Operations Research 2

Course Code: OR2

Instructor: Engr. Cesar Amante Ting

Included in this material are citations from the book Introduction to
Operations Research 7th edition authored by F. S. Hillier/G. J.
Lieberman and Operations Management by J. Heizer/B. Render.
INVENTORY MANAGEMENT &
MODELS
INVENTORY

- stock of good being held for future use or sale.

- it is one of the most expensive assets of many companies,
representing as much as 50% of total invested capital.

FUNCTIONS OF INVENTORY:

1. To decouple or separate various parts of the production

process. For example, if a firm supplies fluctuate, extra
inventory may be necessary to decouple the production process
from suppliers.
2. To decouple the firm from fluctuations in demand and provide
a stock of goods that will provide a selection for customers.
Such inventories are typical in retail establishments.
3. To take advantage of quantity discounts, because purchases in
larger quantities may reduce the cost of goods or their delivery.
4. To hedge against inflation and upward price change.
TYPES OF INVENTORY:

1. Raw Material Inventory – materials that are usually purchased

but have yet to enter the manufacturing process.
2. Work-in-process (WIP) Inventory – this is component/s or raw
material that have undergone some change but are not
completed. It exists because of the time it takes for a product
to be made (called cycle time).
3. Maintenance, Repair and Operating Materials (MROs) – these
are inventories devoted to maintenance/repair/operating
supplies necessary to keep machinery and processes
productive. They exist because the need and timing for
maintenance and repair of some equipment are unknown.
Although the demand for MRO inventories is often a function
of maintenance schedules, other unscheduled MRO demands
must be anticipated.
4. Finished-goods Inventory – it is a completed product ready to
be sold. It may be inventoried because future customer
demands are unknown.
INVENTORY MANAGEMENT:

ABC ANALYSIS

- It divides on-hand inventory into three classifications on the

basis of annual dollar volume.
- It is the application of Pareto Principle on inventories which
states that there are ‘critical few and trivial many’.
- The idea is to establish inventory policies that focus resources
on the few critical inventory parts and not the many trivial ones.
It is not realistic to monitor inexpensive items with the same
intensity as very expensive items.

HOW TO DETERMINE ANNUAL DOLLAR VOLUME FOR ABC

ANALYSIS:

1. Measure the annual demand of each inventory item times the

cost per unit.
Class A Items – these are items on which the annual
dollar volume is high. Although such items may represent
only about 15% of the total inventory items, they
represent 70-80% of the total dollar volume.
Class B Items – these are items of medium annual dollar
volume. These items may represent only about 30% of
the total inventory items and 15-25% of the total value.
Class C Items – these are items with low dollar volume,
which may represent only 5% of the annual dollar volume
but about 55% of the total inventory items.
Criteria other than annual dollar volume can determine dollar
classification. For instance, anticipated engineering changes,
delivery problems, quality problems, or high unit cost may
dictate upgrading items to a higher classification. The advantage
of dividing inventory items into classes allows policies and
controls to be established for each class.
Policies that may be based on ABC analysis include the following:

1. Purchasing resources expended on supplier development

should be much higher for individual A items than for C
items.
2. A items, as opposed to B and C items, should have tighter
physical inventory control; perhaps they belong in a more
secure area, and perhaps the accuracy of inventory records
for A items should be verified more frequently.
3. Forecasting A items may warrant more care than forecasting
other items.

RECORD ACCURACY:
• Accuracy of records is a critical ingredient in production and
inventory systems.
• Record accuracy allows organizations to focus on those items
that are needed, rather than settling for being sure that “some
of everything” is in inventory.
• Only when an organization can determine accurately what is
has on hand can it make precise decisions about ordering,
scheduling and shipping.
To ensure record accuracy, the following measures should be
done:

1. Incoming and outgoing record keeping must be good.

Receiving, shipping and internal transfer documentation /
record keeping should be religiously followed.
2. A well-organized stockroom will have limited access, good
housekeeping, and storage areas that hold fixed amounts of
inventory.
3. Bins, shelf space and other storage materials should be
labeled properly.

CYCLE COUNTING & PERIODIC PHYSICAL COUNT:

• Even though an organization may have made substantial
efforts to record inventory accurately, these records must be
verified through a continuing audit. Such audits are known
as cycle counting.
• Historically, many firms performed annual physical
inventories, some companies do it semi-annually.
• This practice often meant shutting down the facility and
having inexperienced people help in counting inventories with
CYCLE COUNTING & PERIODIC PHYSICAL COUNT:

• This practice often meant shutting down the facility and

having inexperienced people help in counting inventories with
assigned counters usually are accountants or inventory
analysts.
• Cycle counting usually divided into two phases, pre-count and
final physical count.
• Cycle counting uses inventory classifications developed
through ABC analysis.
• With cycle counting procedures, items are counted, records
are verified and inaccuracies are periodically documented. The
cause of inaccuracies is then traced and appropriate remedial
action taken to ensure integrity of the inventory system.

Cycle counting has the following advantages:

1. Eliminates the shutdown and interruption of production
necessary for annual physical inventories.
2. Eliminates annual inventory adjustments.
3. Trained personnel audit the accuracy of inventory.
Cycle counting has the following advantages:
1. Eliminates the shutdown and interruption of production
necessary for annual physical inventories.
2. Eliminates annual inventory adjustments.
3. Trained personnel audit the accuracy of inventory.
4. Allows the cause of the errors to be identified and remedial
action to be taken.
5. Maintains accurate inventory records.

APPLICABLE TECHNIQUES FOR INVENTORY ACCURACY &

CONTROL:

1. Good personnel selection, training and discipline. These are

never easy but very necessary in food-service, wholesale and
retail operations, where employees have access to direct
consumable merchandise.
2. Tight control of incoming shipments. This task is being
addressed by many firms through the use of barcode and
radio frequency ID systems that read every incoming
shipment and automatically check tallies against purchase
orders. When properly designed, these systems are hard to
defeat. Each item has its own stock keeping unit (SKU), an alpha-
numeric code for each type of item.
3. Effective control of all goods leaving the facility. This jobs is
accomplished with barcodes on items being shipped, magnetic
strips on merchandise, or via direct observation. Direct
observation can be personnel stationed at exits and in potentially
high-loss areas or can take the form of one-way mirrors and
video surveillance.

