DECISION THEORY

Decision Theory
Decision making is the process which results in selection from a set of alternative courses of
action.
Terminologies
1. Decision maker; this is the person charged with the responsibility of making a
decision.
2. Action Space; one of the first actions the decision maker must take of determining the
alternative course of action.
3. The Acts; strategies available for the decision maker
4. Events or state of nature; situations that affect the outcome of the decision but beyond
the control of the decision maker
5. Outcome; end results of the decision-making process e.g profit or loss
6. Pay off; measure of the net benefit to the decision maker
7. Opportunity loss; Decision maker might experience regret after the decision has been
taken. Thus, is known as the opportunity loss.

Types of decisions
There are many types of decision making

1. Decision making under uncertainty

These refer to situations where more than one outcome can result from any single decision

2. Decision making under certainty

Whenever there exists only one outcome for a decision we are dealing with this category e.g. linear
programming, transportation assignment and sequencing e.t.c.

3. Decision making using prior data

It occurs whenever it is possible to use past experience (prior data) to develop probabilities for the
occurrence of each data

4. Decision making without prior data

No past experience exists that can be used to derive outcome probabilities in this case the decision
maker uses his/her subjective estimates of probabilities for various outcomes

Decision making under uncertainty

Several methods are used to make decision in circumstances where only the pay offs are known and the
likelihood of each state of nature are known

a) Maximin Method
This criteria is based on the “conservative approach’ to assume that the worst possible is going to
happen. The decision maker considers each strategy and locates the minimum pay off for each and then
selects that alternative which maximizes the minimum payoff

Illustration
Rank the products A B and C applying the Maximin rule using the following payoff table showing
potential profits and loses which are expected to arise from launching these three product in three market
conditions
(see table 1 below)

Page 1 of 8
 Pay off table in £ 000’s
Boom condition Steady state Recession Mini profits row
minima
Product A +8 1 -10 -10
Product B -2 +6 +12 -2
Product C +16 0 -26 -26

Table 1
Ranking the MAXIMIN rule = BAC

b) MAXIMAX method
This method is based on ‘extreme optimism’ the decision maker selects that particular strategy which
corresponds to the maximum of the maximum pay off for each strategy

Illustration
Using the above example
Max. profits rw maxima
Product A +8
Product B +12
Product C +16

Ranking using the MAXIMAX method = CBA

c) MINIMAX regret method

This method assumes that the decision maker will experience ‘regret’ after he has made the decision
and the events have occurred. The decision maker selects the alternative which minimizes the
maximum possible regret.

Illustration
Regret table in £ 000’s
Boom condition Steady state Recession Mini regret row
maxima
Product A 8 5 22 22
Product B 18 0 0 18
Product C 0 6 38 38

A regret table (table 2) is constructed based on the pay off table. The regret is the ‘opportunity loss’
from taking one decision given that a certain contingency occurs in our example whether there is
boom steady state or recession
The ranking using MINIMAX regret method = BAC

d) The expected monetary value method

The expected pay off (profit) associated with a given combination of act and event is obtained by
multiplying the pay off for that act event combination by the probability of occurrence of the given
event. The expected monetary value (EMV) of an act is the sum of all expected conditional profits
associated with that act

Example
A manager has a choice between
i. A risky contract promising shs 7 million with probability 0.6 and shs 4 million with
probability 0.4 and
ii. A diversified portfolio consisting of two contracts with independent out comes each
promising Shs 3.5 million with probability 0.6 and shs 2 million with probability 0.4
Can you arrive at the decision using EMV method?

Page 2 of 8
 Solution
The conditional payoff table for the problem may be constructed as below.
(Shillings in millions)
Event Ei Probability Conditional pay offs decision Expected pay off decision
(Ei)
(i) Contract (ii) Portfolio(iii) Contract (i) Portfolio (i) x
x (ii) (iii)
Ei 0.6 7 3.5 4.2 2.1
E2 0.4 4 2 1.6 0.8
- -
EMV 5.8 2.9

