Decision Making

• Contents
• The decision process
• Causes of poor decisions
• Decision environments
• Decision theory
• Decision Trees
• Expected value of Perfect Information (EVPI)
• Sensitivity analysis
The decision process
• Successful decision making involves following steps:
• Identify the problem

• Specify the objectives and the criteria for choosing a solution

• Develop alternatives

• Analyze and compare alternatives

• Select the best alternative

• Implement the chosen alternative

• Monitor the results to ensure that desired results are achieved

The decision process (cont.)
• Identifying the problem
• Focal point of the process
• Don’t remove symptoms of the problem
• Solution must address the basic problem, not the symptoms

• Identify the criteria

• By which proposed solutions will be judged
• Costs, profits, return on investment, increased productivity, risk, company image, impact on
demand, etc.

• Developing suitable alternatives

• Danger of one or more potentially superior alternatives will be overlooked
• Depends on experience and creativity of the decision maker, and nature of situation
• One alternative is to do nothing
• No time or effort, no costs, no implementations
The decision process (cont.)
• Analyzing the alternatives and comparing alternatives
• Use mathematical and statistical techniques

• Selection of best alternative

• Depends on the objectives of decision maker and the criteria that are being used to evaluate alternatives

• Implementing a solution
• Carrying out the actions indicated by the chosen alternative.
• If the alternative selected is to do nothing, approach used by many decision makers, no action will be required to
implement, but by the time they get around to making a decision, it is too late.

• Monitoring
• Effective decision making requires monitoring to make sure that desired consequences have been
achieved. If not,
• Repeat the entire process
• Review the situation to identify error in implementation
• Identify error in calculations
• Wrong assumption of the situation
Causes of poor decisions
• Decisions occasionally turns out poorly due to unforeseeable circumstances
• In most of the cases failure occurs due to
• Manager’s failure
• Bonded rationality
• Sub-optimization

• Manager’s failure
• Skip a decision making step,
• Style of making quick decisions
• Failure to recognize consequences of a poor decision
• Ego – can do no wrong
• Unwilling to admit a mistake
• Inability to make decision
Causes of poor decisions (Cont.)
• Bounded rationality
• The limitations on decision making caused by cost, human abilities, time technology and
availability of information
• Mangers cannot reach optimal decision (highest profit, least cost) but try to achieve a
satisfactory decision.

• Sub-optimization
• Poor decisions caused by the departmentalized dictions.
• The result of different departments each attempting to reach a solution that is optimum
for that department
Decision Environments
• Three basic categories according to the certainty present.
• Certainty
• Risk
• Uncertainty

• Certainty
• Environment in which relevant parameters such as costs, capacity, and demand 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 (cont.)
• Consider the situations
1. Profit per unit is Rs. 100/-. Based on previous experience, there is 50% chance of an
order of 100 units and 50% chance of an order of 200 units. What is expected profit?
2. Profit per unit is Rs. 100/-. You have an order of 200 units. How much profit will you
make?
3. Profit per unit is Rs. 100/-. The probabilities of potential demands are unknown.

• All these decision environments require different analysis techniques

Decision Theory
• General approach to decision making about capacity planning, product and
service design, equipment selection and location planning, etc.

• Decision theory characterized by elements

• A set of possible future conditions which can effect results of the decision
• A list of alternatives for the manager to choose from
• A known payoff for each alternative under each possible alternative future condition
Decision Theory (Cont.)
• To use this approach, follow the underlines process
• Identify the future conditions (states of nature)
• demand will be LOW, MEDIUM, HIGH
• the number of contracts awarded would be ONE, TWO, THREE
• Develop a list of possible alternatives
• Determine or estimate the payoff associated with each alternative for every
possible future condition
• If possible, estimate the likelihood of each possible future condition
• Evaluate alternatives according to some decision criterion (e.g. maximize
expected profit) and select the best alternative.
Decision Theory (Cont.)
• Payoff table
• Table showing expected payoffs for each alternative in every possible state of nature.
• Helpful in choosing alternatives by facilitating caparison of alternatives.

• Example: capacity planning problem- equivalent dollar values (millions) of

expected future income less costs
Possible future demand
Alternatives Low Moderate High
Small facility $10* $10 $10
Medium facility 7 12 12
Large facility -4 2 16
Decision Theory (Cont.)
• Decision making under Certainty
• When it is known for certain which of the possible future conditions will
actually happen
• simply choose the alternative that has the best payoff under that stage of
nature.

Decision making under Certainty

Possible future demand
Alternatives Low Moderate High
Small facility $10* $10 $10
Medium facility 7 12 12
Large facility -4 2 16
Decision Theory (Cont.)
• Decision making under Uncertainty
• No information is available on how likely the various states of nature are.
• Four possible decision criteria:
• Maximim
• Maximax
• Laplace
• Minimax regret
• Maximin
• Choose the alternative with the best of the worst possible payoff –pessimistic approach – guaranteed minimum
• Maximax
• Choose the alternative with the best possible payoff –optimistic approach-go for it strategy
• Laplace
• Choose the alternative with best average payoff of any of the alternative – traets the nature as equalily likely
• Minimax regret
• Choose the alternative that has the least of the worst regrets

• Weakness of all except Lalace approach is that

• Focus on worst or best, do not take into account of all payoffs
• Weakness of Laplace approach
• For a give set of circumstances some have certain merits
Decision Theory (Cont.)
• Maximin
• Compare worst for each alternative
• Select the best of the worst i.e. “Small facility”

Maximin Approach
Possible future demand
Alternatives Low Moderate High Worst
Small facility $10* $10 $10 10
Medium facility 7 12 12 7
Large facility -4 2 16 -4
Decision Theory (Cont.)
• Maximax
• Best payoff of each alternative
• Select the best of the best i.e. “Large facility”

Maximax Approach
Possible future demand
Alternatives Low Moderate High Best
Small facility $10* $10 $10 10
Medium facility 7 12 12 12
Large facility -4 2 16 16
Decision Theory (Cont.)
• Laplace
• Take average of payoff of each alternative
• Select the best one i.e. “Medium facility”

Laplace Approach
Possible future demand
Alternatives Low Moderate High Total Average
(Total/Number of state of nature)
Small facility $10* $10 $10 30 10
Medium facility 7 12 12 31 10.33
Large facility -4 2 16 14 4.76
Decision Theory (Cont.)
• Minimax regret
• Prepare opportunity losses or regrets
• Subtract each payoff in each column from the largest positive payoff in that column.
• Select the least of worst i.e. “Medium facility”

Minimax Regret
Possible future demand Regrets
Alternatives Low Moderate High Low Moderate High Worst
Small facility $10* $10 $10 0 (10-10) 2 (12-10) 6 (16-10) 6
Medium facility 7 12 12 3 (10-7) 0 (12-12) 4 (16-12) 4
Large facility -4 2 16 14 (10+4) 10 (12-2) 0 (16-16) 14
Decision Theory (Cont.)
• Decision making under Risk
• The probability of occurrence of each state of nature can be estimated
(totaling 1.00)
• Widely used approach is expected monetary value (EMV) criterion.
• Expected monetary value is the sum of the payoffs for the alternative where
each payoff is weighted by the probability for the relevant state of nature.
• Alternative with the best expected value among the alternatives.
Decision Theory (Cont.)
• Multiply probability of state to payoff value of relevant state of an
alternative and sum up them
• Select the alternative with best EMV i.e. “Medium facility”

Decision making under Risk

Possible future demand
Alternatives Low Moderate High Expected monetary value
Probability 0.30 0.50 0.20
Small facility $10* $10 $10 10 (0.30x10+0.50x10+0.20x10)
Medium facility 7 12 12 10.5 (0.30x7+0.50x12+0.20x12)
Large facility -4 2 16 3 (0.30x(-4)+0.50x2+0.20x16)
Decision Trees
• Decision Tree
• A schematic representation of the available alternatives and their possible consequences.
• Can be used instead of payoff tables but more useful for analyzing situations that involve
sequential decisions.
• Decision tree composed of number of
• Nodes
• Square nodes – decision points
• Circular
• Branches emerging from nodes
• nodes – chance event
• Read the tree from left to right
• Branches leaving from square points (decision points) – alternatives
• Branches leaving from circular nodes (chance events) – possible states of nature
• Analyze from right to left
• For a decision-Choose the alternative that will yield greatest return (lowest cost)
• For chance vent follow a decision- choose the alternative that has the highest expected
monetary value (lowest expected cost)
Decision Trees (Cont.)
Decision Trees (Cont.)
• Example
• Analyze the decision tree
below and determine
which initial alternative
(build small or build large)
should be chosen in order
to maximize expected
monetary value.
Decision Trees (Cont.)
• Solution
• Read decision tree from right to left