INDEPENDENT VERSUS DEPENDENT DEMANDS:

Independent Demand – it is where the usage of an item does not

affect the usage of other item. For example, the demand for
refrigerators is independent of the demand of oven toasters.

Dependent Demand – it is where the usage of an item does affect

the usage of other item/s. For example, the demand for oven
toaster components is dependent on the demand requirements
for oven toasters.
HOLDING, ORDERING and SET-UP COSTS:

Holding Costs – these are costs associated with holding or

“carrying” inventory over time. It includes obsolescence and costs
related to storage, such as insurance, extra staffing, and interest
payments. The following table shows the kinds of costs that need
to be evaluated to determine holding cost.
HOLDING, ORDERING and SET-UP COSTS:

Ordering Costs – it includes cost of supplies, forms, order

processing, clerical support and so forth. It also includes cost
when orders are being manufactured.

Set-up Cost – it is the cost to prepare a machine or process for

manufacturing an order. It includes time and labor to clean and
change tools or holders, which is highly correlated to set-up time.

INVENTORY MODELS FOR INDEPENDENT DEMAND:

I. ECONOMIC ORDER QUANTITY (EOQ) MODELS:

1. Basic EOQ Model
2. Production Order Quantity Model
3. Quantity Discount Model
BASIC ECONOMIC ORDER QUANTITY (EOQ) MODEL:

It is one of the oldest and most commonly known inventory

control techniques that yields the minimum total inventory cost.
This model answers the question ‘how much to order’. This
technique is relatively easy to use but is based on several
assumptions.

1. Demand is known, constant and independent.

2. Lead time or the time between placement and receipt of the
order is known and constant.
3. Receipt of inventory is instantaneous and complete. The
inventory from an order arrives in one batch at a time.
4. Quantity discounts are not possible.
5. The only variable costs are the cost of setting up or placing
an order (setup cost) and the cost of holding or storing
inventory over time (holding or carrying cost).
6. Stock-outs (or shortages) can be completely avoided if orders
are placed at the right time.
Inventory usage over time:

Total inventory cost as a function of order quantity:

With the EOQ model,

the optimal order
quantity will occur at
a point where the
total set-up cost is
equal to the total
holding cost.
Expected number of orders and expected time between orders:
Total annual inventory cost:
Total annual inventory cost:
PRODUCTION ORDER QUANTITY MODEL:

In Basic EOQ model, we assumed that the entire inventory order was
received at one time. There are times, however, when the firm may
receive its inventory over a period of time. Such cases require a different
model, one that does not require the instantaneous-receipt assumption.

This model is applicable under two situations:

1. When inventory continuously flows or builds up over a period of time
after an order has been placed; or
2. When units are produced and sold simultaneously.

Under these circumstances, we take into account daily production (or

inventory-flow) rate and daily demand rate. The following figure shows
inventory levels as a function of time:
Because this model is especially suitable for the production
environment, it is commonly called the production order quantity model.
It is useful when inventory continuously builds up over time, and
traditional economic order quantity assumptions are valid.

Using the following symbols, we can determine the expression for annual
inventory holding cost for the production order quantity model.
Example Problem:
QUANTITY DISCOUNT MODEL:

• Quantity discount is simply a reduced price (P) for an item when it is

purchased in larger quantities. Many companies offer quantity
discounts to their customers to increase sales.
• Discount schedules with several discounts for large orders are
common.

When we include the cost of the product, the equation for the total
annual inventory cost can be calculated as follows:
Example of Quantity Discount Schedule:

To get the EOQ for item with quantity discounts, the process involves four
steps:

Step 1: For each discount, calculate a value for optimal order size Q*,
using the following equation:
Note that the holding cost is IP instead of H. Because the price of the item
is a factor in annual holding cost, we cannot assume that the holding cost
is constant when the price per unit changes for each quantity discount.
Thus, it is common to express the holding cost (I) as a percent of unit
price (P) instead of as a constant cost per unit per year, H.

Step 2: For any discount, if the order quantity is too low to qualify for the
discount, adjust the order quantity upward to the lowest quantity that will
qualify for the discount. For example, if Q* for discount 2 in the table were
500 units, you would adjust this value up to 1,000 units. Look at the
second discount in the table. Order quantities between 1,000 and 1,999
will qualify for the 4% discount. Thus, if Q* is below 1,000 units, we will
adjust the order quantity up to 1,000 units.
The reasoning for this step may not be obvious. If the order
quantity, Q*, is below the range that will qualify for the discount, a
quantity within this range may still result in the lowest possible cost.
As shown in the graph, the total cost curve is broken into three
different total cost curves. There is a total cost curve for the first
(0<=Q<=999), second (1,000<=Q<=1,999), and third (Q=>2,000) discount.
Look at the total cost (TC) curve for discount 2. Q* for discount 2 is less
than the allowable discount range, which is from 1,000 to 1,999 units. As
the figure shows, the lowest allowable quantity in this range, which is
1,000 units, is the quantity that minimizes total cost.
Thus, the second step is needed to ensure that we do not discard an order
quantity that may indeed produce the minimum cost. Note that an order
quantity computed in Step 1 is greater than the range that would qualify
it for a discount may be discarded.

Step 3: Using the preceding total cost equation, compute a total cost for
every Q* determined in steps 1 & 2. If you had to adjust Q* upward
because it was below the allowable quantity range, be sure to use the
adjusted value for Q*.

Step 4: Select the Q* that has the lowest total cost, as computed in Step 3.
It will be the quantity that will minimize the total inventory cost.
Total Cost Curve for Quantity Discount Model:
Example Problem:

Wohl’s Discount Store stocks toy race cars. Recently, the store has been
given a quantity discount schedule for these cars. This quantity discount
was shown in the following table. Thus, the normal cost for the toy race
cars is $5.00. For orders between 1,000 and 1,999 units, the unit cost
drops to $4.80; for orders of 2,000 or more units, the unit cost is only
$4.75. Furthermore, ordering cost is $49.00 per order, annual demand is
5,000 race cars, and inventory carrying charge, as percent of cost I is
20%. What order quantity will minimize the total inventory cost?
SOLUTION:

Step 1:

Step 2:
Step 3:

Step 4: The lowest total inventory cost is $24,725 under Discount 2 with
order quantity of 1,000 toy race cars per order.
ROBUST MODEL:

- A model that gives satisfactory answers even with substantial variation

in its parameters.

