Basic Ideas

To make a systematic study of probability, we need some terminology. An experiment is a process that
results in an outcome that cannot be predicted in advance with certainty. The following are all examples of
experiments.
 tossing a coin
 rolling a die
 measuring the diameter of a bolt
 weighing the contents of a box of cereal
 measuring the breaking strength of a length of fishing line
To discuss an experiment in probabilistic terms, we must specify its possible outcomes:

The set of all possible outcomes of an experiment is called the

sample space for the experiment.

A subset of a sample space is called an event.

Example 1:
An electrical engineer has on hand two boxes of resistors, with four resistors in each box. The resistors in the
first box are labeled 10 (ohms), but in fact their resistances are 9, 10, 11, and 12. The resistors in the second
box are labeled 20, but in fact their resistances are 18, 19, 20, and 21. The engineer chooses one resistor
from each box and determines the resistance of each.

Let A be the event that the first resistor has a resistance greater than 10, let B be the event that the second
resistor has a resistance less than 19, and let C be the event that the sum of the resistances is equal to 28.
Find a sample space for this experiment, and specify the subsets corresponding to the events A, B, and C.

Solution:
A good sample space for this experiment is the set of ordered pairs in which the
first component is the resistance of the first resistor and the second component is the
resistance of the second resistor. We will denote this sample space by S.
S = {(9, 18), (9,19), (9, 20), (9,21), (10,18), (10,19), (10, 20), (10,21), (11, 18), (11, 19), (11, 20), (11,
21), (12,18), (12, 19), (12, 20), (12, 21)}

The events A, B, and C are given by

A = {(11, 18), (11, 19), (11, 20), (11, 21), (12, 18), (12, 19), (12, 20), (12, 21)}
B = {(9, 18), (10, 18), (11, 18), (12, 18)}
C = {(9, 19), (10, 18)}

Combining Events
We often construct events by combining simpler events. Because events are subsets of sample spaces, it is
traditional to use the notation of sets to describe events constructed in this way. We review the necessary
notation here.

 The union of two events A and B, denoted A ∪ B, is the set of outcomes that belong either to A, to B,
or to both. In words, A ∪ B means “A or B.” Thus, the event A ∪ B occurs whenever either A or B (or
both) occurs.
  The intersection of two events A and B, denoted A ∩ B, is the set of outcomes that belong both to A
and to B. In words, A ∩ B means “A and B.” Thus, the event A ∩ B occurs whenever both A and B
occur.
 The complement of an event A, denoted Ac, is the set of outcomes that do not belong to A. In words,
Ac means “not A.” Thus, the event Ac occurs whenever A does not occur.

Events can be graphically illustrated with Venn diagrams. Figure 2.1 illustrates the
events A ∪ B, A ∩ B, and B ∩ Ac.

Example 2:
Refer to Example 1 above. Find B ∪ C and A ∩ Bc.

Solution:
The event B ∪ C contains all the outcomes that belong either to B or to C, or to both.
Therefore,
B ∪ C = {(9, 18), (10, 18), (11, 18), (12,18), (9,19)}

The event Bc contains those outcomes in the sample space that do not belong to B. It follows that the event
A ∩ Bc contains the outcomes that belong to A and do not belong to B. Therefore,
A ∩ Bc = {(11, 19), (11,20), (11,21), (12, 19), (12, 20), (12,21)}

Mutually Exclusive Events

There are some events that can never occur together. For example, it is impossible that a coin can come up
both heads and tails, and it is impossible that a steel pin can be both too long and too short. Events like this
are said to be mutually exclusive.

 The events A and B are said to be mutually exclusive if they have no outcomes in common.
 More generally, a collection of events A1, A2, . . ., An is said to be mutually exclusive if no
two of them have any outcomes in common.
Example 3:
Refer to Example 1. If the experiment is performed, is it possible for events A and B both to occur? How
about B and C? A and C? Which pair of events is mutually exclusive?

Solution
If the outcome is (11, 18) or (12, 18), then events A and B both occur. If the outcome is (10, 18), then both B
and C occur. It is impossible for A and C both to occur, because these events are mutually exclusive, having
no outcomes in common.

Probabilities
Each event in a sample space has a probability of occurring. Intuitively, the probability is a quantitative
measure of how likely the event is to occur. Formally speaking, there are several interpretations of
probability; the one we shall adopt is that the probability of an event is the proportion of times the event
would occur in the long run, if the experiment were to be repeated over and over again.

We often use the letter P to stand for probability. Thus, when tossing a coin, the notation “P(heads) = 1/2”
means that the probability that the coin lands heads is equal to 1/2.

