When Does a Count Have a Binomial Setting?
In Exam Practice on Mean and Standard Deviation of Random Variables, you combined a random variable’s meaning with summaries of its probability distribution. A common next step is to recognize when a count of outcomes can be described by a binomial distribution. Before using a binomial model, check how the chance process is set up.
A free-throw sequence is a useful example. A player takes a specified number of shots, and we count how many go in. Whether a binomial model is appropriate depends on more than the fact that the result is a count. Each shot must fit the same set of conditions.
The four conditions are often summarized as binary outcomes, independence, fixed number of trials, and constant probability of success. The word “success” is just a label for the outcome being counted. It does not have to mean something desirable: a study could define a missed shot, a late bus, or a defective part as a success if that is the outcome of interest.
When all four conditions are reasonably satisfied, we can identify \(X\) as a binomial random variable with parameters \(n\) and \(p\). Here, \(n\) is the fixed number of trials, and \(p\) is the probability of success on each trial. We write \(X\sim\operatorname{Binomial}(n,p)\). Recognizing the model comes before using it to find probabilities.
Check Each Condition in the Situation
The conditions describe the chance process, not just the wording of a question. In a written response, name what counts as one trial, state which outcome is success, identify the number of trials and the proposed success probability, and explain whether independence and constant probability are reasonable.
For each trial, there are two relevant outcomes: success or failure. Define success clearly so the counted event is unambiguous.
The outcome of one trial does not change the probabilities for the other trials. Consider whether earlier results, shared conditions, or sampling without replacement could create dependence.
The number \(n\) is determined before the process begins. The process does not end early because a success or failure occurs.
Each trial has the same probability \(p\) of success. The person, conditions, or selection process should not change that probability from one trial to the next.
Independence and constant probability are related but are not the same condition. Independence concerns whether trial outcomes affect one another. Constant probability concerns whether each trial has the same chance of success. A setup can meet one condition and fail the other.
These conditions are often idealizations. For example, a player’s concentration or fatigue could change over a sequence of shots. A binomial model does not claim those factors are impossible; it treats the shots as having a common success probability and independent outcomes when that is a reasonable approximation for the question.
Worked Example: Counting Made Free Throws
Worked Example: Counting Made Free Throws
A player takes 12 free throws during an independent practice session. For this model, suppose each shot has a 0.75 probability of going in, and the result of one shot does not affect the next. Let \(X\) be the number of made shots. Is a binomial model appropriate? If so, identify its parameters.
State. One trial is one free-throw attempt. Define success as making the shot and failure as missing it. The random variable \(X\) counts made shots among the 12 attempts.
Plan. Check whether the shots have two relevant outcomes, whether they are independent, whether there are a fixed 12 attempts, and whether the probability of making a shot stays at 0.75.
Do. Each attempt results in a made shot or a missed shot, so the binary-outcomes condition is met. The number of attempts is fixed at \(n=12\), so the fixed-number condition is met. The situation explicitly assumes that the outcome of one shot does not affect the next, supporting independence. It also specifies the same probability, \(p=0.75\), for every shot, so the constant-probability condition is met.
Conclude. Under the stated assumptions, a binomial model is appropriate: \(X\sim\operatorname{Binomial}(12,0.75)\). This says \(X\) counts the number of made shots in 12 trials, with the same success probability on each trial. The model is conditional on the assumptions; it does not guarantee that a real player’s ability or circumstances never change.
What the Free-Throw Model Assumes
The free-throw example makes a useful distinction between the information given by a scenario and the assumptions needed for a model. Knowing that a player takes 12 shots establishes a fixed number of attempts. It does not, by itself, establish independence or the same probability of making each shot.
For instance, a player could become tired, change technique, or gain confidence after several shots. Those changes might make later outcomes more or less likely. If a problem explicitly says to assume independence and a constant success probability, use those assumptions. If it does not, explain why a binomial model is plausible as an approximation rather than claiming the conditions have been proved.
Worked Example: A Free-Throw Sequence That Is Not Binomial
Worked Example: A Free-Throw Sequence That Is Not Binomial
A training plan asks a player to take shots until the player makes 5 free throws. Let \(Y\) be the total number of attempts needed. Each attempt is either made or missed, but the session ends as soon as the fifth shot is made. Is \(Y\) binomial?
State. One trial is one free-throw attempt. A success is a made shot. The variable \(Y\) counts attempts until the fifth made shot.
Plan. Check the two-outcome, independence, fixed-number, and constant-probability conditions. In particular, determine whether the total number of attempts is set before the sequence begins.
Do. Each attempt has two outcomes, made or missed, so the binary condition is met. If we assume the player’s shot outcomes are independent and the probability of making each shot stays constant, those two conditions are also met. However, the number of attempts is not fixed in advance. It depends on when the fifth made shot occurs; the session can end after different numbers of attempts.