A benefit of the EOQ model is that it is robust. By robust we mean that it

gives satisfactory answers even with substantial variation in its
parameters. As we have observed, determining accurate ordering costs
and holding costs for inventory is often difficult. Consequently, a robust
model is advantageous. Total cost of the EOQ changes little in the
neighborhood of the minimum. The curve is very shallow. This means that
variations in setup costs, holding costs, demand, or even EOQ make
relatively small differences in total cost.

RE-ORDER POINT:

It is the inventory level at which an order should be placed. It answers

the question ‘when to order’.
The Re-order Point Curve:

Q* is the optimum order quantity, and lead time represents the time
between placing and receiving an order.
Formula for Re-order Point:

This equation for ROP assumes that demand during lead time and lead
time itself are constant. When this is not the case, extra stock, often
called safety stock should be added.

The demand per day, d, is found by dividing the annual demand, D, by

the number of working days in a year:
Formula for Re-order Point:

Thus, when inventory stock drops to 96, an order should be placed. The
order will arrive 3 days later, just as the firm’s stock is depleted.

Safety stock is especially important in firms whose raw material deliveries

may be uniquely unreliable. For example, San Miguel Corp. in the
Philippines uses cheese curd imported from Europe. Because the normal
mode of delivery is lengthy and variable, safety stock may be substantial.
PROBABILISTIC MODELS & SAFETY STOCK:

• This inventory model apply when product demand is not known but
can be specified by means of a probability distribution. These types of
models are called probabilistic models.
• An important concern of management is maintaining an adequate
service level in the face of uncertain demand.
• The service level is the complement of the probability of a stockout.
For instance, if the probability of a stockout is 0.05, then the service
level is 0.95.
• Uncertain demand raises the probability of a stockout. One method of
reducing stockout is to hold extra units in inventory, which is usually
referred to as safety stock.
• As we recall from our previous discussion:

ROP = dL
where: d = daily demand
L = lead time, or the number of working days it takes to
deliver an order

The inclusion of safety stock (SS) changes the expression to:

ROP = dL + SS
Example Problem:
Solution:
PROBABILISTIC MODELS & SAFETY STOCK:

• When it is difficult or impossible to determine the cost of being out of

stock, a manager may decide to follow a policy of keeping enough
safety stock on hand to meet a prescribed customer service level.
• The manager may want to define the service level as meeting 95% of
the demand (or conversely, having stockouts only 5% of the time).
Assuming that demand during lead time (the reorder period) follows a
normal curve, only the mean and the standard deviation are needed to
define the inventory requirements for any given service level.
• Sales data are usually adequate for computing the mean and standard
deviation.
• To determine the reorder point and safety stock necessary for any
given service level, we use the formula:
Example Problem:
OTHER PROBABILISTIC MODELS:
Example Problem:
Example Problem:
Example Problem:
END OF PRESENTATION
DECISION MAKING TOOLS
THE DECISION PROCESS IN OPERATIONS:

• Operations managers are decision makers. To achieve the goals

of their organization, managers must understand how
decisions are made and know which decision tools to use.
• To a great extent, the success or failure of both people and
companies depends on the quality of their decisions.
• A good decision uses analytical decision making process that is
based on logic and considers all the available data and possible
alternatives.
• Decision making process usually follows these six steps:
1. Clearly define the problem and the factors that
influence it.
2. Develop specific and measurable objectives.
3. Develop a model that shows a relationship between
objectives and variables which are both measurable
quantities.
4. Evaluate each alternative solution based on its merits
and drawbacks.
5. Select the best alternative.
6. Implement the decision and set a timetable for
completion.
FUNDAMENTALS OF DECISION MAKING:

• There are variety of tools that can be used in decision making

such as decision tables and decision trees.
• Regardless of the complexity of a decision or the sophistication
of the technique used to analyze it, all decision makers are
faced with alternatives and states of nature.
• Alternative is a course of action or strategy that may be
chosen by a decision maker (for example, not carrying an
umbrella).
• State of nature is an occurrence or a situation over which the
decision maker has little or no control (for example, tomorrow
weather).
• Symbols used in a decision tree:
1. Square – decision node from which one of several
alternatives may be selected.
2. Circle – a state of nature node out of which one state of
nature will occur.
• To present a manager’s decision alternatives, we can develop
decision trees using the above symbols.
FUNDAMENTALS OF DECISION MAKING:

• When constructing a decision tree, we must be sure that all

alternatives and states of nature are in their correct and logical
places and that we include all possible alternatives and states
of nature.
EXAMPLE OF DECISION TREE:
DECISION TABLES:

• Decision or payoff table can also be developed to define

alternatives.
• For any alternative and a particular state of nature, there is a
consequence or outcome which is usually expressed as
monetary value. This is called conditional value.
• Usually, alternatives are listed down the left side of the table,
the states of nature are listed across the top, and the
conditional values (payoffs) are in the body of decision table.
EXAMPLE: Decision table for Getz Products including payoffs:
TYPES OF DECISION-MAKING ENVIRONMENTS:

• The types of decisions people make depend on how much

knowledge or information they have about the situation.
• There are 3 decision making environments:
1. Decision making under uncertainty
2. Decision making under risk
3. Decision making under certainty
DECISION MAKING UNDER UNCERTAINTY:

• When there is complete uncertainty as to which state of nature

in a decision environment may occur where we cannot even
assess probabilities for each possible outcome, we rely on
these 3 methods:
• 1. Maximax – this method finds an alternative that maximizes
the maximum outcome for every alternative. First, we find the
maximum outcome within every alternative, and then we pick
the alternative with the maximum number. Because this
decision criterion locates the alternative with the highest
possible gain, it has been called an ‘optimistic’ decision
criterion.
DECISION MAKING UNDER UNCERTAINTY:

• 2. Maximin – this method finds the alternative that maximizes

the minimum outcome for every alternative. First, we find the
minimum outcome within every alternative, and then we pick
the number with the maximum number. Because this decision
criterion locates the alternative that has the least possible loss,
it has been called a ‘pessimistic’ decision criterion.