Given any experiment and any event A:

 The expression P(A) denotes the probability that the event A occurs.
 P(A) is the proportion of times that event A would occur in the long run, if the
experiment were to be repeated over and over again.

In practice, scientists and engineers estimate the probabilities of some events on the basis of scientific
understanding and experience and then use mathematical rules to compute estimates of the probabilities
of other events. In the rest of this unit, I will explain some of these rules and show how to use them.

Axioms of Probability
The subject of probability is based on three commonsense rules, known as axioms. They are:

The Axioms of Probability

1. Let S be a sample space. Then P(S) = 1.
2. For any event A, 0 ≤ P(A) ≤ 1.
3. If A and B are mutually exclusive events, then P(A ∪ B) = P(A) + P(B).
More generally, if A1, A2, . . . are mutually exclusive events, then P(A1
∪ A2 ∪ · · ·) = P(A1) + P(A2) + · · · .

For any event A,

P(Ac) = 1 - P(A) (2.1)
Let ∅ denote the empty set.
Then,
P(∅) = 0 (2.2)
Example 4:
A target on a test firing range consists of a bull’s-eye with two concentric rings around it. A projectile is fired
at the target. The probability that it hits the bull’s-eye is 0.10, the probability that it hits the inner ring is 0.25,
and the probability that it hits the outer ring is 0.45. What is the probability that the projectile hits the target?
What is the probability that it misses the target?

Solution
Hitting the bull’s-eye, hitting the inner ring, and hitting the outer ring are mutually exclusive events, since it
is impossible for more than one of these events to occur. Therefore, using Axiom 3,

P(hits target) = P(bull’s-eye) + P(inner ring) + P(outer ring)

= 0.10 + 0.25 + 0.45
= 0.80

We can now compute the probability that the projectile misses the target by using Equation (2.1):
P(misses target) = 1 - P(hits target)
= 1 - 0.80
= 0.20

If A is an event containing outcomes O1, . . ., On, that is, if A = {O1, . . ., On},

then
P(A) = P(O1) + P(O2) + · · · + P(On) (2.3)

Example 5:
The following table presents probabilities for the number of times that a certain computer system will crash
in the course of a week. Let A be the event that there are more than two crashes during the week, and let B
be the event that the system crashes at least once. Find a sample space. Then find the subsets of the sample
space that correspond to the events A and B. Then find P(A) and P(B).

Number of Crashes Probability

0 0.60
1 0.30
2 0.05
3 0.04
4 0.01

Solution
A sample space for the experiment is the set {0, 1, 2, 3, 4}. The events are A = {3, 4} and B = {1, 2, 3, 4}. To
find P(A), notice that A is the event that either 3 crashes happen or 4 crashes happen. The events “3 crashes
happen” and “4 crashes happen” are mutually exclusive. Therefore, using Axiom 3, we conclude that
P(A) = P(3 crashes happen or 4 crashes happen)
= P(3 crashes happen) + P(4 crashes happen)
= 0.04 + 0.01
= 0.05
We will compute P(B) in two ways. First, note that Bc is the event that no crashes happen. Therefore, using
Equation (2.1),
P(B) = 1 - P(Bc)
= 1 - P(0 crashes happen)
= 1 - 0.60
= 0.40
For a second way to compute P(B), note that B is the event that 1 crash happens or 2 crashes happen or 3
crashes happen or 4 crashes happen. These events are mutually exclusive. Therefore, using Axiom 3, we
conclude that
P(B) = P(1 crash) + P(2 crashes) + P(3 crashes) + P(4 crashes)
= 0.30 + 0.05 + 0.04 + 0.01
= 0.40

Sample Spaces with Equally Likely Outcomes

For some experiments, a sample space can be constructed in which all the outcomes are equally likely. A
simple example is the roll of a fair die, in which the sample space is {1, 2, 3, 4, 5, 6} and each of these
outcomes has probability 1/6.

Another type of experiment that results in equally likely outcomes is the random selection of an item from
a population of items. The items in the population can be thought of as the outcomes in a sample space, and
each item is equally likely to be selected.

A population from which an item is sampled at random

can be thought of as a sample space with equally likely
outcomes.

If S is a sample space containing N equally likely outcomes, and if A

is an event containing k outcomes, then
𝒌𝒌
𝑷𝑷(𝑨𝑨) = (2.4)
𝑵𝑵

Example 6:
An extrusion die is used to produce aluminum rods. Specifications are given for the length and the diameter
of the rods. For each rod, the length is classified as too short, too long, or OK, and the diameter is classified
as too thin, too thick, or OK. In a population of 1000 rods, the number of rods in each class is as follows:

Diameter
Length Too Thin OK Too Thick
Too Short 10 3 5
OK 38 900 4
Too Long 2 25 13

A rod is sampled at random from this population. What is the probability that it is too short?

