In discrete probability, every possible outcome has a chance, with each individual probability between 0 and 1. The key rule is that these probabilities add up to exactly 1, ensuring one outcome occurs. This idea underpins IB Mathematics probability concepts and helps clear common misconceptions about how likelihoods fit together.

Multiple Choice

What key property must the probabilities in a discrete probability distribution satisfy?

In a discrete probability distribution, a crucial requirement is that the sum of the probabilities of all possible outcomes must equal 1. This reflects the principle that one of the outcomes must occur. Each individual probability within the distribution can range from 0 to 1, inclusive, but when added together, they must total exactly 1. This ensures that the model accurately represents a complete set of outcomes for a particular random experiment. The option mentioning that probabilities should be less than 1 does not encompass the necessity for their total to equal 1. Stating that probabilities must be equal to 0 suggests that no outcomes have a chance of occurring, which contradicts the nature of probability distributions. Likewise, asserting that probabilities must exceed 1 is fundamentally incorrect, as probabilities cannot go beyond 1 in any proper distribution.

When you’re modeling randomness with a discrete probability distribution, there’s one rule that holds everything together like the spine of a good story: the probabilities must sum to 1. It’s the quiet, essential constraint that guarantees the model accounts for every possible outcome and nothing is left out.

A gentle reminder of the basics

Think of a simple experiment—like rolling a fair six-sided die. The possible outcomes are the six faces, and each face has a probability of 1/6. If you add up the probabilities of all outcomes (1/6 + 1/6 + 1/6 + 1/6 + 1/6 + 1/6), you land on 1. That’s not a coincidence. It’s exactly what a discrete probability distribution is designed to do: cover all the ways things could turn out, with no gaps and no overlaps. Each individual probability must be between 0 and 1, inclusive. Some outcomes might be impossible (probability 0), others quite likely (probability closer to 1), but together they form a complete, honest accounting of the experiment’s chances.

Non-negativity and normalization: two side-by-side principles

Two guiding principles keep everything honest:

  • Non-negativity: p(x) ≥ 0 for every outcome x. You can’t have a negative chance of something happening.

  • Normalization: the sum of p(x) over all possible outcomes x equals 1. This is the mathematical way of saying, “one of these things will happen.”

These aren’t just pedantic rules. They’re what makes probabilities behave the way we expect. If the sum were less than 1, there’d be some mysterious “missing” outcome with a chance to occur. If the sum exceeded 1, we’d be overcounting the chance that something happens. Neither situation makes sense in the real world.

A quick look at a few concrete examples

Let’s walk through some everyday scenarios to see the rule in action.

  • A biased coin

Imagine a coin that lands heads with probability 0.6 and tails with probability 0.4. The two outcomes cover all possibilities. 0.6 + 0.4 = 1. Easy, right? The distribution is fully normalized.

  • A hand of cards

Suppose you draw one card from a standard deck and you care about getting a heart, a diamond, a club, or a spade. If you assign p(heart) = 13/52, p(diamond) = 13/52, p(club) = 13/52, p(spade) = 13/52, you’ll see the sum is 1. Each suit has a fair share, and the total reflects the complete set of outcomes.

  • A more nuanced case: rolling a weighted die

What if the die isn’t fair and the probabilities are p(1) = 0.05, p(2) = 0.10, p(3) = 0.15, p(4) = 0.20, p(5) = 0.25, p(6) = 0.25? Add them up: 0.05 + 0.10 + 0.15 + 0.20 + 0.25 + 0.25 = 1. It’s still a perfectly valid discrete distribution, just with a bias toward the higher faces.

What if we toss in an impossible outcome?

If you consider a die-like experiment but declare that the outcome “7” has probability 0, while the outcomes 1 through 6 have their usual probabilities that sum to 1, you’re simply stating that 7 cannot happen. The rule still holds: the sum of the probabilities of all possible outcomes included in the model must be 1. If you tried to include an extra outcome with a positive probability without reducing the others, you’d push the total over 1, which is not allowed.

Why this property matters in statistics and beyond

Distributions aren’t just pretty tables of numbers. They’re the language we use to describe randomness. When you understand that the probabilities in a discrete distribution add up to 1, you gain several practical footholds:

  • Expectation and variability become meaningful. The expected value relies on p(x) and x, and since the p(x) sum to 1, you can interpret the average outcome over many repetitions with confidence.

  • Probability rules stay consistent. Whether you’re combining independent events or conditioning on something new, the normalization rule keeps the math neat and predictable.

  • Modeling stays coherent. If you change the outcome set, you adjust p(x) accordingly so the total remains 1. It’s a built-in consistency check.

A quick detour into the broader picture

Probabilities aren’t just about a single experiment; they’re the backbone of data-laden storytelling. When researchers describe a population’s characteristics, they often use discrete distributions to capture categories: colors, brands, test results, or survey choices. Each category gets a probability, and the collection must be complete. That completeness is what makes comparisons meaningful and inferences credible.

From the math to the intuition

students often ask, “Why exactly must the sum be 1?”. Here’s a way to feel the logic. Suppose you know the truth that one outcome will happen. The universe of all possible outcomes partitions the entire sample space into disjoint pieces. If you assign a nonnegative number to each piece representing its chance, the total “chance mass” across the whole space has to be exactly 1. If the total were less than 1, you’d be implying there’s a sliver of chance somewhere unaccounted for. If it were more than 1, you’d be implying chance could somehow exceed certainty. Neither makes sense in a world where randomness is governed by well-defined probabilities.

Connecting to real-life data work

In practical data work, you’ll often see the distribution described in a compact form, sometimes with a probability mass function (pmf) p(x). The pmf is just a function that assigns a nonnegative probability to each possible value x, and the sum over all x in the support equals 1. The math is clean, but the impact is human. It tells you there’s a complete story about what could happen, and that story is statistically sound.

A few quick notes you can hold onto

  • The support matters. If you’re only interested in a subset of outcomes, you can focus on that subset and still preserve the normalization within it, as long as you account for the rest being impossible (probability 0).

  • Dead or alive outcomes? A valid distribution can have many outcomes with probability 0. That simply means those outcomes don’t occur under this model. The essential thing is that the nonzero probabilities add up to 1.

  • Discrete vs. continuous. In discrete distributions, outcomes are countable (think integers: 0, 1, 2, …). In continuous distributions, outcomes form an interval of real numbers, and the idea of “probability of a single point” becomes 0, with probability mass described through densities and integrals. The common thread is still the normalization principle, but the tools and interpretations shift a bit.

A handful of memorable takeaways

  • The sum-to-one rule is the heartbeat of any discrete distribution.

  • Individual probabilities live between 0 and 1; together they must total 1.

  • This isn’t a picky constraint; it’s what makes probabilities coherent and models believable.

  • Whether you’re modeling a simple coin flip or a subtle bias in a game, the same principle applies: cover all possibilities and nothing more.

A final thought while we wander a moment

Probability is often about balance—how likely something is, and how that likelihood stacks up against all other possibilities. That balance is what gives us a lens to predict, compare, and understand random phenomena. When you see a distribution laid out, it’s not just a list of numbers; it’s a compact map of the chances that shape our outcomes. And the reason that map makes sense is wonderfully simple: all the possible routes add up to one shared destination.

If you’re ever unsure, a quick check can help. Add up all the listed probabilities. If they come to 1, you’re in good shape. If they don to something else, you know you’re either missing an outcome or you’ve misallocated a probability somewhere. It’s a small check, but it keeps the math honest—and that honesty shows up in every conclusion you draw from the data.