

Probability: Part I
Grayson White
Math 141
Week 6 | Fall 2025
Probability Theory: the study and quantification of uncertainty/randomness in outcomes of repeated experiments.
A random process is one which we know what results could happen, but don’t know which particular result will happen.
An event is a potential result of some particular random process.
We will soon define the word “probability.” Before that, consider the following questions in small groups:
Neither definition is “correct”. Statisticians are divided on how to interpret probability.
The “frequentist” (most common) definition of probability is…
Definition: Probability
The probability of a particular event is the proportion of times the event would occur, if we observed the random process an arbitrarily large number of times.
To say that a coin has 50% probability of landing heads, means that…
Since probabilities are defined as a proportion, they will always be values between 0 and 1.
For brevity, we’ll represent statements like the probability of the event “the coin lands heads” is 50% using the notation: \[ P(\textrm{Heads})=0.5 \quad \textrm{or} \quad P(H)=0.5 \]
The probability of an event refers to the long-run tendency of the proportion.

The proportion of heads deviates wildly from 0.5 during the first 200 flips
The probability of an event refers to the long-run tendency of the proportion.

The proportion of heads gets closer to 0.5 after 1000 flips
The probability of an event refers to the long-run tendency of the proportion.

The proportion of heads is very close to 0.5 after 2000 flips!
The probability of an event refers to the long-run tendency of the proportion.

This is the Law of Large Numbers in effect!
Suppose you’re interested in winning the Powerball Jackpot, where the chance of winning is 1 in 292 million. You buy a single lottery ticket and wait to see if you won.
When playing the lottery (a random process), what are the two possible outcomes (events)?
Write down the probability of your two events.
If you played the lottery 1 million times, should you expect to win? What about 292 million times?
If you played the lottery 1 billion times, are you guaranteed to win?
We can formalize the connection between random processes/events with probabilities using a Probability Model
A probability model has two components:
Example: Probability Model for a Coin Toss:
When discussing probability, we always (explicitly or implicitly) define a probability model.
Two events \(A\) and \(B\) are said to be mutually exclusive (or disjoint) if it is not possible for both to occur at the same time.
Example: Single Roll of a Die
Theorem: Addition Rule
The probability that at least one event occurs in a pair of disjoint events is the sum of their individual probabilities: \[ P(A \textrm{ or } B) = P(A) + P(B) \]
Q: When a die is rolled, what is the probability that an odd number is rolled?
\[ \begin{aligned} P( \textrm{Odd} ) &= P(\textrm{roll a 1, 3, or 5})\\ &= P(\textrm{roll a 1}) + P(\textrm{roll a 3}) + P(\textrm{roll a 5}) && \text{(by the Addition Rule)}\\ &= \frac{1}{6} + \frac{1}{6} + \frac{1}{6} \\ &= \frac{3}{6} \end{aligned} \]
Theorem: Addition Rule
The probability that at least one event occurs in a pair of disjoint events is the sum of their individual probabilities: \[ P(A \textrm{ or } B) = P(A) + P(B) \]
Q: In a die-roll, what’s the probability we get an even number?
Q: In a die-roll, what’s the probability we get at least a 3?
Q: What’s the probability you get an even number OR at least a 3?
The complement to an event \(A\) (denoted \(A^c\)) is the event that occurs exactly when the original does not.
Theorem: Complement Rule
The probability that the complement of an event occurs is 1 minus the probability of the event: \[ P(A^c) = 1 - P(A) \]
What is the probability that any number other than a 1 is rolled on a fair 6-sided die?
\[ P(\textrm{roll something other than a 1}) = 1 - P( \textrm{roll a 1}) = 1 - \frac{1}{6} = \frac{5}{6} \]
Theorem: Addition Rule
The probability that at least one event occurs in a pair of disjoint events is the sum of their individual probabilities: \[ P(A \textrm{ or } B) = P(A) + P(B) \]
Theorem: Complement Rule
The probability that the complement of an event occurs is 1 minus the probability of the event: \[ P(A^c) = 1 - P(A) \]
In the city of Portland, there’s rain on 60% of days and snow on 1% of days.
On a given day, what’s the probability that it doesn’t rain?
On a given day, what’s the probability that it rains OR doesn’t rain?
What if I told you it rains OR snows 61% of days in Portland. What’s the flaw in my reasoning?
In the city of Portland, there’s rain on 60% of days and snow on 1% of days.
\[P(\text{Doesn't Rain}) = P(\text{Rain}^c) = 1-P(\text{Rain}) = 1-0.6=0.4\]
\[P(\text{Rain or Doesn't Rain}) = P(\text{Rain}) + P(\text{Doesn't Rain}) = 0.6+0.4=1\]
It can rain and snow in one day! Thus, we can’t use the addition rule!
Today we defined: