Intuition can only take us so far when dealing with uncertainty. We know that a sample space, $S$, contains all possible outcomes of an experiment. But how do we assign probabilities in a consistent, logical way, especially when dealing with complex or even infinite outcomes? To build a robust framework, we need a set of fundamental rules, or axioms, that probability must always obey. These aren't just suggestions; they are the bedrock of modern probability theory.

A probability measure, $P$, is a function that assigns a probability to an event. For this function to be valid, it must satisfy three simple axioms.

First, the probability of any event can't be negative. This is the axiom of non-negativity.

Second, the probability of the entire sample space is 1. If our experiment happens, one of the possible outcomes must occur. This is the axiom of unit measure.

Third, and most powerful, is the axiom of additivity. If you have a set of events that are mutually exclusive (meaning no two can happen at the same time), the probability that any of them occurs is the sum of their individual probabilities.

The axioms tell us how a probability measure behaves, but we also need to define what it can act upon. In simple cases like a coin toss, the set of all events is easy: {Heads}, {Tails}, {Heads or Tails}, and {Neither} (the empty set).

But what about picking a random number between 0 and 1? The sample space is infinite. Can we assign a probability to any subset of numbers? It turns out the answer is no, due to some mathematical paradoxes. We need a more careful definition of what constitutes a valid 'event'.

This is where the sigma-algebra comes in. A sigma-algebra (or $\sigma$-field), denoted $\mathcal{F}$, is a collection of subsets of the sample space $S$ that we consider to be the official events. To qualify as a sigma-algebra, this collection must satisfy three properties:

A sample space $S$ paired with a sigma-algebra $\mathcal{F}$ is called a measurable space, $(S, \mathcal{F})$. This is the stage upon which we can apply our probability measure $P$. The full, formal structure of a probability space is therefore a triplet: $(S, \mathcal{F}, P)$.

The third axiom is great for mutually exclusive events, but what if events overlap? Consider a company assessing risks. The event 'Supply Chain Disruption' ($A$) and the event 'Regulatory Change' ($B$) are not mutually exclusive; both could happen.

If we simply add their probabilities, $P(A) + P(B)$, we double-count the scenario where both occur, $A \text{ and } B$. To find the probability of $A \text{ or } B$, we must correct for this by subtracting the probability of their intersection. This is known as the Inclusion-Exclusion Principle.

The principle extends to three or more events. For three events $A, B,$ and $C$, the formula gets more involved:

This pattern of adding individual probabilities, subtracting pairwise intersections, adding three-way intersections, and so on, is the essence of the general principle. It's a crucial tool for calculating probabilities in complex systems where multiple failures or successes can overlap, ensuring that our calculations remain consistent with the axioms.

This axiomatic foundation allows us to build the entire structure of probability theory, from random variables to complex stochastic processes, on solid ground.