 Hurwicz Criterion (Realism)
a) Maximin Criterion
The maximin strategy is a conservative one; it consists of identifying the worst
payoff for each alternative, and, then selecting the alternative that has the best of
the worst payoffs. The maximin payoff to Tabor Ceramic Manufacturing Company is
a loss of Birr 10,000 corresponding to the alternative ‘subcontract’.

4
 High Moderate Low Nil Row minimum

Expand Br.50,000 25,000 -25,000 -45,000 -45,000

Construct 70,000 30,000 -40,000 -80,000 -80,000
Subcontrac 30,000 15,000 -1,000 -10,000 -10,000 maximin payoff
Many
t people view the maximin criterion as pessimistic because they believe that
the decision maker must assume that the worst will occur.
b) Maximax Criterion
This criterion provides the decision maker with optimistic criterion that is, choosing
the alternative that could result in the best payoff. The decision maker finds the
best possible payoff for each possible alternative and then selects the alternative
with best of the best payoff within this group. The maximax payoff for Tabor
Ceramic Manufacturing Company is a profit of Birr 70,000 corresponding to the
alternative ‘construct’.
High Moderate Low Nil Row maximum

Br.50,000 25,000 -25,000 -45,000 50,000

Expand
70,000 30,000 -40,000 -80,000 70,000 maximum payoff
Construct
30,000 15,000 -1,000 -10,000 30,000
Subcontrac
t

c) Minimax Regret Criterion

Both the maximax and maximin strategies can be criticized because they focus only
on a single, extreme payoff and exclude the other payoffs. An approach that does
take all payoffs into account is minimax regret. The decision maker might
experience regret after the decision has been made and the states of nature
(events) have occurred. Thus, the decision maker should attempt to minimize regret
before actually selecting a particular alternative. In order to use this approach, it is
necessary to develop an opportunity loss (regret) table. The opportunity loss
reflects the difference between each payoff and the best possible payoff in a given
state of nature (column). Hence, opportunity loss amounts are found by identifying
the best payoff in a column and, then subtracting each of the other values in the
column from that payoff.
The basic steps involved in minimax regret are:
i) Determine the amount of regret (opportunity loss) to each event for every
alternative
ith regret = (best payoff – ith payoff) for the jth event
ii) Determine the maximum regret amount for each alternative
iii) Select the alternative that corresponds to the minimum of the maximum
regrets.

5
 Opportunity loss table
High Moderate Low Nil Maximum Loss/regret

35,000 minimum regret

Expand Br 20,000 5,000 24,000 35,000

Construct 0 0 39,000 70,000 70,000

Subcontra
ct 40,000 15,000 0 0 40,000

This table shows that Tabor Ceramic Manufacturing Company will minimize its
regret to Birr 35,000 by selecting alternative “Expand”.
In minimax regret criterion, a decision maker could select an alternative in such a
way as to minimize the maximum possible regret. This requires identifying the
maximum opportunity loss in each row and, then, selecting the alternative that
would yield the best (minimum) of these regrets.
d) Principle of Insufficient Reason
Since the probabilities associated with occurrence of various events are unknown,
there is not enough information to conclude that these probabilities will be different.
Therefore, this criterion assigns equal probabilities to all events of each alternative
and selects the alternative associated with the best average payoff. That is, it
treats the states of nature as if each were equally likely, and it focuses on the
average payoff for each row, selecting the alternative that has the best row
average. The basis for the criterion of insufficient reason is that under complete
uncertainty, the decision maker should not focus on either high or low payoffs, but
should treat all payoff /actually, all states of nature), as if they were equally likely.
High Moderate Low Nil Row Average

50,000 25,000 -25,000 -45,000 1,250

Expand

70,000 30,000 -40,000 -80,000 -5,000

Construct

30,000 15,000 -1,000 -10,000 8,500 ←best

Subcontrac payoff

Thus, alternative ‘subcontract’ results in best average payoff of Birr 8,500

e) Hurwitz Criterion (Criterion of Realism)
It is a compromise between the maximax (optimistic) and maximin (pessimistic)
decision criteria. This criterion is based Hurwitz’s concept of coefficient of optimism