• 3. Equally likely – this method finds the alternative with the

highest average outcome. First, we calculate the average
outcome for every alternative, which is the sum of all outcomes
divided by the number of outcomes. We then pick the
alternative with the maximum number. The equally likely
approach assumes that each state of nature is equally likely to
occur.
EXAMPLE:
ANSWER:

1. The maximax choice is to construct a large plant. This is the

maximum of the maximum number within each row, or
alternative.
2. The maximin choice is to do nothing. This is the maximum of
the minimum number within each row, or alternative.
3. The equally likely choice is to construct a small plant. This is
the maximum of the average outcome of each alternative. This
approach assumes that all outcomes for any alternative are
equally likely.

DECISION MAKING UNDER RISK:

• Decision making under risk, a more common occurrence,

relies on probabilities.
• Several possible states of nature may occur, each with an
assumed probability. The states of nature must be mutually
exclusive and collectively exhaustive and their probabilities
must sum to 1.
DECISION MAKING UNDER RISK:

• Given a decision table with conditional values and probability

assessments for all states of nature, we can determine the
expected monetary value (EMV) for each alternative.
• This figure represents the expected value or mean return for
each alternative if we could repeat the decision a large number
of times.
• The EMV for an alternative is the sum of all possible payoffs
from the alternative, each weighted by the probability of that
payoff occurring.
DECISION MAKING UNDER RISK:

EXAMPLE:

Getz Products operations manager believes that the probability of

a favorable market is exactly the same as that of an unfavorable
market; that is, each state of nature has a 0.50 chance of
occurring. We can now determine the EMV for each alternative.
DECISION MAKING UNDER CERTAINTY:
DECISION MAKING UNDER CERTAINTY:
DECISION MAKING UNDER CERTAINTY:
DECISION TREES:

• Although the use of a decision table is convenient in problems

having one set of decisions and one set of states of nature,
many problems include sequential decisions and states of
nature.
• When there are two or more sequential decisions, and later
decisions are based on the outcome of prior ones, the decision
tree becomes appropriate.
• Decision tree is a graphic display of the decision process that
indicates decision alternatives, states of nature and their
respective probabilities and payoff for each combination of
decision alternative and state of nature.
• Expected monetary value (EMV) is the most commonly used
criterion for decision tree analysis. One of the first steps in
decision analysis is to graph the decision tree and to specify
the monetary consequences of all outcomes for a particular
problem.
DECISION TREES:

• Analyzing problems with decision trees involves five steps:

1. Define the problem.
2. Structure or draw the decision tree.
3. Assign probabilities to the states of natures.
4. Estimate payoff for each possible combination of
decision alternatives and states of nature.
5. Solve the problem by computing expected monetary
values (EMV) for each state-of-nature node. This is done
by working backward-that is, by starting at the right of
the tree and working back to decision nodes on the left.
EXAMPLE OF A DECISION TREE:
A MORE COMPLEX DECISION TREE:

• When a sequence of decisions must be made, decision trees are

much more powerful tools than the decision tables.

EXAMPLE:

Let’s say that Getz Products has two decisions to make, with the
second decision dependent on the outcome of the first. Before
deciding about building a new plant, Getz has the option of
conducting its own marketing research survey, at a cost of
$10,000. The information from this survey could help it decide
whether to build a large plant, to build a small plant, or not to
build at all.. Getz recognizes that although such a survey will not
provide it with perfect information, it may be extremely helpful.
END OF PRESENTATION
GAME THEORY
INTRODUCTION TO GAME THEORY:

• Life have competitions. Numerous examples involving

adversaries in conflict include parlor games, military battles,
political campaigns, advertising and marketing campaigns by
competing business firms, and so forth.
• A basic feature in many of these situations is that the final
outcome depends primarily upon the combination of strategies
selected by the adversaries.
• Game theory is a mathematical theory that deals with the
general features of competitive situations like these in a
formal, abstract way. It places particular emphasis on the
decision-making processes of the adversaries.
• The focus in this chapter is on the simplest case, called two-
person, zero-sum games. As the name implies, these games
involve only two adversaries or players (who may be armies,
teams, firms, and so on). They are called zero-sum games
because one player wins whatever the other one loses, so that
the sum of their net winnings is zero.
THE FORMULATION OF TWO-PERSON, ZERO-SUM GAMES:

• To illustrate the basic characteristics of two-person, zero-sum

games, consider the game called odds and evens. This game
consists simply of each player simultaneously showing either
one finger or two fingers.

• If the number of fingers matches, so that the total number for

both players is even, then the player taking evens (say, player
1) wins the bet (say, $1) from the player taking odds (player 2).
If the number does not match, player 1 pays $1 to player 2.

• Thus, each player has two strategies: to show either one finger
or two fingers.

• In general, a two-person game is characterized by:

1. The strategies of player 1
2. The strategies of player 2
3. The payoff table
THE FORMULATION OF TWO-PERSON, ZERO-SUM GAMES:

• Before the game begins, each player knows the strategies she
or he has available, the ones the opponent has available, and
the payoff table. The actual play of the game consists of each
player simultaneously choosing a strategy without knowing
the opponent’s choice.

• In more complicated games involving a series of moves, a

strategy is a predetermined rule that specifies completely how
one intends to respond to each possible circumstance at each
stage of the game. For example, a strategy for one side in chess
would indicate how to make the next move for every possible
position on the board, so the total number of possible
strategies would be astronomical.

• The payoff table shows the gain (positive or negative) for

player 1 that would result from each combination of strategies
for the two players. It is given only for player 1 because the
table for player 2 is just the negative of this one, due to the
zero-sum nature of the game.
THE FORMULATION OF TWO-PERSON, ZERO-SUM GAMES:

• The entries in the payoff table may be in any units desired,

such as dollars, provided that they accurately represent the
utility (or benefit) to player 1 of the corresponding outcome.
However, utility is not necessarily proportional to the amount
of money (or any other commodity) when large quantities are
involved.

• For example, $2 million (after taxes) is probably worth much

less than twice as much as $1 million to a poor person. In
other words, given the choice between (1) a 50 percent chance
of receiving $2 million rather than nothing and (2) being sure
of getting $1 million, a poor person probably would much
prefer the latter.

