3.2.1. Overview

The Axiomatic Definition of probability was introduced by Andrey Kolmogorov in 1933. It provides a mathematically rigorous foundation for probability and can handle infinite sample spaces and non-uniform probabilities.

In 1933, mathematician Andrey Kolmogorov developed the Axiomatic Definition of probability to create a more formal and structured approach to the concept of probability. Unlike earlier methods, this definition does not just rely on simple equal outcomes but is designed to accommodate complex situations, including those with infinite possibilities. This definition lays out a strong mathematical framework that can deal with different types of probabilities, including non-uniform ones.

Think of the Axiomatic Definition of Probability as the rules of a game that can vary drastically based on how many players you have and the rules they follow. In some cases, all players might have equal chances of winning, much like a simple board game. However, in other scenarios—like in real-world markets—each player (or outcome) may have different probabilities of winning, akin to a dynamic game where some players hold advantages over others.

A probability space consists of three elements:

• Sample space (S): The set of all possible outcomes.
• Set of events (F): A collection of subsets of S, including S and the empty set.
• Probability function (P): A function that assigns a probability to each event in F.

A probability space is the fundamental structure in the Axiomatic definition. It is made up of three parts:

1. Sample Space (S): This is like a complete list of everything that could happen in a given experiment (e.g., flipping a coin gives you two possible outcomes: heads or tails).
2. Set of Events (F): This includes not just the entire sample space but also any specific grouping of outcomes (like just getting heads). This allows for the analysis of specific scenarios within the broader sample space.
3. Probability Function (P): This assigns a numerical value (the probability) to each event, essentially giving us a way to measure how likely each event is to occur based on the established framework.

Consider a jar filled with different colored marbles—this is your sample space (S). The event set (F) might include events like picking a red marble, or picking a marble of any color. The probability function (P) would then determine how likely those events are based on the proportions of marbles of each color in the jar.

Let P:F → [0,1] be a probability function. Then it must satisfy: Axiom 1 (Non-negativity): P(E) ≥ 0 for every event E ∈ F Axiom 2 (Normalization): P(S) = 1 Axiom 3 (Additivity): If E1, E2, E3,… are mutually exclusive events, then: P(⋃Ei) = ∑P(Ei) for i=1 to ∞.

Kolmogorov introduced three foundational axioms that form the basis of probability:

1. Non-negativity: This states that the probability of any event cannot be negative—it must always be zero or positive. Intuitively, you can't have a negative chance of something happening.
2. Normalization: This implies that the probability of the entire sample space happening is equal to 1, which serves as a benchmark or reference point. It's like saying 'something must happen; there's no chance it won't.'
3. Additivity: This axiom specifies that if you have several events that cannot happen at the same time (mutually exclusive), the probability of at least one of these events happening is the sum of their individual probabilities. This allows us to combine probabilities in a logical way and is crucial for dealing with complex events.

Think of an event like a concert, where you have different bands performing. Each band's performance is a distinct event (E1, E2, etc.) that cannot happen at the same time. According to the additivity axiom, if you wanted to know the total chance of seeing one of those bands, you could simply add together the individual chances of each band performing. Non-negativity ensures you never get a ‘negative chance’ of seeing a band, while normalization assures you that at least one performance will happen at the concert.

Example: Tossing a fair coin Let S = {H, T} Define F = {∅, {H}, {T}, {H, T}} Assign P({H}) = 0.5, P({T}) = 0.5 This satisfies all three axioms:

• Non-negativity: 0.5 ≥ 0
• Normalization: P({H, T}) = 0.5 + 0.5 = 1
• Additivity: P({H} ∪ {T}) = P({H}) + P({T})

This example illustrates the application of the Axiomatic Definition of Probability using a fair coin toss. Here, the sample space (S) consists of two outcomes: heads (H) or tails (T). The event space (F) includes all potential combinations of outcomes, including the empty event (∅) and the entire sample space itself. The probability function assigns a probability of 0.5 to both outcomes since it's a fair coin. The situation satisfies the axioms: Non-negativity shows both probabilities are greater than or equal to zero; normalization confirms that adding the probabilities results in one; and additivity holds since choosing either heads or tails adds up to the total probability of outcomes in the sample space.

When flipping a coin, you might think of the situation as a 50/50 decision, like choosing between two flavors of ice cream. If the ice cream shop has only vanilla and chocolate, and you want one scoop, you have an equal chance of getting either flavor. That’s the fairness of the coin toss reflected in the equal probabilities of 0.5 for heads and tails.

• Applicable to both finite and infinite sample spaces. • Handles both discrete and continuous cases. • Can model real-world probabilities where outcomes are not equally likely.

The Axiomatic Definition of Probability offers several significant advantages:

1. Applicability to Various Types of Sample Spaces: It can be used to model scenarios where there are a finite number of outcomes as well as those with an infinite range. This is especially useful in fields like statistics and engineering, where many real-world phenomena do not conform to simple models.
2. Handling Different Cases: It is versatile enough to accommodate both discrete events (like rolling dice) and continuous ones (like measuring time or distance).
3. Modeling Real-World Scenarios: Many situations in real life have probabilities that are not uniform (not all outcomes are equally likely), such as predicting weather patterns or economic data. The axiomatic framework can accurately calculate these probabilities, making it an invaluable tool in various fields.

Imagine a weather forecasting system. Axiomatic probability allows meteorologists to predict events like rain or sunshine accurately, considering many outcomes and their differing probabilities. Unlike a simple coin toss where the outcomes are equally likely, weather predictions involve complex models that account for numerous variables, creating a rich probability landscape that Axiomatic Definition can navigate effectively.

The classical definition is a special case of the axiomatic definition where all outcomes are equally likely, and the sample space is finite.

The classical definition of probability and the axiomatic definition are interconnected, with the classical being a simpler, specific instance of the axiomatic framework. The classical definition applies when we can assume that all outcomes are equally likely and that our sample space is finite. In other words, the axiomatic definition generalizes the classical approach, allowing for a broader range of applications that include cases where outcomes are not uniformly likely or where the sample space might be infinite.

Think back to a roulette wheel in a casino. The classical definition applies easily here, as each number and color has an equal chance (in a fair game). However, if you consider the casino's profit system or the odds based on certain outcomes, you're operating within the realm of the axiomatic definition, which gives you a richer understanding of the probabilities involved, accounting for more complex factors.

--