6
(or pessimism). This concept allows the decision maker to take into account both
the best and worst payoffs for each alternative and assign weights according to
his/her degree of optimism (or pessimism). The alternative, which has the best sum
of these weighted payoffs is then selected.
The criterion of realism consists of the following steps:
 Choose an appropriate degree of optimism , so that (1-) represents the
degree of pessimism.  is called coefficient or index of optimism.
 Determine the best as well as the worst of each alternative and obtain:
P=× best payoff + ( 1− ) × worst payoff
Where , P=row weighted payoff
 Choose the alternative that yields the best payoff of P.
Assume that α =0.8 , then
1−α=0.2
P=0.8× maximum value+ 0.2× minimum value

High Moderate Low Nil Row weighted payoff

Expand 50,000 25,000 -25,000 -45,000 31,000

Construct
70,000 30,000 -40,000 -80,000 40,000 best payoff

30,000 15,000 -1,000 -10,000 22,000

Subcontract

4.7. Decision Making Under Risk

The essential difference between decision making under complete uncertainty and
decision making under risk is the presence of probabilities for the occurrence of the
various states of nature under risk.
The probabilities may be subjective estimates from managers or from experts in a
particular field, or they may reflect historical frequencies. If they are reasonably
correct, they provide a decision maker with additional information that can
dramatically improve the decision making process. The sum of the probabilities for
all states of nature must be 1.00. For example, Tabor Ceramic Manufacturing
Company might estimate that probability of high demand at 0.25, the probability of
moderate demand at 0.40, the probability of low demand at 0.20 and the
probability of no demand at 0.15. (Note that 0.25 + 0.4 + 0.2 + 0.15 = 1.00).
4.7.1.Expected Monetary Value
The expected monetary value (EMV) approach provides the decision maker with a
value, which represents an average payoff for each alternative. The best alternative
is, then, the one that has the highest expected monetary value.
The average or expected payoff of each alternative is a weighted average: the state
of nature probabilities are used to weight the respective payoffs.
Thus, the expected monetary value is:
EMV i=¿

7
Where:
EMVi= the expected monetary value for the ith alternative
Pj = the probability of the jth state of nature
Vij = the estimated payoff for alternative ‘i’ under state of nature ‘j’.
For example, the Expected Monetary Value of the ‘expand’ alternative is:
EMVe = 0.25 x 50,000 + 0.4 x 25,000 + 0.2 x -25,000 + 0.15 x -45,000
= 12,500 + 10,000 – 5,000 – 6,750
= 10,750
The Expected Monetary Value of the ‘construct’ alternative is:
EMVc = 0.25 x 70,000 + 0.4 x 30,000 + 0.2 x -40,000 + 0.15 x -80,000
= 17,500 + 12,000 – 8,000 – 12,000
= 8,500
The Expected Monetary Value of the ‘subcontract’ alternative is:
EMVs = 0.25 x 30,000 + 0.4 x 15,000 + 0.2 x -1,000 + 0.15 x -10,000
= 7500 + 600 – 200 – 1500
= 11,800

(0.25) (0.4) (0.2) (0.15) Expected

High Moderate Low Nil payoff
50,000 25,000 -25,000 -45,000 10,750
Expand
70,000 30,000 -40,000 -80,000 8,500
Construct
30,000 15,000 -1,000 -10,000 11,800 ←best payoff
Subcontrac
t

Because the ‘subcontract’ alternative has the largest expected monetary value, it
would be selected using EMV criterion. The expected monetary value approach is
more suited to an ongoing decision strategy.
4.7.2.Expected Opportunity Loss
Expected opportunity loss criterion is nearly identical to the EMV approach, except
that a table of opportunity losses is used rather than a table of payoffs. Hence, the
opportunity losses for each alternative are weighted by the probabilities of their
respective states of nature to compute a long run average opportunity loss, and the
alternative with the smallest expected loss is selected as the best choice.
High Moderate Low Nil Expected Losses