• A primary objective of game theory is the development of

rational criteria for selecting a strategy. Two key assumptions
are made:
1. Both players are rational.
2. Both players choose their strategies solely to promote
their own welfare (no compassion for the opponent).
THE FORMULATION AS A TWO-PERSON, ZERO-SUM GAME:

• To formulate this problem as a two-person, zero-sum game, we

must identify the two players (obviously the two politicians),
the strategies for each player, and the payoff table.

• As the problem has been stated, each player has the following
three strategies:
Strategy 1 - spend 1 day in each city.
Strategy 2 - spend both days in Bigtown.
Strategy 3 - spend both days in Megalopolis.

• Each entry in the payoff table for player 1 represents the utility
(or benefit) to player 1 (or the negative utility to player 2) of the
outcome resulting from the corresponding strategies used by
the two players.
• From the politician’s viewpoint, the objective is to win votes,
and each additional vote (before he learns the outcome of the
election) is of equal value to him.
THE FORMULATION AS A TWO-PERSON, ZERO-SUM GAME:

• Therefore, the appropriate entries for the payoff table for

politician 1 are the total net votes won from the opponent (i.e.,
the sum of the net vote changes in the two cities) resulting
from these 2 days of campaigning.
• Using units of 1,000 votes, this formulation is summarized in
the table.
• Game theory assumes that both players are using the same
formulation (including the same payoffs for player 1) for
choosing their strategies.
DERIVING FOR THE SOLUTION TO TWO-PERSON, ZERO-SUM
GAMES:

• This situation is a rather special one, where the answer can be

obtained just by applying the concept of dominated strategies
to rule out a succession of inferior strategies until only one
choice remains.
• A strategy is dominated by a second strategy if the second
strategy is always at least as good (and sometimes better)
regardless of what the opponent does. A dominated strategy
can be eliminated immediately from further consideration.
• At the outset, the table includes no dominated strategies for
player 2. However, for player 1, strategy 3 is dominated by
strategy 1 because the latter has larger payoffs (1>0, 2>1, 4>1)
regardless of what player 2 does. Eliminating strategy 3 from
further consideration yields the following reduced payoff table:
DERIVING FOR THE SOLUTION TO TWO-PERSON, ZERO-SUM
GAMES:

• Because both players are assumed to be rational, player 2 also

can deduce that player 1 has only these two strategies
remaining under consideration. Therefore, player 2 now does
have a dominated strategy — strategy 3, which is dominated
by both strategies 1 and 2 because they always have smaller
losses for player 2 (payoffs to player 1) in this reduced payoff
table (for strategy 1: 1<4, 1<5; for strategy 2: 2<4, 0<5).
Eliminating this strategy yields.

• At this point, strategy 2 for player 1 becomes dominated by

strategy 1 because the latter is better in column 2 (2 > 0) and
equally good in column 1 (1 = 1). Eliminating the dominated
strategy leads to
DERIVING FOR THE SOLUTION TO TWO-PERSON, ZERO-SUM
GAMES:

• Strategy 2 for player 2 now is dominated by strategy 1 (1<2), so

strategy 2 should be eliminated.
• Consequently, both players should select their strategy 1.
Player 1 then will receive a payoff of 1 from player 2 (that is,
politician 1 will gain 1,000 votes from politician 2).
• In general, the payoff to player 1 when both players play
optimally is referred to as the value of the game. A game that
has a value of 0 is said to be a fair game. Since this
particular game has a value of 1, it is not a fair game.
DERIVING FOR THE SOLUTION TO TWO-PERSON, ZERO-SUM
GAMES:

• The concept of a dominated strategy is a very useful one for

reducing the size of the payoff table that needs to be
considered and, in unusual cases like this one, actually
identifying the optimal solution for the game. However, most
games require another approach to at least finish solving, as
illustrated by the next two variations of the example.
VARIATION 2 OF THE EXAMPLE:

• Now suppose that the current data given in the table as the
payoff table for player 1 (politician 1). This game does not have
dominated strategies, so it is not obvious what the players
should do. What line of reasoning does game theory say they
should use?

• Consider player 1. By selecting strategy 1, he could win 6 or

could lose as much as 3. However, because player 2 is rational
and thus will seek a strategy that will protect himself from
large payoffs to player 1, it seems likely that player 1 would
incur a loss by playing strategy 1.
VARIATION 2 OF THE EXAMPLE:

• Similarly, by selecting strategy 3, player 1 could win 5, but

more probably his rational opponent would avoid this loss and
instead administer a loss to player 1 which could be as large
as 4. On the other hand, if player 1 selects strategy 2, he is
guaranteed not to lose anything and he could even win
something.
• Therefore, because it provides the best guarantee (a payoff of
0), strategy 2 seems to be a “rational” choice for player 1
against his rational opponent. (This line of reasoning assumes
that both players are averse to risking larger losses than
necessary, in contrast to those individuals who enjoy gambling
for a large payoff against long odds.)
• Now consider player 2. He could lose as much as 5 or 6 by
using strategy 1 or 3, but is guaranteed at least breaking even
with strategy 2. Therefore, by the same reasoning of seeking
the best guarantee against a rational opponent, his apparent
choice is strategy 2.
VARIATION 2 OF THE EXAMPLE:

• If both players choose their strategy 2, the result is that both

break even. Thus, in this case, neither player improves upon
his best guarantee, but both also are forcing the opponent into
the same position. Even when the opponent deduces a player’s
strategy, the opponent cannot exploit this information to
improve his position. Stalemate.
VARIATION 2 OF THE EXAMPLE:

• The end product of this line of reasoning is that each player

should play in such a way as to minimize his maximum losses
whenever the resulting choice of strategy cannot be exploited.
• This so-called minimax criterion is a standard criterion
proposed by game theory for selecting a strategy.
• In effect, this criterion says to select a strategy that would be
best even if the selection were being announced to the
opponent before the opponent chooses a strategy.
• In terms of the payoff table, it implies that player 1 should
select the strategy whose minimum payoff is largest, whereas
player 2 should choose the one whose maximum payoff to
player 1 is the smallest.
• This criterion is illustrated in table, where strategy 2 is
identified as the maximin strategy for player 1 and strategy 2 is
the minimax strategy for player 2. The resulting payoff of 0 is
the value of the game, so this is a fair game.
VARIATION 2 OF THE EXAMPLE:

