Decision Theory

Cruz, Eduard
Salamat, Ramil Paolo
DECISION THEORY
• Represents a general approach to decision making
which is suitable for a wide range of operations
management decisions.

• The reason for studying decision theory is that most

management decisions are made in an environment
of uncertainty and thus, it provides an orderly way of
choosing among several alternatives or alternative
strategies when decisions are to be made under
either uncertainty or risk.
 Components of Decision Theory

• Decision Alternatives
• States of Nature- actual events that may occur in the
future
• Payoff table
 EXAMPLE
Suppose a distribution company is considering purchasing a
computer to increase the number of orders it can process
and thus increase its business. If economic conditions
remain good, the company will realize a large increase in its
profit, however, if the economy takes a downturn, the
company will lose money.
Decision Alternatives:
-To purchase the computer
- Not to purchase the computer
States of Nature:
- Good economic conditions
- Bad economic conditions
 EXAMPLE
Payoff Table:
State of Nature

Decision Alternatives A B

1 Payoff 1A Payoff 1B

2 Payoff 2A Payoff 2B
 TYPES OF DECISION-MAKING
ENVIRONMENTS
I.Decision Making under Certainty

- Environment in which relevant parameters have

known values.
- There is only one type of event that can take place.
- For these kinds of problems, linear programming shall
be used.
 TYPES OF DECISION-MAKING
ENVIRONMENTS
II. Decision Making under Uncertainty (without Probabilities)
- Environment in which it is impossible to assess the
likelihood of various future events.

- 5 Decision-Making Criteria:
1.Maximax Criterion- decision maker selects the decision that
will result in the maximum of the maximum payoffs; said to be
“very optimistic”
2.Maximin Criterion- decision maker selects the decision that
will reflect the maximum of the minimum payoffs; said to be
“pessimistic”
 TYPES OF DECISION-MAKING
ENVIRONMENTS
II. Decision Making under Uncertainty (without Probabilities)

5 Decision-Making Criteria:
3.Minimax Regret Criterion- decision maker attempts to avoid
regret by selecting the decision alternative that minimizes the
maximum regret.
4.Hurwicz Criterion- is a compromise between the maximax and
maximin criteria; decision maker is neither totally optimistic (as
maximax criterion assumes) nor totally pessimistic (as maximin
criterion assumes).
5.Equal Likelihood/ LaPlace Criterion- weights each state of
nature equally, thus assuming that the states of nature are
equally likely to occur.
 TYPES OF DECISION-MAKING
ENVIRONMENTS
III. Decision Making under Risk (with Probabilities)
- Environment in which certain future events have probable
outcomes.

3 Decision-Making Criteria:
1.Expected Value- the decision maker must first estimate the
probability of occurrence of each state of nature, and then the
expected value for each decision alternative can be
computed.
2.Expected Opportunity Loss- expected value of the regret for
each decision.
3.Expected Value of Perfect Information (EVPI)- the
maximum amount a decision maker would pay for additional
information.
 THE SIX STEPS IN DECISION THEORY
1. Clearly define the problem.
2. List the possible decision alternatives.
3. Identify the possible outcomes/payoffs.
4. List the payoff or profit of each combination of
alternatives and outcomes.
5. Select one of the Mathematical Decision Theory
Models.
6. Apply the model and make your decision.
 ACTUAL EXAMPLE:

An investor is to purchase one among three types of

real estate: an apartment building, an office building
and a warehouse. The future states of nature that will
determine how much profit the investor will make are
good economic conditions and poor economic
conditions. The profits that will result from each
decision in the event of each state of nature are shown
in the table below:
 ACTUAL EXAMPLE:

State of Nature
Decision
(purchase) Good Economic Poor economic
Conditions conditions
Apartment
Building $ 50,000 $ 30,000
Office Building 100,000 -40,000
Warehouse 30,000 10,000
 Decision Making under Uncertainty
(without Probability)
1.Maximax Criterion (Max profit) – choose the
alternative with the best possible payoff or the decision
that will result in the maximum of the maximum payoffs.
(risk seeking)

Steps:
1)Identify the maximum payoff for each alternative.
2)Pick the largest maximum payoff from the choices.

Conclusion: Among the three choices, purchasing an

office building will yield the largest payoff of 100,000.
 Decision Making under Uncertainty
(without Probability)
2. Minimin Criterion (Min cost) – choose the
alternative with the smallest payoff or the decision that
will result in the minimum of the minimum profit.

Steps:
1) Identify the minimum payoff for each alternative.
2) Pick the smallest minimum payoff.
 Decision Making under Uncertainty
(without Probability)
3.Maximin Criterion – choose the alternative with the best of
the worst possible payoffs or the decision that will reflect the
maximum of the minimum payoffs (risk averse).
Steps:
1)Identify the minimum payoff for each alternative.
2)Pick the largest minimum payoff

Conclusion: The minimum payoffs for each alternative are

30,000, -40,000 and 10,000. The maximum of these three
payoffs is 30,000, thus, the decision arrived by using the maximin
criterion would be to purchase the warehouse.
 Decision Making under Uncertainty
(without Probability)
4.Minimax Regret Criterion – choose the alternative that has
the least of the worst regrets. Criterion that minimizes the loss
incurred by not selecting the optimal alternative.
Steps:
1)Identify the maximum payoff per state of nature (column).
2)Subtract the maximum payoff per column from all the other
payoffs from each state of nature. The difference is called the
degree of regret.
3)Construct a regret table showing the degree of regrets.
4)Identify the maximum regret for each alternative and then
choose the alternative with the smallest maximum regret, the
chosen value is called the minimax regret value.
 Decision Making under Uncertainty
(without Probability)

Conclusion: According to the minimax regret criterion,

the decision should be to purchase the apartment
building rather than the office building or warehouse.
This decision is based on the philosophy that the
investor will experience the least amount of regret by
purchasing the apartment building.
 Decision Making under Uncertainty
(without Probability)
5. Hurwicz Criterion- the decision payoffs are
weighted by a coefficient of realism called alpha (α),
which is a decimal number between 0 and 1, also
called the coefficient of optimism, on the other
hand, 1- α is the coefficient of pessimism.
Steps:
1)For each decision alternative, the maximum
payoff shall be multiplied by α and the minimum
payoff be multiplied by 1- α. Then add the results to
come up with the criterion of realism/optimism.
2)From the calculated numbers in each decision
alternative, we select the highest one.
 Decision Making under Uncertainty
(without Probability)

Conclusion: The Hurwicz criterion specifies selection of

the decision alternative corresponding to the maximum
weighted value, in this particular example, it is 38,000.
Thus, the decision would be to purchase the apartment
building.
 Decision Making under Uncertainty
(without Probability)
6. Equal Likelihood or Laplace Criterion – choose the alternative
that has the least of the worst regrets
- makes the assumption that any of the States of Nature could
occur, but does not give preference to anyone.
Steps:
1) Average the payoffs for each decision alternative.
2) Select the alternative with the highest average payoff.

Conclusion: Using the equal likelihood criterion, the investor

should purchase the apartment building because it has the
highest average payoff.
Decision Making under Uncertainty
(without Probability)
 Decision Making under Risk (with
Probability)
1.Expected Value (EV)
Steps:
1)Estimate the probability of occurrence of each state of nature.

State of Nature

Decision Good Economic Poor Economic

(purchase) Conditions Conditions
. 60 .40

Apartment Building 50,000 30,000

Office Building 100,000 -40,000
Warehouse 30,000 10,000
 Decision Making under Risk (with
Probability)
1.Expected Value (EV)
Steps:
2)The expected value (EV) can be computed by multiplying
each decision payoff under each state of nature by
corresponding probability of occurrence and then summing
up these products.
3)Select the best decision with the greatest expected value.

Conclusion: Because the greatest expected value is 44,000,

the best decision is to purchase the office building.
 Decision Making under Risk (with
Probability)
2. Expected Opportunity Loss
Steps:
1)Use the regret table (from Minimax Regret criterion)
showing the degree of regrets with probabilities for each
state of nature.
2)Multiply the probabilities by the regret for each
decision outcomes (rather than the actual decision
outcomes as we did with the expected value), then add
up the products.
3)Select the best decision that would be from minimizing
the regret, for this method we select the smallest
expected opportunity loss.
 Decision Making under Risk (with
Probability)
2. Expected Opportunity Loss

Conclusion: Because 28,000 is the minimum expected

regret, the decision is to purchase the office building.

Note: The expected value and the expected

opportunity loss methods will always yield the same
result. Thus, it is repetitious to apply both methods to a
decision situation when one of the two will suffice.
 Decision Making under Risk (with
Probability)
3.Expected Value of Perfect Information (EVPI)- the
maximum amount that would be paid to gain information that
would result in a decision better than the one made without
perfect information.
Steps:
1)Hypothetical “perfect” decisions are weighted by getting
the maximum payoff for each state of nature.
2)Get the expected value of the decision (maximum payoff
chosen above) by multiplying it to its respective probability,
then add up the products.
3)Get the expected value decision without perfect
information. (This was computed earlier in the Expected Value
with the decision of purchasing the office building since it has
the greatest EV of 44, 000).
 Decision Making under Risk (with
Probability)
3.Expected Value of Perfect Information (EVPI
Steps:
4)Compute the EVPI by subtracting the expected value (EV)
without perfect information from the expected value given
perfect information.

Conclusion: 28,000 is the maximum amount the investor

would pay to purchase perfect information from some other
source, such as an economic forecaster.

Note: The EVPI result is equal to the expected opportunity

loss.
 DECISION TREE
- Another useful technique for analyzing a decision
situation; provides an illustration (or picture)
representing the sequence of events in the decision-
making process.

- graphical diagram consisting of nodes and branches:

*circles and square are referred to as nodes
> square is the decision node
> circles are probability or event nodes
* branches are the decision alternatives
 DECISION TREE
State of Nature

Decision Good Economic Poor Economic

(purchase) Conditions Conditions
. 60 .40

Apartment
50,000 30,000
Building
Office Building 100,000 -40,000

Warehouse 30,000 10,000

 DECISION TREE
Determining the Best Decision using a Decision Tree:
1) Compute the expected value at each probability
node.
2) The values, now shown as the expected payoffs at
nodes 2, 3, and 4, is the outcome of a possible decision
that can occur at node 1. Select the branch with the
highest expected payoff.

Conclusion: The best decision is to purchase the office

building, with an expected payoff of $44,000.
 DECISION TREE

Note: When only one decision is to be made, the

decision tree will always yield the same decision and
expected payoff as the expected value criterion. Thus,
in this type of decision situation, a decision tree may
not be very useful.
 SEQUENTIAL DECISION TREE

-If a decision situation requires a series of decisions,

then a payoff table cannot be created (since it is usually
limited to a single decision situation), and a decision
tree becomes the best method for decision analysis.
SEQUENTIAL DECISION TREE EXAMPLE
*Real estate investment to encompass a 10-year period, during
which several decisions must be made.
 SEQUENTIAL DECISION TREE
Determining the Best Decision using a Decision Tree:
Steps:
1)Compute the expected values at nodes 6 and 7.
2)At decision nodes 4 and 5, we must make a decision.
Choose the one with the greatest expected value after
subtracting the cost of each decision.
3)Compute the expected values at nodes 2 and 3.
4)Make the final decision at node 1. Select the decision
with the greatest expected value after the cost of each
decision is subtracted out.
Conclusion: The best decision is to purchase the land with
the highest net expected value of $1,160,000.
REFERENCES
• Management Science by Bernard Taylor
• http://www.yourarticlelibrary.com/decision-
making/decisions-making-environments-
certainty-uncertainty-and-risk/10269/