Solutions to Tutorial on Risk and Uncertainty

1. Suppose that a decision maker having four decision alternatives and four states of nature has developed the
following payoff table.

State of Nature
Decision    
Alternative

a1 15 13 12 9
a2 12 11 9 10
a3 9 10 7 12
a4 7 10 8 12

You may assume that the decision maker knows nothing concerning the probability of occurrence of the various
states of nature.

(a) Using the maximin decision criteria, which decision alternative would be selected?
(b) Using the maximax decision criteria, which decision alternative would be selected?
(c) Using the minimax regret decision procedure, which decision alternative would be selected?
(d) Assume the following estimates of probability of occurrence of the states of nature can be made. Which
decision would be selected based on the EMV criteria?

P ( P P P

Solution

The minimums for each of the decision alternatives are a1 = 9, a2 = 9, a3 = 7 and a4 = 7. Therefore the decision-
maker will narrow the choice down to a1 and a2. To decide between these alternatives the next lowest case must
be looked at. For a1 this is 12. For a2 this is 10. Therefore the decision-maker will choose a1.
The maximums for each of the decision alternatives are a1 = 15, a2 = 12, a3 = 12 and a4 = 12. The maximum of
these is 15 for a1. Therefore choose a1.
The opportunity loss matrix is as follows

State of Nature
Decision    
Alternative

a1 0 0 0 3
a2 3 2 3 2
a3 6 3 5 0
a4 8 3 4 0
And the maximum regret for each alternative is a1 = 3, a2 = 3, a3 = 6 and a4 = 8. Therefore the minimum is 3 and
so the decision-maker will narrow the choice down to a1 and a2. To decide between these alternatives the next
lowest case must be looked at. For a1 this is 0. For a2 this is 3. Therefore the decision-maker will choose a1.
EMV1 = 0.3 * 15 + 0.2 * 13 + 0.2 * 12 + 0.3 * 9 = 12.2
EMV1 = 0.3 * 12 + 0.2 * 11 + 0.2 * 9 + 0.3 * 10 = 10.6
EMV1 = 0.3 * 9 + 0.2 * 10 + 0.2 * 7 + 0.3 * 12 = 9.7
EMV1 = 0.3 * 7 + 0.2 * 10 + 0.2 * 8 + 0.3 * 12 = 9.3
Therefore the highest EMV is for alternative a1.

Engineering Economics Risk and Uncertainty – Tutorial Solution Page 1 of 7

2. You have decided to dispose of your second-hand car. Unfortunately it has failed its inspection and will
require $100 of repair to obtain a pink slip. A wrecker has offered you $50 for the car as it is. You have little
idea what the market is like for your type of car; if the market is strong with a pink slip you would get $250 if
moderate $150 and if weak $80. Without a pink slip you could sell the car for $120, $80 and $50 in strong,
moderate and weak markets, respectively. If you decide to advertise your car for sale you incur a $10
advertising cost.

(a) Identify the three possible strategies and construct the payoff matrix.
(b) What decision would you adopt if,
(i) you are a pessimist
(ii) you are an optimist
(iii) if you wish to minimize your regret
(c) Suppose the probability of a strong, moderate and weak market are given by 0.3, 0.3 and 0.4
respectively. Which decision maximizes your expected value.

Solution

The three possible strategies are to sell to the wrecker, to get pink slip and sell privately, and to sell privately
with no pink slip. Therefore the payoff matrix is given by:
States of nature
Alternatives Strong Moderate Weak
Wrecker $50 $50 $50
Pink slip $250 - $110 = $140 $150 - $110 = $40 $80 - $110 = -$30
No pink slip $120 - $10 = $110 $80 - $10 = $70 $50 - $10 = $40

Use Maximin, choose between $50, -$30 and $40. Choose $50, ie sell to a wrecker.
Use Maxaimax, look for best result = $140. Therefore get a pink slip and sell privately
Construct table
States of nature
Alternatives Strong Moderate Weak
Wrecker $90 $20 $0
Pink slip $0 $30 $80
No pink slip $30 $0 $10
Choose minimum of $90, $80 and $30, ie sell the car privately without a pink slip
EMV (wrecker) = $50, EMV (pink slip) = 0.3 * $140 + 0.3 * $40 + 0.4 * -$30 = $42
EMV (no pink slip) = 0.3 * $110 + 0.3 * $70 + 0.4 * $40 = $70. Therefore sell privately with no pink slip.

3. From Dandy and Warner, 'Planning and Design of Engineering Systems'.

A construction company must decide on the best plan to follow for the installation of a pipeline. Plan A is a
minimum-time approach which will require 3 months for the installation at a cost of $200,000. Plan B is to
follow normal construction procedures; the time required for installation is 4 months and the cost is $160,000.
Plan C involves a slower approach to installation; the time required is 5 months and the cost is $150,000.

