Probability space

A probability space is a way that mathematicians study chance and random events. It gives us rules to think about what might happen when we do an experiment that has many possible results. This idea was created by a Soviet mathematician named Andrey Kolmogorov in the 1930s.

Imagine you roll a dice. The possible results are the numbers 1 through 6. This set of results is called the sample space. An event could be something simple, like rolling a 5, or something more complex, like rolling an even number.

We also need a way to measure how likely each event is. This is done with a probability function, which gives each event a number between 0 and 1. For example, the chance of rolling a 5 is 1 out of 6, or about 0.17. The chance of rolling an even number (2, 4, or 6) is 3 out of 6, or 0.5. These tools together — the sample space, the events, and the probabilities — make up a probability space. It helps us understand and predict the world of chance and randomness.

Introduction

A probability space is a way to study chance and uncertainty using math. It has three main parts:

The sample space is all the possible results of an experiment. For example, when flipping a coin, the sample space has two outcomes: Head or Tail.

The event space is a group of special collections of these results. For example, one event might be "getting a Head," while another could be "getting either Head or Tail."

The probability measure tells us how likely each event is. Probabilities are numbers between 0 (something that can't happen) and 1 (something that will almost certainly happen). For instance, the chance of flipping a Head is 0.5, because it's equally likely to get a Head or a Tail.

Definition

A probability space is a special way to study chance and randomness. It has three main parts:

The sample space is all the possible results of an experiment. For example, when flipping a coin, the sample space is heads or tails.

The event space is a collection of these possible results we might be interested in. For example, we might care about getting heads when flipping a coin.

The probability measure tells us how likely each event is. For a fair coin, the chance of heads is 1 out of 2, or 0.5.

Discrete case

When we talk about chances or probabilities for simple, countable situations, we use something called a discrete probability distribution. This helps us assign probabilities — chances — to each possible outcome. These probabilities must add up to 1, meaning they cover all possible results together.

In this setup, every possible group of outcomes can be treated as an event. The probabilities are set up so that they give a clear picture of all the information we can have about the outcomes. Sometimes, certain outcomes might have a probability of zero, meaning they almost never happen, and we can ignore them in practice.

P ( A ) = ∑ ω ∈ A p ( ω ) for all A ⊆ Ω . {\displaystyle P(A)=\sum _{\omega \in A}p(\omega )\quad {\text{for all }}A\subseteq \Omega .}

General case

When the set of all possible outcomes, called Ω, is very large and cannot be counted, some outcomes might still have a chance of happening. These special outcomes are called atoms. There can only be a limited number of these atoms. If the total chance of all atoms adds up to 1, we can ignore all other outcomes. If the total chance of atoms is less than 1, the probability space splits into two parts: one with countable outcomes and one without.

Non-atomic case

When each possible outcome has zero probability, there must be an uncountable number of outcomes. In this case, the usual way of adding up probabilities doesn't work because it only applies to countable numbers of items. Instead, a more advanced math tool called measure theory is used. We start by assigning probabilities to certain basic groups of outcomes, then use a special process to assign probabilities to more complex groups. These groups are part of something called a σ-algebra, which includes many more complicated sets than the basic ones but avoids very tricky sets that are hard to work with. For more details, see Carathéodory's extension theorem.

Complete probability space

A probability space ( Ω , F , P ) is called a complete probability space if it includes every possible small group of outcomes. This means that if there is a group of outcomes with zero chance of happening, then every smaller group inside it is also included in the study. Usually, experts focus only on these complete probability spaces to make their work easier.

Examples

Discrete examples

Example 1

Imagine flipping a fair coin once. The result can either be heads or tails. We call this our list of possible outcomes.

Example 2

Now let’s flip the coin three times. There are eight possible results, like HHH, HHT, and so on. Each result is an event we can think about.

Example 3

Suppose we randomly pick 100 voters from California to ask who they will vote for. The list of all possible groups of 100 voters is our sample space.

Non-atomic examples

Example 4

Imagine picking a number between 0 and 1 at random. The set of all possible numbers is our sample space.

Example 5

Think of flipping a coin forever, recording each result. The list of all possible endless sequences of heads and tails is our sample space. Each short part of this endless list can be an event we study.

Related concepts

Probability distribution

Main article: Probability distribution

Random variables

Main article: Random variable

A random variable is a way to connect outcomes of a random process to numbers or other values. It helps us understand and work with the results of random events.

Defining the events in terms of the sample space

When we have a small list of possible outcomes, we can easily describe all the events we might be interested in. But when there are too many outcomes to list, we need to use special tools to keep things organized and understandable.

Conditional probability

Main article: Conditional probability

Conditional probability helps us understand the chance of something happening, given that we already know some other information. It tells us how probabilities change with new knowledge.

Independence

Main article: Statistical independence

Two events are independent if knowing that one happened doesn’t change the chance that the other happened. This idea helps us understand when different parts of a random process don’t affect each other.

Mutual exclusivity

Main article: Mutual exclusivity

Two events are mutually exclusive if they cannot happen at the same time. When events are mutually exclusive, we can simply add their probabilities together to find the chance that either one happens.