What Is The Binomial Distribution?
The binomial distribution is the probability distribution of the number of successes in a fixed number of repeated, independent trials that each have the same chance of success. Each trial is a Bernoulli trial: it has exactly two outcomes, usually labelled success and failure. Counting how many successes appear across all the trials gives a random whole number, and the binomial distribution tells you how likely each possible count is.
Two numbers define the whole distribution:
$n$, the number of trials (how many times the experiment is repeated).
$p$, the probability of success on a single trial (the same for every trial).
We write $X \sim B(n, p)$ to mean the random variable $X$ follows a binomial distribution with those two parameters. The value of $X$ can be any whole number from $0$ (no successes) up to $n$ (all successes). Because $X$ counts things, the binomial is a discrete distribution, one built on a discrete random variable rather than a continuous measurement.
What Are The Four Conditions For A Binomial Distribution?
A situation follows a binomial distribution only when all four Bernoulli-trial conditions hold at once. Miss one and the formula no longer applies.
Fixed number of trials. You decide $n$ in advance, such as "flip the coin 10 times." The number of trials cannot depend on the results.
Two outcomes per trial. Every trial ends in one of two categories: success or failure, heads or tails, defective or fine.
Constant probability of success. The value of $p$ is the same on every trial. A fair coin keeps $p = 0.5$ throughout.
Independent trials. The result of one trial does not change the probability of any other. One coin flip tells you nothing about the next.
Table: A quick test for whether a situation is binomial.
Condition | Question to ask | Coin-flip check |
|---|---|---|
Fixed $n$ | Is the number of trials set in advance? | Yes, 10 flips |
Two outcomes | Only success or failure each time? | Yes, heads or tails |
Constant $p$ | Same chance every trial? | Yes, $p = 0.5$ |
Independence | Does one result affect another? | No |
When you draw cards without putting them back, the probability shifts after each draw, so independence fails and the model does not fit. That single check prevents the most common misuse of the formula.
What Is The Binomial Distribution Formula?
The probability of getting exactly $k$ successes in $n$ trials is given by the binomial probability mass function:
$$P(X = k) = \binom{n}{k} , p^{k} , (1-p)^{n-k}$$
Each piece earns its place in the formula.
Table: What each symbol in the binomial formula means.
Symbol | Meaning |
|---|---|
$n$ | Number of trials |
$k$ | Number of successes you want ($0 \le k \le n$) |
$p$ | Probability of success on one trial |
$1-p$ | Probability of failure on one trial (often written $q$) |
$\binom{n}{k}$ | Number of ways to choose which $k$ of the $n$ trials are the successes |
The term $\binom{n}{k}$, read "$n$ choose $k$," is the nCr combination count, the number of ways to choose which $k$ of the trials are successes. It matters because the $k$ successes can land in many different orders, and every order is equally likely. The factor $p^{k}$ is the chance of the $k$ successes, and $(1-p)^{n-k}$ is the chance of the remaining $n-k$ failures. Drop that last factor and the answer is wrong, which is exactly where many students slip.
How Do You Calculate Binomial Probabilities?
To use the formula, identify $n$, $p$, and $k$, then substitute and compute. Two worked examples show the pattern.
Example 1: A fair coin is flipped 5 times. What is the probability of exactly 3 heads?
Here $n = 5$, $p = 0.5$, and $k = 3$. Substitute into the formula:
$$P(X = 3) = \binom{5}{3}(0.5)^{3}(0.5)^{2}$$
Work out each part on its own line:
$$\binom{5}{3} = 10$$
$$(0.5)^{3} = 0.125, \qquad (0.5)^{2} = 0.25$$
$$P(X = 3) = 10 \times 0.125 \times 0.25 = 0.3125$$
Final answer: the probability of exactly 3 heads is $0.3125$ (or $31.25%$).
Example 2: A machine produces items with a 2% defect rate. In a box of 10, what is the probability of exactly 2 defective items?
Here $n = 10$, $p = 0.02$ (defective is the "success" we are counting), and $k = 2$:
$$P(X = 2) = \binom{10}{2}(0.02)^{2}(0.98)^{8}$$
$$\binom{10}{2} = 45, \qquad (0.02)^{2} = 0.0004, \qquad (0.98)^{8} \approx 0.8508$$
$$P(X = 2) = 45 \times 0.0004 \times 0.8508 \approx 0.0153$$
Final answer: about $0.0153$ (or $1.53%$). Notice the failure factor $(0.98)^{8}$ carries most of the weight here, which is precisely the term beginners forget.