• Step 1: First select best alternative for decision level 2

• For small facility “expand”
• For large facility – “reduce price”

• Sep 2: determine product of chance probabilities and their respective payoffs of the selected
branches i.e.
• Build small: Low demand : 0.4x40=16, High demand: 0.6x55=33
• Build Large: Low demand: 0.4x50=20, High demand: 0.6x70=42

• Determine the expected value of each initial alternative by summing up values

• Build small: 16+33=49
• Build large: 20+42=62

• Decision would be “build Large”, large expected monetary value.

Expected value of Perfect Information (EVPI)
• A decision maker may have the probabilities to delay a decision until it is clear
which state of nature will occur.
• If the state of nature favorable, the option can be exercised; if it is unfavorable, the
option can be allowed to expire.
• Whether the cost of option will be less than the expected gain due to delaying the
decision (expected payoff above the expected value)

• Expected gain is the expected value of perfect information (EVPI)

• The difference between the expected payoff under certain and the expected
payoff under risk.
Expected value of Perfect Information (EVPI) (cont.)

• Two ways to determine EVPI

• Compute the expected payoff under certain and subtract the expected payoff
under risk
𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑝𝑎𝑦𝑜𝑓𝑓 𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑝𝑎𝑦𝑜𝑓𝑓
• = −
𝑝𝑒𝑟𝑓𝑒𝑐𝑡 𝑖𝑛𝑓𝑜𝑟𝑚𝑎𝑡𝑖𝑜𝑛 𝑢𝑛𝑑𝑒𝑟 𝑐𝑒𝑟𝑡𝑎𝑖𝑛𝑖𝑡𝑦 𝑢𝑛𝑑𝑒𝑟 𝑟𝑖𝑠𝑘

• Use regret table to compute EVPI

Expected value of Perfect Information (EVPI) (cont.)
Possible future demand
Alternatives Low Moderate High Expected monetary value
Probability 0.30 0.50 0.20
Small facility $10* $10 $10 10 (0.30x10+0.50x10+0.20x10)
Medium facility 7 12 12 10.5 (0.30x7+0.50x12+0.20x12)
Large facility -4 2 16 3 (0.30x(-4)+0.50x2+0.20x16)

• Compute expected payoff under certainty

• Identify best payoff under each state of nature, multiply with respective probability of that state and then add
all. i.e.
• 0.30x10 (low) + 0.50x12 (moderate) + 0.20x16(high) = 12.2
• Expected payoff under risk i.e. 10.5
• EVPI = 12.2-10.5 = 1.7
• Upper limit on the amount the decision maker should be willing to obtain perfect information in
this case.
• If the cost equals of exceeds this amount, the decision maker would be better off not spending
additional money and go for alternative with highest payoff
Expected value of Perfect Information (EVPI) (cont.)
• Use of Regret table
• Find expected regret for each alternative
• The minimum expected regret is equal to EVPI
• Small facility: 0.30x0 + 0.50x2 + 0.20x6 = 2.2
• Medium facility: 0.30x3 + 0.50x0 + 0.20x4 = 1.7
• Large facility: 0.30x14 + 0.50x10 + 0.20x0 = 9.2
• EVPI=1.7 (Minimum), associated with second alternative.

Possible future demand Regrets

Alternatives Low Moderate High Low Moderate High Worst
Probability 0.30 0.50 0.20
Small facility $10* $10 $10 0 (10-10) 2 (12-10) 6 (16-10) 6
Medium facility 7 12 12 3 (10-7) 0 (12-12) 4 (16-12) 4
Large facility -4 2 16 14 (10+4) 10 (12-2) 0 (16-16) 14
Sensitivity Analysis
• Both payoff and probabilities are estimated values.
• Good to have indication how sensitive the choice of a alternative is to
change in one or more of these values
• Determining the range of probability for which an alternative has the
best expected payoff.
 Example
• Given the following table, determine the range of probability for state
nature #2, that is P(2), for which each alternative is optimal under the
expected –value approach.

State of nature
#1 #2
Alternatives A 4 12
B 16 2
C 12 8