Maximin strategy A strategy where the player seeking to maximize the value of the

game selects the strategy that maximizes the minimum payoff obtainable by the other
player.
Minimax strategy A strategy where the player seeking to minimize the value of the
game selects the strategy that minimizes the maximum payoff obtainable by the other
player.
Pure strategy When one of the available strategies is optimal and the player always
selects this strategy regardless of the strategy selected by the other player.
Saddle point A condition that exists when pure strategies are optimal for both players.
Neither player can improve the value of the game by changing from the optimal pure
strategy.
Mixed strategy When a player randomly selects its strategy based on a probability dis-
tribution. The strategy selected can vary each time the game is played.
Expected value In a mixed strategy game, a value computed by multiplying each payoff
by its probability and summing. It can be interpreted as the long-run average payoff for the
mixed strategy.
PROBLEMS
Note: The following problems have been designed to give you an understanding and apprecia-
tion of the broad range of problems that can be formulated as linear programs. You should be
able to formulate a linear programming model for each of the problems. However, you will need
access to a linear programming computer package to develop the solutions and make the
requested interpretations.
1. In Section 5.1 data envelopment analysis was used to evaluate the relative efciencies of
four hospitals. Data for three input measures and four output measures were provided in
Tables 5.1 and 5.2.
a. Use these data to develop a linear programming model that could be used to evaluate
the performance of General Hospital.
b. The following solution is optimal. Does the solution indicate that General Hospital is
relatively inefcient?
248 Chapter 5 Advanced Linear Programming Applications
Objective Function Value 1.000
Variable Value Reduced Costs
E 1.000 0.000
WG 1.000 0.000
WU 0.000 0.000
WC 0.000 0.331
WS 0.000 0.215
c. Explain which hospital or hospitals make up the composite unit used to evaluate Gen-
eral Hospital and why.
2. Data envelopment analysis can measure the relative efciency of a group of hospitals. The
following data from a particular study involving seven teaching hospitals include three
input measures and four output measures:
a. Formulate a linear programming model so that data envelopment analysis can be used
to evaluate the performance of hospital D.
b. Solve the model.
c. Is hospital D relatively inefcient? What is the interpretation of the value of the objec-
tive function?
Problems 249
Input Measures
Full-Time Supply Bed-Days
Equivalent Expense Available
Hospital Nonphysicians (1000s) (1000s)
A 310.0 134.60 116.00
B 278.5 114.30 106.80
C 165.6 131.30 65.52
D 250.0 316.00 94.40
E 206.4 151.20 102.10
F 384.0 217.00 153.70
G 530.1 770.80 215.00
Output Measures
Patient-Days Patient-Days
(65 or older) (under 65) Nurses Interns
Hospital (1000s) (1000s) Trained Trained
A 55.31 49.52 291 47
B 37.64 55.63 156 3
C 32.91 25.77 141 26
D 33.53 41.99 160 21
E 32.48 55.30 157 82
F 48.78 81.92 285 92
G 58.41 119.70 111 89
d. How many patient-days of each type are produced by the composite hospital?
e. Which hospitals would you recommend hospital D consider emulating to improve the
efciency of its operation?
3. Refer again to the data presented in Problem 2.
a. Formulate a linear programming model that can be used to perform data envelopment
analysis for hospital E.
b. Solve the model.
c. Is hospital E relatively inefcient? What is the interpretation of the value of the objec-
tive function?
d. Which hospitals are involved in making up the composite hospital? Can you make a
general statement about which hospitals will make up the composite unit associated
with a unit that is not inefcient?
4. The Ranch House, Inc., operates ve fast-food restaurants. Input measures for the restau-
rants include weekly hours of operation, full-time equivalent staff, and weekly supply
expenses. Output measures of performance include average weekly contribution to prot,
a. Develop a linear programming model that can be used to evaluate the performance of
the Clarksville Ranch House restaurant.
b. Solve the model.
c. Is the Clarksville Ranch House restaurant relatively inefcient? Discuss.
d. Where does the composite restaurant have more output than the Clarksville restau-
rant? How much less of each input resource does the composite restaurant require
when compared to the Clarksville restaurant?
e. What other restaurants should be studied to nd suggested ways for the Clarksville
restaurant to improve its efciency?
5. Reconsider the Leisure Airlines problem from Section 5.2. The demand forecasts shown
in Table 5.3 represent Leisure Airs best estimates of demand. But because demand cannot
be forecasted perfectly, the number of seats actually sold for each origin-destination-
itinerary fare (ODIF) may turn out to be smaller or larger than forecasted. Suppose that
Leisure Air believes that economic conditions have improved and that their original fore-
cast may be too low. To account for this possibility, Leisure Air is considering switching
the Boeing 737-400 airplanes that are based in Pittsburgh and Newark with Boeing 757-
200 airplanes that Leisure Air has available in other markets. The Boeing 757-200 airplane
has a seating capacity of 158 in the coach section.
a. Because of scheduling conicts in other markets, suppose that Leisure Air is only able
to obtain one Boeing 757-200. Should the larger plane be based in Pittsburgh or in
Newark? Explain.
b. Based upon your answer in part (a), determine a new allocation for the ODIFs. Briey
summarize the major differences between the new allocation using one Boeing 757-
200 and the original allocation summarized in Figure 5.5.
c. Suppose that two Boeing 757-200 airplanes are available. Determine a new allocation
for the ODIFs using the two larger airplanes. Briey summarize the major differences
between the new allocation using two Boeing 757-200 airplanes and the original allo-
cation shown in Figure 5.5.
d. Consider the new solution obtained in part (b). Which ODIF has the highest bid price?
What is the interpretation for this bid price?
250 Chapter 5 Advanced Linear Programming Applications
Input Measures
Restaurant Hours of Operation FTE Staff Supplies ($)
Bardstown 96 16 850
Clarksville 110 22 1400
Jeffersonville 100 18 1200
New Albany 125 25 1500
St. Matthews 120 24 1600
Output Measures
Restaurant Weekly Prot Market Share (%) Growth Rate (%)
Bardstown $3800 25 8.0
Clarksville $4600 32 8.5
Jeffersonville $4400 35 8.0
New Albany $6500 30 10.0
St. Matthews $6000 28 9.0
market share, and annual growth rate. Data for the input and output measures are shown in
the following tables:
6. Reconsider the Leisure Airlines problem from Section 5.2. Suppose that as of May 1 the
following number of seats have been sold:
ODIF 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Seats Sold 25 44 18 12 5 9 20 33 37 11 5 8 27 6 35 7
a. Determine how many seats are still available for sale on each ight leg.
b. Using the original demand forecasted for each ODIF, determine the remaining de-
mand for each ODIF.
c. Revise the linear programming model presented in Section 5.2 to account for the
number of seats currently sold and a demand of one additional seat for the Pittsburgh
Myrtle Beach Q class ODIF. Resolve the linear programming model to determine a
new allocation schedule for the ODIFs.
7. Hanson Inn is a 96-room hotel located near the airport and convention center in Louisville,
Kentucky. When a convention or a special event is in town, Hanson increases its normal
room rates and takes reservations based on a revenue management system. The Classic
Corvette Owners Association scheduled its annual convention in Louisville for the rst
weekend in June. Hanson Inn agreed to make at least 50% of its rooms available for con-
vention attendees at a special convention rate in order to be listed as a recommended hotel
for the convention. Although the majority of attendees at the annual meeting typically re-
quest a Friday and Saturday two-night package, some attendees may select a Friday night
only or a Saturday night only reservation. Customers not attending the convention may
also request a Friday and Saturday two-night package, or make a Friday night only or Sat-
urday night only reservation. Thus, six types of reservations are possible: Convention
customers/two-night package; convention customers/Friday night only; convention cus-
tomers/Saturday night only; regular customers/two-night package; regular customers/
Friday night only; and regular customers/Saturday night only. The cost for each type of
reservation is shown here:
Problems 251
Two-Night Friday Night Saturday Night
Package Only Only
Convention $225 $123 $130
Regular $295 $146 $152
Two-Night Friday Night Saturday Night
Package Only Only
Convention 40 20 15
Regular 20 30 25
The anticipated demand for each type of reservation is as follows:
Hanson Inn would like to determine how many rooms to make available for each type of
reservation in order to maximize total revenue.
a. Dene the decision variables and state the objective function.
b. Formulate a linear programming model for this revenue management application.
c. What are the optimal allocation and the anticipated total revenue?
d. Suppose that one week before the convention, the number of regular customers/
Saturday night only rooms that were made available sell out. If another nonconvention
customer calls and requests a Saturday night only room, what is the value of accept-
ing this additional reservation?