20,000 5,000 24,000 35,000 17,050

Expand

8
 0 0 39,000 70,000 18,300

Construct

40,000 15,000 0 0 16,000 ←worst payoff

Subcontrac

The alternative ‘subcontract’ would be selected, because it has the smallest

expected opportunity loss (EOL).
Note that the EOL approach resulted in the same alternative as the EMV approach.
The two methods will always result in the same choice because they are equivalent
ways of combining the values; maximizing payoffs is equivalent to minimizing the
opportunity losses.
4.7.3. Expected Value of Perfect Information
The expected value of perfect information (EVPI) is a measure of the difference
between the certain payoff that could be realized under condition of certainty and
the expected payoff under a condition involving risk. Consider the payoff that Tabor
Ceramic Manufacturing Company could expect under certainty. If the company
knew that the demand is high, ‘construct’ would be chosen with a payoff of Birr
70,000; if the company knew that the demand is moderate, ‘construct’ would be
chosen with payoff of Birr 30,000, if the company knew that the demand is low,
‘subcontract’ would be chosen with a payoff of Birr -1,000, and if the company
knew that there is no demand, ‘subcontract’ would be chosen with a payoff of Birr -
10,000. Hence, if it were possible to remove the uncertainty surrounding the state
of nature, the decision maker could capitalize on that knowledge. Obviously, before
investing time or money in eliminating the probabilities, it will be impossible for the
decision maker to say which state of nature will turnout to be the one that will
occur. However, what can be said is that the probabilities that perfect information
will indicate for each state of nature is their original probabilities. Thus, these
probabilities can be used to weight the payoffs, one of which will occur under
certainty. This is called the expected payoff under certainty (EPC), and computed in
the following way;
EPC = 0.25 x 70,000 + 0.4 x 30,000 – 0.2 x 1,000 – 0.15 x 10,000
 17,500 + 12,000 – 200 – 1,500 = Birr 27,800
EVPI = EPC – EMV
EVPI = 27,800 – 11,800 = Birr 16,000
Note that the EVPI is exactly equal to EOL. In fact, these two quantities will always
be equal. The EOL indicates the expected opportunity losses due to imperfect
information, which is another way of saying the expected payoff that could be
achieved by having perfect information.

9
The EVP1 represents an upper bounded on the amount of money the company
would be justified in spending to obtain perfect information. Thus, the company
would be justified in spending up to Birr 16,000 to find out for certain which state of
nature will prevail.

10
 4.8. Decision Trees
Decision trees sometimes are used by decision makers to obtain a visual portray of
decision alternatives and their possible outcomes. The term gets its name from the
tree-like appearance of the diagram. A tree composed of squares, circles and lines.
The lines or branches that emanate from a square represent alternatives, while the
branches which emanate from a circle represent states of nature. The expected
payoffs and the probabilities associated with alternative course of action and the
chance events are shown along side these branches. At the terminal of the chance
branches are shown the expected value of outcome. The general approach used in
decision tree analysis is to work backward through the tree from right to left,
computing the expected value of each chance
Although decision trees represent an alternative approach to payoff tables, they are
not commonly used for problems that involve a single decision. Rather, their
greatest benefit lies in portraying sequential decisions (i.e., a series of chronological
decisions). In the case of a single decision, constructing a tree can be cumbersome
and time consuming. Conversely, situations that involve sequential decisions are
difficult to represent in payoff tables.

Example
Ato Sheferaw had to decided whether or not drill a well on his farm. In his village,
only 40% of the wells drilled were successful at 200 feet depth. Some of the farmers
who did not get water at 200 feet, drilled up to 250 feet but only 20% struck water
at 250 feet. Cost of drilling is Birr 50 per foot. Ato Sheferaw estimated that he would
pay Birr 18,000 during a 5 year period in the present value terms, if he continues to
buy water from the neighbour rather than go for the well which would have a life of
5 years. Ato Sheferaw has three decisions to make
a) Should he drill up to 200 feet?
b) In no water is found at 200 feet, should he drill up to 250 feet?
c) Should he continue to buy water from his neighour?
Solution
The cost associated with each outcome is written on the decision tree.
EMV of node B = Birr (0.2 x 12,500 + 0.8 x 30,400) = Br 26,900
EMV of node 2 = Birr 26,900 (lesser of the two values of Birr 26,900 and Birr
28,000)
EMV of node A = Birr (0.4 x 10,000 + 0.6 x 26,900) = Birr 20,140
EMV of node 1 = Birr 18,000 (Lesser of the two values of Birr 20,140 and Birr
18,000)
Thus the optimal (least cost) course of action for Ato Sheferaw is not to drill the will
and pay Birr 18,000 for water to his neightbour for five years.
4.9. Decision Making with Additional Information
Decision makers can sometimes improve decision making by bringing additional
information into the process. The additional information can come from a variety of
sources. For example, either a market survey might be used to acquire additional
information or a forecasting technique might be employed. In certain situations, it