The installation must be undertaken during winter and additional costs will be incurred if bad weather occurs.
Essentially, only three weather conditions need to be distinguished: a dry winter; an average winter; and a wet
winter. Additional construction costs for each plan and set of weather conditions are given in Table 9.8.

Assuming that the probability of occurrence of each weather type cannot be estimated from existing data, which
plan should be chosen under each of the following criteria:

(a) pessimistic
(b) optimistic
(c) minimum regret
(d) Laplace

What do you think is the best construction plan and why?

Engineering Economics Risk and Uncertainty – Tutorial Solution Page 2 of 7

Additional Construction Costs Construction Plan
A B C
Dry winter 0 0 0
Average winter $15 000 $60 000 $90 000
Wet winter $20 000 $90 000 $120 000

Solution

Firstly we must construct the payoff matrix

Construction Plan Dry winter Average winter Wet winter

A $200 000 $215 000 $220 000
B $160 000 $220 000 $250 000
C $150 000 $240 000 $270 000

Pessimistic - use maximin. Find minumum (since we are looking at costs not benefits) of $220,000, $250,000
and $270,000. Therefore choose option A
Optimistic - use maximax. Best result is $150 000 for option C.
Construct the opportunity loss table:
Construction Plan Dry winter Average winter Wet winter
A $50 000 $0 $0
B $10 000 $5 000 $30 000
C $0 $25 000 $50 000
Choose the minimum of $50 000, $30 000 and $50 000, ie choose $30 000 - Plan B.
This assumes that each situation (dry, average and wet winters has an equal chance)
Laplace (A) = ($200 000 + $215 000 + $220 000) / 3 = $211 667
Laplace (B) = ($160 000 + $220 000 + $250 000) / 3 = $210 000
Laplace (C) = ($150 000 + $240 000 + $270 000) / 3 = $220 000
Therefore choose construction plan B.
Construction Plan A is probably the overall best plan. Firstly it is the maximin result, so there is a lower cap on
any loss that could be made. Secondly it is absolutely better in both the average winter and wet winter
categories. If there is a dry winter then the company still spends less than it would in worse weather conditions
for any project.

4. Newcastle Knights Rugby League Club orders programmes for its home games in batches of 10,000, 20,000
or 30,000 at costs of $1,000, $1,600 or $2,000 respectively. Programmes sell at 20¢ each. The order for each
match has to be placed a week in advance and any programme not sold on the day has no value. Sales of
programmes depend on the match attendance. For simplicity assume that the programme sales will be 10,000,
20,000 or 30,000. Find the maximin and minimax regret strategies for programme orders.

Next Saturday, Newcastle will be playing their arch rival Canberra. The general manager of Newcastle
estimates the probability of 30,000 sales = 0.6, 20,000 = 0.3 and 0.1 of 10,000.

What is the optimum EMV strategy under these conditions?

Solution
Firstly construct the payoff table
Attendance
Programmes printed 10 000 20 000 30 000
10 000 $2000 - $1000 = $1000 $1000
$1000
20 000 $2000 - $1600 = $400 $4000 - $1600 = $2400
$2400
30 000 $2000 - $2000 - $0 $4000 - $2000 = $6000 - $2000 =
$2000 $4000

The maximin strategy requires the maximum of $1000, $400 and $0, which is $1000. Therefore choose to print
10 000 programmes.
Engineering Economics Risk and Uncertainty – Tutorial Solution Page 3 of 7
The minimax regret strategy requires an opportunity loss table:
Attendance
Programmes printed 10 000 20 000 30 000
10 000 $0 $1400 $3000
20 000 $600 $0 $1600
30 000 $1000 $400 $0
We then choose the minimum of $3000, $1600, and $1000. Therefore we would choose to print 30 000
programmes.

EMV (10 000) = $1000

EMV (20 000) = 0.1 * $400 + 0.9 * $2400 = $2200
EMV (30 000) = 0.1 * $0 + 0.3 * $2000 + 0.6 * $4000 = $3000
Therefore choose to print 30 000 programmes.

5. The Government is planning a major motorway development in an area where there is land available to build
a hotel/restaurant. Local protesters are raising objections to the Government's plans and it is by no means certain
that they will go ahead. If the development is confirmed, the hotel will yield a profit equivalent to a present
value of $20 million. On the other hand if the land is bought but the development does not go ahead the present
value of the resultant loss is $5 million. There is an alternative site for this investment which will yield a return
of $7 million irrespective of the Government's decision on the motorway development.

What is the smallest value of the probability that the development will go ahead which will cause you to buy the
land? How sensitive is your answer to the assumed return of $20 million?

