JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

The Multiplication Principle

If a task splits into independent stages, you multiply the number of choices at each stage. This one humble idea is the engine behind every counting formula you are about to meet.

Why counting is the heart of classical probability

In the previous rung you learned the classical definition of probability: when every outcome in the sample space is equally likely, the probability of an event is just the number of outcomes that make it happen, divided by the total number of outcomes. That looks simple, but it quietly hands you a hard job. To compute P(A) you must count two things — the size of the event and the size of the whole sample space — and the outcomes are usually far too many to write out by hand.

So classical probability is really a counting problem in disguise. The whole rung you are starting now is a toolkit for counting cleverly, so that questions like "how likely is a full house?" or "what are the chances two people in a room share a birthday?" become arithmetic rather than guesswork. And every tool in that kit — permutations, combinations, sampling models, even the famous birthday problem — is built on a single foundation we set down here.

That foundation is the multiplication principle, also called the fundamental counting principle. It is the first thing combinatorics teaches and, honestly, almost everything later is this idea wearing a costume. Get it deep into your bones now and the rest of the rung will feel like bookkeeping.

The principle itself: stages multiply

Here is the statement. Suppose you build an outcome by making a sequence of choices: first choice 1, then choice 2, and so on up to choice k. If choice 1 can be made in n_1 ways, and no matter how it turns out choice 2 can be made in n_2 ways, and so on, then the total number of ways to complete the whole sequence is the product n_1 * n_2 * ... * n_k. You multiply the counts, stage by stage.

Picture a tiny example. You are choosing an outfit from 3 shirts and 2 pairs of trousers. Pick a shirt — 3 ways. Then for each shirt you still have 2 ways to pick trousers. So the outfits are 3 * 2 = 6. You can see it as a tree: 3 branches at the top, each splitting into 2, giving 6 leaves at the bottom. Counting the leaves of a branching tree IS the multiplication principle, and that mental image will carry you a long way.

shirt A --- trousers 1   (A,1)
        \-- trousers 2   (A,2)
shirt B --- trousers 1   (B,1)
        \-- trousers 2   (B,2)
shirt C --- trousers 1   (C,1)
        \-- trousers 2   (C,2)

        3 shirts x 2 trousers = 6 outfits
The choice tree: 3 first-stage branches, each with 2 second-stage branches, gives 3 * 2 = 6 leaves.

The fine print: the count per stage must not change

The principle has a condition that is easy to gloss over, so read it carefully. The number of ways to make a later choice may depend on which choice you are at, but it must not depend on which earlier options you actually picked. In the outfit example, you always had 2 trouser options regardless of the shirt — that is why the multiplication is clean. Whenever the count at a stage is genuinely the same number for every path leading into it, you may multiply.

Watch what happens when the count does change. Suppose you seat 3 different people in 3 chairs. The first chair: 3 people to choose from. The second chair: only 2 people are left, because one is already seated — but crucially it is always 2, whoever sat first. The third chair: always 1. So the answer is 3 * 2 * 1 = 6. The per-stage number shrank, yet it shrank by the same amount on every path, so multiplication still applies. That product 3 * 2 * 1 is exactly the factorial 3!, and this is precisely the bridge into the next guide on permutations.

Multiply for "and," add for "or"

The multiplication principle has a partner that handles a different shape of problem: the addition rule for counting. The two split the work by a simple linguistic test. When an outcome is built by doing one stage and then another stage (shirt and trousers, first chair and second chair), you multiply. When an outcome falls into one category or a separate category with no overlap, you add the sizes of those categories.

A worked mix makes it click. How many ways can you travel from town X to town Z, if you must pass through town Y, and there are 3 roads from X to Y and 4 roads from Y to Z? That is X-to-Y AND Y-to-Z, so 3 * 4 = 12 routes. Now a different question: how many direct trips leave X today, if there are 3 buses OR 4 trains? Those are exclusive categories, so 3 + 4 = 7. The word "and" reached for multiplication; the word "or" reached for addition.

Order built in: why sampling is multiplication

Here is the deep reason this guide opens the rung. Almost every counting model you will study is a sequence of choices in disguise, and the multiplication principle is what turns that sequence into a number. Consider drawing items one at a time from a bag of 10 distinct tokens. If you put each token back before the next draw — sampling with replacement — every one of, say, 3 draws has 10 options, giving 10 * 10 * 10 = 10^3 = 1000 ordered results. The count per stage never changes, so the product is a clean power.

Now do not replace the token after each draw — sampling without replacement. The first draw has 10 options, the second has 9 (one is gone), the third has 8, giving 10 * 9 * 8 = 720 ordered results. The shrinking-by-one pattern is the factorial idea again, and the distinction between these two setups is exactly the topic of sampling with and without replacement later in this rung. Notice that both counts are of ordered sequences; whether order should matter at all is the ordered-versus-unordered question that separates permutations from combinations in the next two guides.

  1. Decide whether you are counting and-stages (multiply) or or-categories (add). Most build-it-up problems are and-stages.
  2. List the stages in order, and write down how many options each stage has.
  3. Check the fine print: is each stage's count the same for every earlier path? If not, split into cases first.
  4. Multiply the per-stage counts to get the number of ordered ways.
  5. Ask whether order should count. If it should not, you will later divide out the duplicate orderings — that is the leap to combinations.

A first honest probability calculation

Let us close the loop back to probability, because counting is only the means. Roll a fair die twice. By the multiplication principle the sample space has 6 * 6 = 36 equally likely ordered outcomes. What is P(both rolls show a 6)? There is exactly 1 favorable ordered outcome, so P = 1/36. What is P(the two rolls differ)? Count the favorable outcomes directly with multiplication: 6 choices for the first roll, then 5 choices for the second that avoid matching it, giving 6 * 5 = 30, so P = 30/36 = 5/6.

That last move — count the favorable outcomes as a fresh multiplication, then divide by the size of the sample space — is the rhythm of nearly every classical probability problem in this course. The whole sample space is one product; the event is another product (or a sum of products); the probability is the ratio. Honest warning: this only works when the outcomes you are counting really are equally likely. If you collapse the 36 ordered die-pairs into 21 unordered pairs, those 21 are NOT equally likely (the pair {3,5} can happen two ways while {4,4} can happen only one), and dividing by 21 would give wrong answers. Counting and equal likelihood must agree, or the classical formula lies.