11
 may be possible to delay a decision; the passage of time often allows a decision
maker to obtain a clearer picture of the future because it shortens the time horizon
the decision maker must deal with. Whatever the source of information, the benefit
is that estimates of probabilities of possible future events tend to become more
accurate.
In general, obtaining additional (sample) information includes an associated cost. As
a result, a key issue for a decision maker in such circumstances is whether the
value of additional information is worth the cost of obtaining that information.
Moreover, prior probabilities may then be revised in light of the addition information
by using Bayes theorem to yield posterior probabilities.
Bayes theorem

P ( AB )= P ( A ) × P
i P( A i)× P(B/ Ai )

1
( AB )+ P ( A )× P ( AB )+… P ( A ) × P ( AB )
1
2
2
n
n

Example
The manager of an advertisement firm is trying to decide which of two advertising
proposals to use for an upcoming promotion. The manager has developed the
following payoff table.
Market
0.70 0.3
Strong Weak * At this point, the manager simply could
make a decision using the expected value
Print 40 20 criterion with the information given
media
50 10
Video
media
However, suppose that the manager has the option of testing the market, and this
testing will provide additional information in the form of revised probabilities on
whether the market will be strong or weak. If the manager chooses to test the
market, it will cost Birr 1,000; the manager, therefore, must decide whether the
expected benefit from the test will offset the cost required to conduct the test. If the
manager conducts the test; this will surely alter the probabilities of a strong and
weak market that were originally estimated. In fact, an integral part of the analysis
in assessing the value of this sample information involves computing revised
probabilities.

Reliability of market test (conditional

probabilities)
Results of Actual state of nature
market Strong market Weak market
test
Shows strong market 0.80 0.10
Shows weak market 0.20 0.90

12
These probabilities are known as conditional probabilities because they express the
reliability of market test given the condition of actual market type.
In order to compute the posterior probabilities, we must combine these conditional
probabilities with the original (prior) probabilities (0.70 and 0.30)

i) For a “Strong Market” test result:

Actual Conditional Prior Joint Revised
Market probabilities Probabilities probabilities probabilities
Strong 0.80 x 0.70 = 0.56 0.56/0.59 =0.95
Weak 0.10 x 0.30 = 0.03 0.03/0.59 =
0.59 0.05 1.00

ii)For all a “Weak Market” test result:

Actual Conditional Prior Joint Revised
Market probabilities Probabilities probabilities probabilities
Strong 0.2 x 0.70 = 0.14 0.14/0.41 =0.34
Weak 0.9 x 0.30 = 0.27 0.27/0.41 = 0.66
0.41 1.00

EMV of node B = 0.95 x 40 + 0.05 x 20 = Birr 39,000

EMV of node C = 0.95 x 50 + 0.05 x 10 = Birr 48,000
EMV of node 2 = Birr 48,000 (the larger of Birr 39,000 and Birr 48,000)
EMV of node D = 0.34 x 40 + 0.66 x 20 = Birr 26,800
EMV of node E = 0.34 x 50 + 0.66 x 10 = Birr 23,600
EMV of node 3 = Birr 26,800 (the larger of Birr 26,800 & 23,600)
EMV of node F = 0.70 x 40 + 0.30 x 20 = Birr 34,000
EMV of node G = 0.70 x 50 + 0.30 x 10 = Birr 38,000
EMV of node 4 = Birr 38,000
EMV of node A = 0.59 x 48 + 0.41 x 26.8 = Birr 39,308

Using the market test has an expected value Birr 39,308, whereas not using the
market test has an expected value of Birr 38,000. Thus, using the market test has
an expected value that is Birr 308 (39,308 - 1,000 – 38,000) more than not using
the test.

[ ][
EVSI = expected value wit h – expected value wit h out
sample information sample inforamtion ]
Then, if the cost obtaining the additional information is less than EVSI, it would
seem reasonable to spend the money to obtain information. But if the cost equals or
exceeds the expected value of the information, it would seem reasonable to not
spend the money to obtain information.

13