Answer
This problem requires the use of the EMV criterion. Firstly assume that the probability of the motorway going
ahead is p. Then the probability of the motorway not going ahead is 1-p. Now set up an equation where the
EMV of building a hotel is equal to the EMV of the alternative site:

p * $20M - (1-p) * $5M = $7M

p * $25M = $12M
Therefore p = 48%

To investigate the sensitivity a spreadsheet was used to create the graph below with various values of the
assumed return. Each change in return of $1 million gave slightly under a 2 percent absolute change, which is
about a 4 percent relative change.
1.2
Minimum accptable probability of

1
building motorway

0.8

0.6

0.4

0.2

0
0 5E+06 1E+07 2E+07 2E+07 3E+07 3E+07 4E+07 4E+07
Assumed return

6. A construction company needs to erect a temporary access bridge to a construction site in area of known
seismic activity. They must decide what degree of earthquake resistance should be incorporated in the design to
minimize total costs associated with seismic design and damages. The following information is available.

Engineering Economics Risk and Uncertainty – Tutorial Solution Page 4 of 7

(a) 50 years of earthquake records. 40 earthquakes have occurred.
(b) The cost of making the bridge earthquake resistant.

Earthquake No. in past 50 years Cost of Resistant

Magnitude Design
0-4 24 $1000
4-7 12 $3000
>7 4 $6000

(c) the cost of damage of an earthquake occurs,

Earthquake Magnitude Earthquake Cost of

Designed for Magnitude Occurring Damage ($)
0-4 3000
None 4-7 5000
>7 8000
0-4 0
0-4 4-7 3000
>7 7000
4-7 0-7 0
>7 4000
>7 any 0

Answer

(i) There is only one level of decision that is what level of earthquake to design for. We
can start from one decision node with four possible branches.

(ii) At the end of each branch there is a chance node with two outcomes: an earthquake
probability 0.8, no earthquake probability 0.2.

If an earthquake occurs there are three possible ranges of magnitudes with the following probabilities
p(0-4) = 0.6, p(4-7) = 0.3, p(>7) = 0.1.

The decision tree is shown below.

The EMV's of each node can be calculated as follows:

EMV(F) = 0.6 (-3000) + 0.3 x (-5000) + 0.1 *(-8000) = -$4100
EMV(G) = -$2600
EMV(H) = -$3400
EMV(I) = -$6000

Now looking at the second chance nodes,

EMV(B) = 0.8 x (-4100) + 0.2 x 0 = -$3280
EMV(C) = -$2280
EMV(D) = -$3320
EMV(E) = -$6000

This would indicate that the bridge should be designed to resist an earthquake of magnitude 0-4.

Engineering Economics Risk and Uncertainty – Tutorial Solution Page 5 of 7

 0 - 4 (0.60)
0 - 3000 = -3000

Quake (0.8) 4 - 7 (0.30)

F 0 - 5000 = -5000

> 7 (0.10)
0 - 8000 = -8000
B

No Quake (0.2) 0
0 - 4 (0.60) -1000 - 0 = -1000

No earthquake design Quake (0.8) 4 - 7 (0.30) -1000 - 3000 = -4000

G
> 7 (0.10) -1000 - 7000 = -8000
C
No Quake (0.2) -1000
Design for 0-4 0 - 4 (0.60) -3000 - 0 = -3000
A
Quake (0.8) 4 - 7 (0.30) -3000 - 0 = -3000
Design for 4-7 H

> 7 (0.10) -3000 - 4000 = -7000

D
Design for >7 -3000
No Quake (0.2)
0 - 4 (0.60) -6000 - 0 = -6000

Quake (0.8) 4 - 7 (0.30) -6000 - 0 = -6000

I

> 7 (0.10) -6000 - 0 = -6000

E

No Quake (0.2)
-6000

7. (after Meredith et al., 1973) Consider the problem encountered by a contractor who is performing some
work with special equipment in a river flats area that has been subjected in the past to high-water conditions and
occasional destructive flooding. There is a 4-month period during which there is no use for the equipment either
on this job or on others. The contractor can keep the equipment on the job in the river flats, or else move it out,
store it, and then move it back, at a total cost of $1800.

If the contractor keeps the equipment in the river flats area, there is the option of building a platform for the
equipment at a cost of $500, which will protect it against high water, but not against a destructive flood. The
damage that would be caused by high water amounts to $10,000 if there is no platform. A destructive flood
would entail a loss of $60,000, regardless of whether or not a platform is built. The probability of high water in
the 4-month period is 0.25; the probability of a destructive flood is 0.02.

Determine the alternative courses of action the contractor can take, draw up a decision tree, determine the EMV
of each action. Which action should be chosen?

Engineering Economics Risk and Uncertainty – Tutorial Solution Page 6 of 7

Solution
EMV = -$1800 (1.00) -$1 800

Move the
equipment
EMV = Flood (0.02) -$60 500
-$1 700

Build a -$500
No flood (0.98)
platform

Do nothing Flood (0.02) -$60 000

High Water
EMV =-$3 700 (0.25)
-$10 000

No problem (0.73) $0

Therefore the engineer should build the platform.

Engineering Economics Risk and Uncertainty – Tutorial Solution Page 7 of 7