• Notice the interesting fact that the same entry in this payoff
table yields both the maximin and minimax values. The reason
is that this entry is both the minimum in its row and the
maximum of its column. The position of any such entry is
called a saddle point.
• The fact that this game possesses a saddle point was actually
crucial in determining how it should be played. Because of the
saddle point, neither player can take advantage of the
opponent’s strategy to improve his own position.
• In particular, when player 2 predicts or learns that player 1 is
using strategy 2, player 2 would incur a loss instead of
breaking even if he were to change from his original plan of
using his strategy 2. Similarly, player 1 would only worsen his
position if he were to change his plan.
• Thus, neither player has any motive to consider changing
strategies, either to take advantage of his opponent or to
prevent the opponent from taking advantage of him.
VARIATION 2 OF THE EXAMPLE:

• Therefore, since this is a stable solution (also called an

equilibrium solution), players 1 and 2 should exclusively use
their maximin and minimax strategies, respectively.

VARIATION 3 OF THE EXAMPLE:

• Late developments in the campaign result in the final payoff

table for player 1 (politician 1) given by table. How should this
game be played?
VARIATION 3 OF THE EXAMPLE:

• Suppose that both players attempt to apply the minimax

criterion in the same way as in variation 2. Player 1 can
guarantee that he will lose no more than 2 by playing strategy
1. Similarly, player 2 can guarantee that he will lose no more
than 2 by playing strategy 3.
• However, notice that the maximin value (-2) and the minimax
value (2) do not coincide in this case. The result is that there is
no saddle point.
• What are the resulting consequences if both players plan to
use the strategies just derived?
• It can be seen that player 1 would win 2 from player 2, which
would make player 2 unhappy. Because player 2 is rational
and can therefore foresee this outcome, he would then
conclude that he can do much better, actually winning 2
rather than losing 2, by playing strategy 2 instead. Because
player 1 is also rational, he would anticipate this switch and
conclude that he can improve considerably, from 2 to 4, by
changing to strategy 2. Realizing this, player 2 would then
VARIATION 3 OF THE EXAMPLE:

• consider switching back to strategy 3 to convert a loss of 4 to a

gain of 3. This possibility of a switch would cause player 1 to
consider again using strategy 1, after which the whole cycle
would start over again. Therefore, even though this game is
being played only once, any tentative choice of a strategy
leaves that player with a motive to consider changing
strategies, either to take advantage of his opponent or to
prevent the opponent from taking advantage of him.
• In short, the originally suggested solution (player 1 to play
strategy 1 and player 2 to play strategy 3) is an unstable
solution, so it is necessary to develop a more satisfactory
solution.
• But what kind of solution should it be?
GAMES WITH MIXED STRATEGIES:

• Whenever a game does not possess a saddle point, game

theory advises each player to assign a probability distribution
over her set of strategies. To express this mathematically, let

• where m and n are the respective numbers of available

strategies.
• Thus, player 1 would specify her plan for playing the game by
assigning values to x1, x2, . . . , xm. Because these values are
probabilities, they would need to be non-negative and add to 1.
Similarly, the plan for player 2 would be described by the
values she assigns to her decision variables y1, y2, . . . , yn.
These plans (x1, x2, . . . , xm) and (y1, y2, . . . , yn) are usually
referred to as mixed strategies, and the original strategies are
then called pure strategies.
GAMES WITH MIXED STRATEGIES:

• When the game is actually played, it is necessary for each

player to use one of her pure strategies. However, this pure
strategy would be chosen by using some random device to
obtain a random observation from the probability distribution
specified by the mixed strategy, where this observation would
indicate which particular pure strategy to use.
• To illustrate, suppose that players 1 and 2 in variation 3 of the
political campaign problem (see the preceding table) select the
mixed strategies (x1, x2, x3) = (1/2, 1/2, 0) and (y1, y2, y3) =
(0, 1/2, 1/2) respectively. This selection would say that player
1 is giving an equal chance of choosing strategy 1 & 2 but he is
discarding strategy 3 entirely.
• Similarly, player 2 is randomly choosing between his last two
pure strategies. To play the game, each player could then flip a
coin to determine which of his two acceptable pure strategies
he will actually use.
• Although no completely satisfactory measure of performance is
available for evaluating mixed strategies, a very useful one is
the expected payoff. By applying the probability theory
definition of expected value, this quantity is
GAMES WITH MIXED STRATEGIES:

• where pij is the payoff if player 1 uses pure strategy i and

player 2 uses pure strategy j. In the example of mixed
strategies just given, there are four possible payoffs (-2, 2, 4, -
3), each occurring with a probability of 1/4, so the expected
payoff is ¼ (-2 + 2 + 4 – 3) = ¼.
• Thus, this measure of performance does not disclose anything
about the risks involved in playing the game, but it does
indicate what the average payoff will tend to be if the game is
played many times.
• By using this measure, game theory extends the concept of the
minimax criterion to games that lack a saddle point and thus
need mixed strategies. In this context, the minimax criterion
says that a given player should select the mixed strategy that
minimizes the maximum expected loss to himself.
GAMES WITH MIXED STRATEGIES:

• Equivalently, when we focus on payoffs (player 1) rather than

losses (player 2), this criterion says to maximin instead, i.e.,
maximize the minimum expected payoff to the player.
• By the minimum expected payoff we mean the smallest possible
expected payoff that can result from any mixed strategy with
which the opponent can counter. Thus, the mixed strategy for
player 1 that is optimal according to this criterion is the one
that provides the guarantee (minimum expected payoff) that is
best (maximal). (The value of this best guarantee is the
maximin value, denoted by v. )
• Similarly, the optimal strategy for player 2 is the one that
provides the best guarantee, where best now means minimal
and guarantee refers to the maximum expected loss that can be
administered by any of the opponent’s mixed strategies. (This
best guarantee is the minimax value, denoted by v.)
• Recall that when only pure strategies were used, games not
having a saddle point turned out to be unstable (no stable
solutions). The reason was essentially that < , so that the
players would want to change their strategies to improve their
positions.
GAMES WITH MIXED STRATEGIES:

• Similarly, for games with mixed strategies, it is necessary that

v v for the optimal solution to be stable. Fortunately,
according to the minimax theorem of game theory, this
condition always holds for such games.
• Minimax theorem: If mixed strategies are allowed, the pair of
mixed strategies that is optimal according to the minimax
criterion provides a stable solution with (the value of
the game), so that neither player can do better by unilaterally
changing her or his strategy.
• Although the concept of mixed strategies becomes quite
intuitive if the game is played repeatedly, it requires some
interpretation when the game is to be played just once. In this
case, using a mixed strategy still involves selecting and using
one pure strategy (randomly selected from the specified
probability distribution), so it might seem more sensible to
ignore this randomization process and just choose the one
“best” pure strategy to be used.
• However, we have already illustrated for variation 3 in the
preceding section that a player must not allow the opponent to
deduce what his strategy will be (i.e., the solution procedure
GAMES WITH MIXED STRATEGIES:

• However, we have already illustrated for variation 3 in the

preceding section that a player must not allow the opponent to
deduce what his strategy will be (i.e., the solution procedure
under the rules of game theory must not definitely identify
which pure strategy will be used when the game is unstable).
• Furthermore, even if the opponent is able to use only his
knowledge of the tendencies of the first player to deduce
probabilities (for the pure strategy chosen) that are different
from those for the optimal mixed strategy, then the opponent
still can take advantage of this knowledge to reduce the
expected payoff to the first player.
GRAPHICAL SOLUTION PROCEDURE:

• Consider any game with mixed strategies such that, after

dominated strategies are eliminated, one of the players has
only two pure strategies.
• To be specific, let this player be player 1. Because her mixed
strategies are (x1, x2) and x2 = 1 - x1, it is necessary for her to
solve only for the optimal value of x1. However, it is
straightforward to plot the expected payoff as a function of x1
for each of her opponent’s pure strategies. This graph can then
be used to identify the point that maximizes the minimum
expected payoff. The opponent’s minimax mixed strategy can
also be identified from the graph.
• To illustrate this procedure, consider variation 3 of the
political campaign problem (see the following table). Notice
that the third pure strategy for player 1 is dominated by her
second, so the payoff table can be reduced to the form given in
the table. Therefore, for each of the pure strategies available to
player 2, the expected payoff for player 1 will be
GRAPHICAL SOLUTION PROCEDURE:
GRAPHICAL SOLUTION PROCEDURE:

• Now plot these expected-payoff lines on a graph, as shown in

the figure. For any given values of x1 and (y1, y2, y3), the
expected payoff will be the appropriate weighted average of the
corresponding points on these three lines. In particular,
GRAPHICAL SOLUTION PROCEDURE:

• Remember that player 2 wants to minimize this expected

payoff for player 1. Given x1, player 2 can minimize this
expected payoff by choosing the pure strategy that corresponds
to the “bottom” line for that x1 in the figure (either -3 + 5x1 or 4
- 6x1, but never 5 - 5x1). According to the minimax (or
maximin) criterion, player 1 wants to maximize this minimum
expected payoff. Consequently, player 1 should select the value
of x1 where the bottom line peaks, i.e., where the (-3 + 5x1) and
(4 - 6x1) lines intersect, which yields an expected payoff of

• To solve algebraically for this optimal value of x1 at the

intersection of the two lines -3 + 5x1 and 4 - 6x1, we set
WAITING LINE MODELS &
QUEUEING THEORY
WAITING LINE MODELS & QUEUEING THEORY:

• The body of knowledge about waiting lines, often called

queueing theory, is an important part of operations and a
valuable tools for operations manager.
• Waiting lines are a common situation – they may, for example,
(1) take the form of cars waiting for repair at a Midas Muffler
Shop, (2) copying jobs waiting to be completed at a Kinko’s
print shop, or (3) vacationers waiting to enter Mr. Toad’s Wild
Ride at Disney.
• Waiting-line models are useful in both manufacturing and
service areas. Analysis of queues in terms of waiting length,
average waiting time, and other factors helps us to understand
service systems (such as bank teller stations), maintenance
activities (that might repair broken machinery), and shop-floor
control activities.
• Indeed, patients waiting in a doctors office and broken drill
presses waiting in a repair facility have a lot in common from
an operations management perspective. Both use human and
equipment resources to restore valuable production assets
(people and machine) to good condition.
WAITING LINE MODELS & QUEUEING THEORY:

• The following table lists just few OM uses of waiting-line

models:
CHARACTERISTICS OF A WAITING-LINE SYSTEM:

• The 3 components of a waiting-line or queueing system are

as follows:
1. Arrivals or inputs to the system. These have
characteristics such as population size, behavior and a
statistical distribution.
2. Queue discipline, or the waiting line itself.
Characteristics of the queue include whether it is limited
or unlimited in length and the discipline of people or items
in it.
3. The service facility. Its characteristics include its design
and the statistical distribution of service times.

ARRIVAL CHARACTERISTICS:

• The input source that generates arrivals or customers for a

service system has three major characteristics:
1. Size of the arrival population
2. Behavior of arrivals
3. Pattern of arrivals (statistical distribution)
ARRIVAL CHARACTERISTICS:

• 1. Size of the arrival population: Population sizes are considered

either unlimited (essentially infinite) or limited (finite).
• When the number of customers or arrivals on hand at any
given moment is just a small portion of all potential arrivals,
the arrival population is considered unlimited or infinite.
• Examples of unlimited populations include cars arriving at a
big-city car wash, shoppers arriving at supermarket, and
students arriving to register for classes at a large university.
Most queueing models assume such an infinite arrival
population.
• An example of a limited or finite population is found in copying
shop that has, say, eight copying machines. Each of the
copiers is a potential “customer” that may break down and
require service.
• 2. Pattern of arrivals at the system: Customers arrive at a
service facility according to some known schedule (for
example, one patient for every 15 minutes or one student for
every half an hour) or else they arrive randomly.
ARRIVAL CHARACTERISTICS:

• Arrivals are considered random when they are independent to

one another and their occurrence cannot be predicted exactly.
• Frequently in queueing problems, the number of arrivals per
unit of time can be estimated by a probability distribution
known as poisson distribution. This is a discrete probability
distribution that often describes the arrival rate in queueing
theory.
• For any given arrival time (such as 2 customers per hour or 4
trucks per minute), a discrete poisson distribution can be
established by using the formula:
ARRIVAL CHARACTERISTICS:

• The following table illustrates the poisson distribution for λ=2

and λ=4.