Conclude. \(Y\) is not a binomial random variable because the fixed-number-of-trials condition fails. The fact that the goal is to make five shots does not mean there are five trials: the number of attempts is the variable quantity. Do not describe this count as binomial just because each attempt has two possible outcomes.
This example highlights why it matters to define the random variable precisely. “Number of successes in 12 attempts” has a fixed number of trials. “Number of attempts to get 5 successes” does not. They describe different counts, and only the first has the fixed \(n\) required for a binomial setting.
Worked Example: Same Number of Trials, Changing Probability
Worked Example: Same Number of Trials, Changing Probability
A player takes 10 free throws. A coach’s practice plan assigns a 0.70 chance of making each of the first 5 shots and a 0.60 chance of making each of the last 5 shots because the last group follows a strenuous drill. Assume the outcomes are independent. Let \(Z\) count made shots. Is \(Z\) binomial?
State. One trial is one shot, success is making the shot, and \(Z\) is the number of made shots among 10 attempts.
Plan. Check the conditions separately. The situation gives a fixed total of 10 attempts and says the outcomes are independent. Compare the possible outcomes and the success probabilities across all 10 trials.
Do. Each shot is either made or missed, so the binary condition is met. There are exactly 10 attempts, so the fixed-number condition is met. Independence is given. But the success probability is 0.70 for the first 5 shots and 0.60 for the last 5. Since the probability changes, there is no single common value of \(p\) for every trial.
Conclude. \(Z\) is not binomial under this description because the constant-probability condition fails. Having a fixed number of trials and independent outcomes is not enough. A binomial model for the total number of made shots would require the same success probability on every attempt.
The changing-probability example also shows why “same kind of trial” needs careful thought. All 10 events are free throws, but the assigned chances differ between the two groups. A shared label for the activity does not automatically make the success probability constant.
Common Mistakes and AP Exam Tips
A strong model justification connects each condition to a detail in the setting. Simply writing “the four conditions hold” is less convincing than naming the conditions and explaining how the scenario supports them.
- Leaving success undefined. State exactly which outcome is counted. For the free-throw example, success means making a shot; \(X\) counts made shots.
- Confusing the number of successes with the number of trials. In a fixed sequence of 12 shots, \(n=12\), not the number of shots the player makes. The count \(X\) can vary from 0 to 12.
- Assuming binary outcomes prove the setting is binomial. Made/missed is binary, but a variable stopping point or changing probability can still make the setting non-binomial.
- Treating independence and constant probability as interchangeable. Check both. A problem may state independence while giving different success probabilities across groups.
- Calling a stopping rule a fixed number of trials. “Take shots until 5 are made” fixes the target number of successes, not the number of attempts. Explain that the trial count can vary.
- Claiming assumptions are facts. If a model assumes a player’s success chance stays the same, say that it is an assumption or a reasonable approximation when appropriate. Do not claim the scenario proves a person’s ability cannot change.
- Giving parameters without context. When a binomial model is appropriate, state what \(n\) counts and what \(p\) represents. For example, \(n=12\) free throws and \(p=0.75\) chance of making each shot.
For full-credit communication, a concise but complete justification might say: “Let \(X\) count made free throws in 12 attempts. Each attempt is a made or missed shot; the 12 attempts are fixed; the outcomes are assumed independent; and each shot has the same 0.75 chance of being made. Thus \(X\sim\operatorname{Binomial}(12,0.75)\).” If a condition fails, name it and point to the detail that causes the failure.
Key Takeaway
A count is binomial only when the chance process has all four features: two outcomes per trial, independent trials, a fixed number of trials, and the same probability of success on every trial. In a free-throw setting, define a made shot as success, identify the fixed number of attempts, and check whether independence and a common success probability are reasonable.
Check Your Understanding
For each situation, define the relevant count and decide whether a binomial model is appropriate. Explain which conditions hold or fail.
- A basketball player takes 8 free throws. Assume each shot is made or missed, the outcomes are independent, and the chance of making each shot is 0.65. Define \(X\) and identify \(n\) and \(p\).
- A player takes free throws until missing one shot. Explain why a fixed number of trials is or is not satisfied.
- A machine checks 20 packages, and each package is either correctly sealed or not. Assume the checks are independent, but the chance of a correct seal increases after the machine is adjusted halfway through. Which binomial condition fails?
- In a fixed sequence of 15 independent attempts, the probability of success is the same on every attempt. Describe what \(X\) counts and explain why the number of successes is not fixed even though \(n\) is.
- A student says that a count must be binomial because every trial has a success and a failure. Identify the missing checks the student must make.