How Do You Handle "At Least" And "At Most" Cases?
Many questions do not ask for exactly $k$ successes. They ask for "at least," "at most," or "fewer than," and these need a sum of several binomial probabilities, not a single term.
The key is to read the phrase as a range of values and add the probabilities across that range.
"At most $k$" means $X \le k$: add $P(X=0) + P(X=1) + \dots + P(X=k)$.
"At least $k$" means $X \ge k$: add from $k$ up to $n$, or use the shortcut below.
"Fewer than $k$" means $X < k$, so stop at $k-1$. "More than $k$" means $X > k$, so start at $k+1$.
Example 3: A fair coin is flipped 4 times. What is the probability of at least 1 head?
Adding $P(X=1)$ through $P(X=4)$ works, but the complement is faster. "At least 1" is the opposite of "none":
$$P(X \ge 1) = 1 - P(X = 0) = 1 - (0.5)^{4} = 1 - 0.0625 = 0.9375$$
Final answer: $0.9375$ (or $93.75%$). The complement shortcut, $P(X \ge 1) = 1 - P(X=0)$, turns a four-term sum into one subtraction, and it is worth reaching for whenever "at least one" appears.
What Are The Mean And Variance Of A Binomial Distribution?
The mean (expected value) of a binomial distribution is $np$, and its variance is $np(1-p)$. The standard deviation is the square root of the variance.
$$\text{Mean} = np, \qquad \text{Variance} = np(1-p), \qquad \text{SD} = \sqrt{np(1-p)}$$
The mean has a plain reading: if each trial succeeds with probability $p$, then across $n$ trials you expect about $np$ successes. For 10 coin flips at $p = 0.5$, the expected number of heads is $10 \times 0.5 = 5$, which matches intuition.
For the defective-items example ($n = 10$, $p = 0.02$):
$$\text{Mean} = 10 \times 0.02 = 0.2, \qquad \text{Variance} = 10 \times 0.02 \times 0.98 = 0.196$$
So a box of 10 holds $0.2$ defective items on average, a useful figure for quality control even though no single box can contain a fifth of a defect.
Why Does The Binomial Distribution Work?
The formula is not an arbitrary rule to memorise. It falls straight out of the binomial theorem, the same expansion that gives $(a+b)^{n}$.
Write $q = 1 - p$ for the failure probability. Expanding $(p + q)^{n}$ produces one term for every possible number of successes:
$$(p + q)^{n} = \sum_{k=0}^{n} \binom{n}{k} p^{k} q^{n-k}$$
Each term $\binom{n}{k} p^{k} q^{n-k}$ is exactly $P(X = k)$. The binomial distribution is the expansion of $(p+q)^{n}$, term by term. Two consequences follow at once:
The probabilities must add to $1$, because $p + q = 1$, so $(p+q)^{n} = 1^{n} = 1$. The whole distribution is guaranteed to be complete.
The coefficients $\binom{n}{k}$ are the rows of Pascal's triangle. The number of ways to arrange $k$ successes among $n$ trials is the same counting problem that builds each row of the triangle.
That is the deep reason the distribution carries the name "binomial." It is the arithmetic of "two outcomes, repeated," and the algebra of two-term powers is the same structure viewed from a different angle.
Who Discovered The Binomial Distribution?
The pieces arrived over centuries. Blaise Pascal worked out the triangle of coefficients in the 1650s while trading letters with Pierre de Fermat about gambling problems, and Jacob Bernoulli turned repeated two-outcome trials into a full theory soon after.
Two other mathematicians shaped the same idea:
Blaise Pascal (1623–1662, France) built the triangle of coefficients that supplies every $\binom{n}{k}$ in the formula.
Abraham de Moivre (1667–1754, France and England) showed that for large $n$ the binomial distribution is closely approximated by the smooth bell curve, an early bridge to the normal distribution.
Where Is The Binomial Distribution Used In The Real World?
Any setting with repeated yes-or-no trials at a steady rate is a candidate for the binomial distribution.
Quality control: a factory samples $n$ items and asks how many are likely to be defective, using a known defect rate as $p$.
Medicine and clinical trials: if a treatment works for a fraction $p$ of patients, the binomial models how many of $n$ patients are expected to respond.
Genetics: the number of offspring inheriting a trait across $n$ births follows a binomial pattern when each birth carries the same probability.
Elections and surveys: the count of "yes" responses in a poll of fixed size behaves binomially when each response is independent.