Solution
We can think of each of the 1000 rods as an outcome in a sample space. Each of the 1000 outcomes is equally
likely. We’ll solve the problem by counting the number of outcomes that correspond to the event. The
number of rods that are too short is 10 + 3 + 5 = 18. Since the total number of rods is 1000,

18
𝑃𝑃 (𝑡𝑡𝑡𝑡𝑡𝑡 𝑠𝑠ℎ𝑜𝑜𝑜𝑜𝑜𝑜) =
1000
The Addition Rule
If A and B are mutually exclusive events, then P(A∪ B) = P(A) + P(B). This rule can be generalized to cover the
case where A and B are not mutually exclusive. Example 7 illustrates the reasoning.

Example 7:
Refer to Example 6. If a rod is sampled at random, what is the probability that it is either too short or too
thick?

Solution:
First we’ll solve this problem by counting the number of outcomes that correspond to the event. In the
following table the numbers of rods that are too thick are circled, and the numbers of rods that are too short
have rectangles around them. Note that there are 5 rods that are both too short and too thick.

Of the 1000 outcomes, the number that are either too short or too thick is 10 + 3 + 5 + 4 + 13 = 35. Therefore
35
𝑃𝑃(𝑡𝑡𝑡𝑡𝑡𝑡 𝑠𝑠ℎ𝑜𝑜𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜 𝑡𝑡𝑡𝑡𝑡𝑡 𝑡𝑡ℎ𝑖𝑖𝑖𝑖𝑖𝑖) =
1000

Now we will solve the problem in a way that leads to a more general method. In the sample space, there are
10 + 3 + 5 = 18 rods that are too short and 5 + 4 + 13 = 22 rods that are too thick. But if we try to find the
number of rods that are either too short or too thick by adding 18 + 22, we get too large a number (40 instead
of 35).

The reason is that there are five rods that are both too short and too thick, and these are counted twice. We
can still solve the problem by adding 18 and 22, but we must then subtract 5 to correct for the double counting.

We restate this reasoning, using probabilities:

18 22 5
𝑃𝑃(𝑡𝑡𝑡𝑡𝑡𝑡 𝑠𝑠ℎ𝑜𝑜𝑜𝑜𝑜𝑜) = , 𝑃𝑃(𝑡𝑡𝑡𝑡𝑡𝑡 𝑡𝑡ℎ𝑖𝑖𝑖𝑖𝑖𝑖) = , 𝑃𝑃(𝑡𝑡𝑡𝑡𝑡𝑡 𝑠𝑠ℎ𝑜𝑜𝑜𝑜𝑜𝑜 𝑎𝑎𝑎𝑎𝑎𝑎 𝑡𝑡𝑡𝑡𝑡𝑡 𝑡𝑡ℎ𝑖𝑖𝑖𝑖𝑖𝑖) =
1000 1000 1000

P(too short or too thick) = P(too short) + P(too thick) - P(too short and too thick)
18 22 5
𝑃𝑃(𝑡𝑡𝑡𝑡𝑡𝑡 𝑠𝑠ℎ𝑜𝑜𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜 𝑡𝑡𝑡𝑡𝑡𝑡 𝑡𝑡ℎ𝑖𝑖𝑖𝑖𝑖𝑖) = + −
1000 1000 1000
35
𝑃𝑃(𝑡𝑡𝑡𝑡𝑡𝑡 𝑠𝑠ℎ𝑜𝑜𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜 𝑡𝑡𝑡𝑡𝑡𝑡 𝑡𝑡ℎ𝑖𝑖𝑖𝑖𝑖𝑖) =
1000

Let A and B be any events. Then

P(A U B) = P(A) + P(B) - P(A ∩ B) (2.5)

Example 8:
In a process that manufactures aluminum cans, the probability that a can has a flaw on its side is 0.02, the
probability that a can has a flaw on the top is 0.03, and the probability that a can has a flaw on both the side
and the top is 0.01. What is the probability that a can has a flaw on the top but not on the side?

Solution:
Let A be the event that a can has a flaw on the top and let B be the event that a can has a flaw on the side.
We need to find P(A ∩ Bc).

The following Venn diagram (Figure 2.3) shows that A ∩ B and A ∩ Bc are mutually exclusive, so that
P(A) = P(A ∩ B) + P(A ∩ Bc)

We know that P(A) = 0.03 and P(A ∩ B) = 0.01. Therefore 0.03 = 0.01 + P(A ∩ Bc), so
P(A ∩ Bc) = 0.02.