Counting Outcomes Using Addition and Multiplication Principles

Hello! We’ll begin your probability and statistics refresher with a question that comes before calculating any probability: what exactly are the possible outcomes, and how many are there? This first module develops counting methods from basic rules toward arrangements and selections. Today’s lesson focuses on the two rules underneath them all: addition for alternative outcomes and multiplication for outcomes assembled in stages.

Allow about 35–40 minutes. The examples are small enough to inspect directly; that is useful, because a counting method is trustworthy only when each valid outcome is counted once.

---

Define the outcome before counting it

Suppose a payment request has a channel and a confirmation method. Is an outcome just the channel, or is it the pair consisting of channel and method? Those are different questions with different answers. In a finite counting problem, first specify what information makes two outcomes distinct.

For example, if there are two channels, card and bank transfer, the set of channels has two outcomes. But “card confirmed by email” and “card confirmed by a mobile notification” are distinct outcomes if the question asks for complete channel-and-confirmation configurations.

The same discipline applies before choosing a rule. Ask whether a complete outcome belongs to one of several alternative categories, or whether it requires one choice at each of several stages. We will use “or” and “and” as clues, but the structure of the outcomes—not the wording alone—decides the count.

Alternatives: the addition principle

Suppose you can select one of three report formats or one of two dashboard formats. If every format belongs to exactly one of those categories, there are $3+2=5$ choices. Addition works because the categories do not overlap: counting the reports and then the dashboards counts each possible choice once.

More formally, if finite sets $A$ and $B$ have no outcomes in common, then

$$|A\cup B|=|A|+|B|.$$

Here, $|A|$ means “the number of outcomes in $A$.” The rule extends to any number of mutually exclusive categories.

The video below develops this distinction between alternatives and stages. Its round-trip example is deliberately simplified: when we count all combinations of outward and return journeys, we must assume each stated return option is feasible after every outward option.

“Or” does not automatically mean add. Consider choosing a number from $1$ through $6$ that is even or greater than $3$. The even numbers are $\{2,4,6\}$; the numbers greater than $3$ are $\{4,5,6\}$. Adding $3+3$ counts $4$ and $6$ twice. One way to count correctly with today’s addition rule is to use disjoint categories: there are three even numbers and one odd number greater than $3$, namely $5$. Thus there are $3+1=4$ qualifying outcomes.

That example exposes the key condition: before adding category counts, check that each outcome fits in only one category. We will revisit systematic ways to handle overlaps later in the course.

---

Stages: the multiplication principle

Now suppose each of three available plans can be paired with one of two permitted add-ons. A complete outcome specifies both a plan and an add-on. For each plan there are two complete outcomes, giving

$$3\cdot 2=6.$$

The tree below makes the count visible. The initial $W$ is fixed, so it contributes one choice. Each of its three branches has two endings, producing six distinct paths.

Notice that the labels $P$ and $Y$ repeat. That does not duplicate an outcome: the path through $R$ ending at $P$ differs from the path through $G$ ending at $P$. The outcome records the choices at both stages.

In general, if there are $n$ choices at the first stage and exactly $m$ allowable choices after each first-stage choice, then the number of complete outcomes is

$$n\cdot m.$$

The actual second-stage choices need not be identical across branches. For example, plan A could permit add-ons $X,Y$, plan B could permit $X,Z$, and plan C could permit $Y,Z$. The count is still $3\cdot 2=6$: every first-stage choice has two continuations, and each plan–add-on pair is distinct. This is a counting rule, not an assumption of statistical independence.

Read the short section below for a more formal statement and an example in which the available choices change after the first selection.

Multiplication also works for more than two stages. If a complete configuration needs one choice from each of three stages, with respectively $2$, $3$, and $4$ allowable choices at every relevant branch, its count is $2\cdot 3\cdot 4=24$. The phrase at every relevant branch matters: if an earlier choice changes the number of later choices, a single product may no longer describe the whole problem.

---

When the number of continuations varies

Consider a request that uses either a card channel or a bank-transfer channel. Suppose a card request permits three verification methods, while a bank-transfer request permits two. A complete outcome specifies the channel and its verification method.

There are three outcomes in the card category and two in the bank-transfer category. Since an outcome cannot use both channels, the categories are disjoint. The count is therefore

$$3+2=5.$$

It would be wrong to multiply the two channel choices by $3$, because bank transfer does not have three continuations. The reliable method is to count complete outcomes within each first-stage branch, then add the disjoint branch totals.

If every one of those five configurations can additionally use any one of four notification options, each has four continuations. There are then

$$(3+2)\cdot 4=20$$

complete channel–verification–notification configurations. This is addition within the alternative channel branches, followed by multiplication for the notification stage.

The final video segment shows the same combination of principles in a route-finding example. Watch for the point at which several steps form one route, and the point at which different complete routes are added.

For a general two-stage problem, suppose the first stage has $r$ choices and first-stage choice $i$ has $m_i$ permissible continuations. Each first-stage choice defines a disjoint category of complete outcomes, so

$$\text{total outcomes}=\sum _{i=1}^r{m}_i.$$

If all the $m_i$ are the same number $m$, this becomes $r\cdot m$, the multiplication principle. In this sense, multiplication is a compact way of adding equally sized branches.

---

A quick way to check a count

Before settling on an answer, describe one complete outcome in words. Then check three things:

1. Coverage: Does every allowed outcome appear in a category or branch?
2. Uniqueness: Could the same outcome be counted in two places?
3. Branch size: If you multiplied stage counts, does every earlier choice really have the stated number of allowable continuations?

For a small problem, listing outcomes or drawing a tree can verify the arithmetic. For a large problem, the same questions still apply, even when drawing the whole tree is impractical. They will be especially important when counting outcomes to form a probability: an incorrect outcome count cannot be repaired by a later formula.

---

Key takeaways

Add the sizes of disjoint alternative categories. Multiply stage counts when each earlier choice has the stated number of continuations. If continuation counts differ, count within each branch and add the branch totals. Throughout, define a complete outcome precisely so that you neither omit nor double-count one.

Next, we’ll use these rules to count ordered arrangements of distinct objects. The multiplication principle will remain the foundation; the new challenge is tracking how the available choices shrink as objects are used.