Sports and games: the number of successful free throws in $n$ attempts, at a player's steady success rate, is a binomial variable.
One formula, built from coin-flip logic, quietly runs quality checks, drug trials, and opinion polls. Mathematics turns a child's coin game into a tool that industries depend on.
What Are The Most Common Binomial Distribution Mistakes?
These four errors account for most lost marks on the binomial distribution, verified against StatisticsHowTo, Cuemath's formula guide, and standard statistics revision notes.
Dropping the failure factor $(1-p)^{n-k}$
Where it slips in:
A student computes $\binom{n}{k}p^{k}$ and stops, forgetting that the $n-k$ trials that were not successes each carry a probability too.
Don't do this:
Do not treat $\binom{n}{k}p^{k}$ as the answer. Without $(1-p)^{n-k}$ the probabilities will not add to $1$.
The correct way:
Always write all three parts: the coefficient $\binom{n}{k}$, the success factor $p^{k}$, and the failure factor $(1-p)^{n-k}$, then multiply.
Using the binomial model when trials are not independent or $p$ is not constant
Where it slips in:
A student applies the formula to drawing cards without replacement, or to a situation where the success rate changes partway through.
Don't do this:
Do not use the binomial distribution when sampling without replacement, because each draw changes the probability and independence fails.
The correct way:
Check the four conditions first. If $p$ shifts between trials, or one result affects another, a different model applies, not the binomial.
Swapping $p$ and $1-p$
Where it slips in:
When "success" is the rare event (a defect, a failure), a student plugs in the common outcome's probability by habit.
Don't do this:
Do not assume success means the pleasant outcome. Success is simply the event you are counting.
The correct way:
Define success explicitly at the start. If you are counting defects, then $p$ is the defect rate and $1-p$ is the rate of good items.
Mishandling "at least" and "at most" as a single term
Where it slips in:
A student answers "at least 2 successes" by computing only $P(X = 2)$, ignoring the larger counts the phrase includes.
Don't do this:
Do not treat a cumulative phrase as one exact value. "At least 2" covers $X = 2, 3, \dots, n$.
The correct way:
Translate the words into a range, then sum the terms across it. For "at least one," use the complement $1 - P(X = 0)$ to save work.
Practice Problems On The Binomial Distribution
Try these, then check the answer that follows each. Use the formula $P(X=k) = \binom{n}{k}p^{k}(1-p)^{n-k}$.
A fair coin is flipped 6 times. Find $P(X = 4)$ heads.
(Answer: $\binom{6}{4}(0.5)^{4}(0.5)^{2} = 15 \times 0.015625 = 0.2344$, about $23.44%$.)A die is rolled 4 times. Find the probability of exactly two sixes ($p = \tfrac{1}{6}$).
(Answer: $\binom{4}{2}\left(\tfrac{1}{6}\right)^{2}\left(\tfrac{5}{6}\right)^{2} = 6 \times \tfrac{1}{36} \times \tfrac{25}{36} \approx 0.1157$.)A multiple-choice quiz has 8 questions, each with a $0.25$ guess rate. Find the expected number of correct guesses.
(Answer: mean $= np = 8 \times 0.25 = 2$.)For $X \sim B(20, 0.1)$, find the variance.
(Answer: $np(1-p) = 20 \times 0.1 \times 0.9 = 1.8$.)A basketball player makes 70% of free throws. In 3 attempts, find the probability of at least one make.
(Answer: $1 - (0.3)^{3} = 1 - 0.027 = 0.973$, about $97.3%$.)A machine is 99% reliable per run. Over 5 runs, find the probability that all 5 succeed.
(Answer: $\binom{5}{5}(0.99)^{5}(0.01)^{0} = (0.99)^{5} \approx 0.951$.)
Where Should You Go Next After The Binomial Distribution?
The binomial distribution opens onto the wider world of probability, and several natural doors lead onward from here.
Discrete random variable. Step back to the general idea a binomial variable is one example of, and see how any counting variable gets a probability distribution.
Binomial theorem. Follow the algebra that supplies the coefficients, and see why $(p+q)^{n}$ and the distribution are the same structure.
Negative binomial distribution. Flip the question around: instead of fixing the trials and counting successes, fix the successes and count the trials needed.
If your child is building these foundations, a live Bhanzu trainer teaches the binomial distribution starting from the coin-flip intuition and connecting it back to the algebra, in the Bhanzu algebra program.
Was this article helpful?
Your feedback helps us write better content