Using the EMV method the manager must go in for the risky contract which will yield him a higher
expected monetary value of shs 5.8 million

e) Expected opportunity loss (EOL) method

This method is aimed at minimizing the expected opportunity loss (OEL). The decision maker
chooses the strategy with the minimum expected opportunity loss

f) The Hurwiz method

This method was the concept of coefficient of optimism (or pessimism) introduced by L. Hurwicz.
The decision maker takes into account both the maximum and minimum pay off for each alternative
and assigns them weights according to his degree of optimism (or pessimism). The alternative which
maximizes the sum of these weighted payoffs is then selected

g) The Laplace method

This method uses all the information by assigning equal probabilities to the possible payoffs for each
action and then selecting that alternative which corresponds to the maximum expected pay off

Example
A company is considering investing in one of three investment opportunities A, B and C under certain
economic conditions. The payoff matrix for this situation is economic condition

Investment 1£ 2£ 3£
opportunities
A 5000 7000 3000
B -2000 10000 6000
C 4000 4000 4000

Determine the best investment opportunity using the following criteria

i. Maximin
ii. Maximax
iii. Minimax
iv. Hurwicz (Alpha = 0.3

Solution
Economic condition
Investment 1£ 2£ 3£ Minimum Maximum £
opportunities £
A 5000 7000 3000 3000 7000
B -2000 10000 6000 -2000 10000
C 4000 4000 4000 4000 4000
i. Using the Maximin rule Highest minimum = £ 4000

Page 3 of 8
 Choose investment C
ii. Using the Maximax rule Highest maximum = £ 10000
Choose investment B
iii. Minimax Regret rule

1 2 3 Maximum
regret
A 0 3000 3000 3000
B 7000 0 0 7000
C 1000 6000 2000 6000

Choose the minimum of the maximum regret i.e. £3000

Choose investment A
iv. Hurwicz rule: expected values
For A (7000 x 0.3) + (3000 x 0.7) = 2100 + 2100 = £4200
For B (10000 x 0.3) + (-2000 x 0.7) = 3000 + 1400 = £ 1600
For C (4000 x 0.3) + (4000 x 0.7) = 1200 + 2800 = £ 4000
Best outcome is £ 4200 choose investment A

Value of perfect information

It relates to the amount that we would pay for an item of information that would enable us to forecast the
exact conditions of the market and act accordingly.
The expected value of perfect information EVPI is the expected outcome with perfect information minus
the expected outcome without perfect information namely the maximum EMV

Example
From table 1 above and given that the probabilities are Boom 0.6, steady state 0.3 and recession 0.1 then
When conditions of the market are; boom launch product C: profit = 16
When conditions of the market are; steady state launch product B: profit = 6
When conditions of the market are; recession launch product B: profit = 12
The expected profit with perfect information will be
(16 x 0.6) + (6 x 0.3) + (12 x 0.1) = 12.6
our expected profit choosing product C is 7
the maximum price that we would pay for perfect information is 12.6 – 7 = 5.6

7.2 DECISION TREES AND SUB SEQUENTIAL DECISIONS

A decision tree is a graphic display of various decision alternatives and the sequence of events as if they
were branches of a tree.

- The symbol and indicates the decision point and the situation of uncertainty or event
respectively. The node depicted by a square is a decision node while outcome nodes are depicted
by a circle.
- Decision nodes: points where choices exist between alternatives and managerial decisions is made based
on estimates and calculations of the returns expected.
- Outome nodes are points where the events depend on probabilities

Page 4 of 8
Illustration of a tree diagram
Event

111
event B1
ACT E1 D2 B2 112
A1 E2 C1
D1 D3 122

A2 C2 121
131

For example 111 represents the payoff of the act event combination A1 – E1 – B1
When probabilities of various events are known they are written along the corresponding branches. Joint
probabilities are obtained by multiplying the probabilities along the branches

Example
Kauzi Agro mills ltd (KAM) is considering whether to enter a very competitive market. In case KAM
decided to enter this market it must either install a new forging process or pay overtime wages to the entire
workers. In either case, the market entry could result in
i. high sales
ii. medium sales
iii. low sales
iv. no sales
a) Construct an appropriate tree diagram
b) Suppose the management of KAM has estimated that if they enter the market there is a
60% chance of their stakeholders approving the installation of the new forge. (this means
that there is a 40% chance of using overtime) a random sample of the current market
structure reveals that KAM has a 40% chance of achieving high sales, a 30% chance of
achieving medium sales, a 20% chance of achieving low sales and a 10% chance of
achieving no sales. Construct the appropriate probability tree and determine the joint
probabilities for various branches
c) Market analysts of KAM have indicated that a high level of sales will yield shs 1,000,000
profit; a medium level of sales will result in a shs 600000 profit a low level of sales will
result in a shs 200000 profit and a no sales level will case KAM a loss of shs 500000 apart
from the cost of any equipment. Entering the market will require a cash outlay of either
shs 300000 to purchase and install a forge or shs 10000 for overtime expenses should the
second option be selected.
Draw the appropriate decision tree diagram

