Binomial vs geometric: how to tell them apart

Binomial vs geometric: the one difference that matters

Both the binomial and geometric distributions describe a run of independent trials where each trial ends in success or failure and the probability of success, written $p$ (the success probability), stays the same on every trial.

The difference is what you count.

• Binomial: the number of trials is fixed ahead of time (call it $n$, the number of trials), and you count how many successes occur.
• Geometric: the number of trials is not fixed, and you count how many trials it takes to get the first success.

So when a problem hands you a set number of trials and asks how many successes, it is binomial. When it asks how long you wait for the first success, it is geometric. That single question, fixed count of trials or waiting for the first success, decides which distribution you use.

The conditions binomial and geometric share

Before you pick a formula, check that the setting fits these conditions. Both distributions need all of them.

• Two outcomes per trial. Each trial ends in success or failure. "Success" just labels the outcome you are counting, even if that outcome is something bad like a defect.
• Independent trials. The result of one trial does not change the probability on any other trial.
• Constant success probability. The value of $p$ is the same on every trial.

The two distributions split on one more condition:

• For the binomial distribution, the number of trials $n$ is fixed before you start, and $X$ (the count of successes) can be any whole number from $0$ to $n$.
• For the geometric distribution, you keep going until the first success, so the number of trials is itself a random outcome. Here $X$ (the trial of the first success) can be $1, 2, 3, \ldots$ with no upper limit.

Is this binomial or geometric? A quick check

Read the question and ask what is fixed and what you are counting.

1. Is the number of trials fixed? If yes, and you count successes, use the binomial distribution. Phrases like "in 10 attempts" or "out of 20 patients" point to a fixed $n$.
2. Are you waiting for the first success? If the question asks for the probability that the first success happens on a particular trial, or how many trials it takes, use the geometric distribution. Phrases like "until the first" or "the first defective is the 4th one" point to a geometric setting.

One quick tell: if the answer to "how many trials?" is a number the problem gives you, it is binomial. If "how many trials?" is the thing you are solving for, it is geometric.

Binomial and geometric side by side

This table lines up the pieces you use most. In the geometric column, $X$ counts trials starting from the first one.

Notice that both formulas are built from the same two pieces: a run of failures each with probability $1-p$, and successes each with probability $p$. The binomial counts a fixed batch; the geometric stops at the first success.

The formulas for each distribution

Binomial. For a fixed number of trials $n$ and success probability $p$, the probability of exactly $x$ successes is

where $\binom{n}{x}$ is the number of ways to choose which $x$ of the $n$ trials are the successes. The mean (the expected number of successes) and standard deviation are

Here $\mu_X$ (mu, the mean of $X$) and $\sigma_X$ (sigma, the standard deviation of $X$) summarize the whole distribution. This is the version printed on the AP formula sheet.

Geometric. Let $X$ be the trial on which the first success occurs, so $X$ can be $1, 2, 3, \ldots$. The probability that the first success lands on trial $x$ is

You need $x-1$ failures first, each with probability $1-p$, then one success with probability $p$. The mean number of trials until the first success is $\mu_X = \frac{1}{p}$, and the standard deviation is $\sigma_X = \frac{\sqrt{1-p}}{p}$. This is the standard college form; the AP formula sheet does not list it.

Common mix-ups to avoid

A few slips show up again and again once you know the two formulas.

• **Reading a fixed $n$ into a geometric problem.** If the wording says "until the first success," there is no $n$ to plug in. The trial number is what you solve for.
• **Forgetting the $x-1$ in the geometric exponent.** The first success on trial $x$ needs exactly $x-1$ failures before it, so the exponent is $x-1$, not $x$.
• Mislabeling the success. Pick which outcome counts as a success and keep $p$ tied to that outcome. If the defect is the success, then $p$ is the defect rate, and $1-p$ is the chance of a good item.
• Using the binomial mean for a geometric problem. The binomial mean is $np$; the geometric mean is $1/p$. They are not interchangeable.

When you are unsure, go back to the decision check above and name what is fixed and what you are counting before you touch a formula.

AP note: the geometric distribution and the Fall 2026 exam

The binomial distribution is part of the redesigned AP Statistics course as Unit 2, topic 2.10. You are expected to know its conditions, formula, mean, and standard deviation.

The geometric distribution was removed from the AP Statistics exam starting with the Fall 2026 course (first exam May 2027). It stays a standard topic in college introductory statistics, so it is worth knowing if you plan to take a college stats course or want the fuller picture. If you are studying only for the AP exam, put your time into the binomial distribution instead.

You can run the numbers with the binomial probability calculator, and work through more examples in the binomial probability guide.