• This means that if the average arrival rate is 2 customers per

hour, the probability of 0 customer arriving in any random
hour is about 13%, probability of 1 customer is about 27%, 2
customers about 27%, 3 customers about 18% and so on.
ARRIVAL CHARACTERISTICS:

• Arrivals, of course, are not always poisson distributed. They

may follow some other distribution.
• Patterns, therefore, should be examined to make certain that
they are well approximated by poisson before that distribution
is applied.
• 3. Behavior of arrivals: Most queueing models assume that an
arriving customer is a patient customer. Patient customers are
people or machines that wait in the queue until they are
served and do not switch between lines.
• Unfortunately, life is complicated by the fact that people have
been known to balk, renege or jockey.
• Balking happens when a customer refuse to join the waiting
line because it is too long to suit their interest or needs.
• Reneging happens when a customer who enter the queue but
then become impatient and leave without completing their
transaction.
• Jockeying happens when a customer switch to other queue in
a multi-server queueing system.
• Actually, all of these situations just serve to highlight the need
for queueing theory and waiting-line analysis.
WAITING-LINE CHARACTERISTICS:

• The waiting line itself is the second component of a queueing

system. The length of a line can be either limited or unlimited.
• A queue is limited when it cannot, either by law or because of
physical restrictions, increase to an infinite length.
• A small barbershop, for example, will only have a limited
number of waiting chairs.
• Queueing models are treated in this module under an
assumption of unlimited queue length. A queue is unlimited
when its size is unrestricted, as in the case of toll booth
serving arriving automobiles.
• A second waiting-line characteristics deals with queue
discipline. This refers to the rule by which customers in the
line are to receive service. Most systems use a queue discipline
known as the first-in, first-out (FIFO) rule.
• Patient who are critically injured will move ahead in treatment
priority over patients with broken fingers or noses. Shoppers
with fewer than 10 items may be allowed to enter the express
checkout queue (but are then treated as first-come, first-
served).
SERVICE CHARACTERISTICS:

• The third part of any queueing system are the service

characteristics. Two basic properties are important: (1) design
of the service system and (2) the distribution of service times.
• 1. Basic queueing system designs: Service systems are usually
classified in terms of their number of channels (for example,
number of servers) and number of phases (for example,
number of service stops that must be made).
• A single-channel queueing system, with one server, is typified
by the drive-in bank with only one open teller.
• If, on the other hand, the bank has several tellers on-duty,
with each customer waiting in one common line for the first
available teller, then we would have multiple-channel queueing
system. Most banks, large barbershops, airline ticket counters
and post offices are multi-channel service systems.
• In a single-phase system, the customer receives service from
only one station and then exists the system. A fast-food
restaurant in which the person who takes your order also
brings your food and takes your money is a single-phase
system.
SERVICE CHARACTERISTICS:
• However, say the restaurant requires you to place your order
at one station, pay at a second and pick up your food at a
third; it is a multi-phase system.
SERVICE CHARACTERISTICS:

• 2. Service time distribution: Service patterns are like arrival

patterns in that they may be either constant or random.
• If service time is constant, it takes the same amount of time to
take care of each customer. This is the case in a machine-
performed service operation such as an automatic car wash.
• More often, service times are randomly distributed. In many
cases, we can assume that random service times are described
by the negative exponential probability distribution.
• The following figure shows that if the service times follow a
negative exponential distribution, the probability of any very
long service time is low. For example, when an average service
time is 20 minutes (or 3 customers per hour), seldom if ever
will a customer require more than 1.5 hours in the service
facility.
SERVICE CHARACTERISTICS:
QUEUEING COST:
QUEUEING COST:

• Operations Analyst (or manager) must recognize the trade-off

that takes place between two queueing cost; the cost of
providing good service and the cost of customer or machine
waiting time.
• Operations Analyst (or manager) want queues that are short
enough so that customers do not become unhappy and either
leave without buying or buy but never return.
• However, Operations Analyst (or manager) may be willing to
allow some waiting if it is balanced by a significant savings in
service cost.
• One means of evaluating a service facility is to look at total
expected cost. Total cost is the sum of expected service cost plus
expected waiting cost.
• As shown in the graph on the next page, service costs increase
as a firm attempts to raise its service level.
• Managers in some service centers can vary capacity by having
standby personnel and machines that they can assign to
specific service stations to prevent or shorten excessively long
lines.
QUEUEING COST:

• In grocery stores, for example, managers and stock clerks can

open extra checkout counters.
• Waiting cost may reflect lost productivity of workers because of
poor service and long queues.
THE VARIETY OF QUEUEING MODELS:

• A wide variety of queueing models may be applied in

operations management. In the following table are the most
widely used models.
• All waiting line models presented have three characteristics in
common. (1) Poisson distribution arrivals, (2) FIFO discipline,
(3) A single-service phase.
M/M/1 SYSTEM: SINGLE-CHANNEL QUEUEING MODEL WITH
POISSON ARRIVALS & EXPONENTIAL SERVICE TIMES:
• The most common case of queueing problems involves the
single-channel, or single-server, waiting line. In this situation,
arrivals form a single line to be serviced by a single station.
• We assume that the following conditions exist in this type of
system.
1. Arrivals are served on a first-in first-out (FIFO) basis,
and every arrival waits to be served, regardless of the
length of the line or queue.
2. Arrivals are independent of preceding arrivals, but the
average number of arrivals (arrival rate) does not change
over time.
3. Arrivals are described by a Poisson probability
distribution and come from an infinite (or very, very large
population).
4. Service time vary from one customer to the next and
are independent of one another, but their average rate is
known.
5. Service times occur according to the negative
exponential probability distribution.
6. The service rate is faster than the arrival rate.
M/M/1 SYSTEM: SINGLE-CHANNEL QUEUEING MODEL WITH
POISSON ARRIVALS & EXPONENTIAL SERVICE TIMES:

• When these conditions are met, the following series of

equations can be developed.
EXAMPLE PROBLEM & SOLUTION:
EXAMPLE PROBLEM & SOLUTION:
EXAMPLE PROBLEM & SOLUTION:
END OF PRESENTATION