Solution
a) The tree diagram for this problem is illustrated as follows
The 1st stage of drawing a tree diagram is to show all decision points and outcome done from left to right,
concentrate first on the logic of the problem and on probabilities or values involved. This is called forward
pass.

Page 5 of 8
The resultant is the figure below:
Outcome/event
Act Act/event
High sales
Install forge 5

6 Medium sales
3
7 Low sales

1 8 No sales

0
Use overtime High sales
9

4 Medium sales
10

stop 11 Low sales

2

12 No sales

Do not enter market

tree diagram
The entire sample space of act event choices is available to KAM are summarized in the table shown below
Path Summary of alternative Act event sequence
0–1–3–5 Enter market, install forge, high sales
0–1–3–6 Enter market, install forge, medium sales
0–1–3–7 Enter market, install forge, low sales
0–1–3–8 Enter market, install forge, no sales
0–1–4–9 Enter market, use overtime, high sales
0 – 1 – 4 –10 Enter market, use overtime, medium sales
0 – 1 – 4 – 11 Enter market, use overtime, low sales
0 – 1 – 4 – 12 Enter market, use overtime, no sales
0–2 Do not enter the market

b) The appropriate probability tree is shown in the figure below. The alternatives available to the
management of KAM ad events are identified. The joint probabilities are the result of the path
sequence that is followed. For example, the sequence ‘enter market install forge, low sales’ yields
(0.6) (0.2) = 0.12 = probability to install forge and get low sales.

Page 6 of 8
 Pay offs

HS = 0.24 = 1,000,000
0.4
Install forge
(300,000) 0.3 MS = 0.18 = 600,000
3
0.2
Enter Market 0.6
LS = 0.12 = 200,000
0.1
1
NS = 0.06 = - 500,000

0 0.4
Use overtime
0.4
(10,000) HS = 0.16 = 1,000,000
4 0.3
MS = 0.12 = 600,000
0.2
Don’t enter market
0.1 LS = 0.08 = 200,000
2
NS = 0.04 = - 500,000
(c) The overall decision is determined after analysis of the expected values at various points so the
correct decision (with the highest expected value is made. The stage is worked from right to left
and is known as the backward pass.
- The expected value for a decision is the highest pay off value where as the E.V for an
outcome is the summation of probability x pay off value of each branch. In both cases
any expenditure incurred due to the selection of the said option is deducted.
- In our case
Node 3 = (0.4  1,000,000) + (0.3  600,000) + (0.2  200,000) + (0.1  −50,000)
- 300,000
E.V. = 615,000 – 300,000 = 315,000

Node 4 = (0.4  1,000,000) + (0.3  600,000) + (0.2  200,000) + (0.1  −50,000)

- 10,000
E.V. = 615,000 – 10,000 = 605,000

Node 1 = (0.6 × 315,000) + (0.4 × 605,000)

E.V. = 431,000

Node 0 = The highest of (0;431,000)

Since not entering the market has a 0 expected value
= 431,000 = thus the decision should be to enter the market.

Page 7 of 8
This is represented as below in a tree diagram.

1,000,000
0.4
Install forge
0.3 600,000
3
0.2
Enter Market EV = 315,000 200,000
0.1
1
- 500,000
EV = 431,000
0
Use overtime
0.4
1,000,000
4 0.3
Don’t enter market 600,000
0.2
EV = 605,000

0.1 200,000

- 500,000

Advantages of decision trees

1. it clearly brings out implicit assumptions and calculations for all to see question and revise
2. it is easy to understand

Disadvantages
1. it assumes that the utility of money is linear with money
2. it is complicated by introduction of more variables and decision alternatives
3. it is complicated by presence of interdependent alternatives and dependent variables

Page 8 of 8