Law of Large Numbers in AP Statistics

In AP Statistics, the Law of Large Numbers says that as a random process is repeated many times, the relative frequency of an event (the proportion of trials where it happens) gets closer and closer to the event's true probability. It's the reason simulation can estimate probabilities.

What is the Law of Large Numbers?

The Law of Large Numbers describes what happens when you repeat a chance process over and over. In the short run, results jump around a lot. In the long run, the proportion of times an event happens settles down and homes in on the event's true (theoretical) probability. Flip a fair coin 10 times and getting 7 heads is no big deal. Flip it 10,000 times and the proportion of heads will almost certainly be very close to 0.5.

This is the engine behind Topic 2.3, Estimating Probabilities Using Simulation. When a probability is too messy to calculate, you can simulate the process many times. Then you count how often the event occurs and use that relative frequency as your estimate. The Law of Large Numbers is your guarantee that the estimate is trustworthy, as long as you run enough trials. A good way to think about it: randomness is unpredictable up close but predictable from far away. The law also says nothing about any single trial. The coin has no memory and doesn't "owe" you heads after a run of tails.

Why the Law of Large Numbers matters in AP® Statistics

The Law of Large Numbers lives in Topic 2.3, Estimating Probabilities Using Simulation. It's the idea that turns "I ran a simulation" into "so I can estimate the probability." Without it, a simulated proportion would just be a number. With it, that number becomes a defensible estimate that improves with more trials. The concept also sets up a big theme of the whole course, which is that more data means less variability. The same intuition comes back later when sample proportions and sample means get less variable as sample size grows. If you understand why 10,000 trials beat 100 trials, you already have the core intuition for sampling distributions.

How the Law of Large Numbers connects across the course

Estimating Probabilities Using Simulation (Topic 2.3)

This is the law's home base. A simulation produces a relative frequency, and the Law of Large Numbers is what lets you treat that relative frequency as an estimate of the true probability. More trials give you a better estimate.

Sampling Distributions and Sample Size (Unit 5)

The Law of Large Numbers is the gut-level version of a formula you'll see later. The standard deviation of a sample proportion shrinks as n grows, so estimates cluster tighter around the truth. Big n means less wobble.

Mean (Expected Value) of a Random Variable (Unit 4)

The law applies to averages too. If you repeat a random process many times, the average outcome settles toward the expected value. That's why casinos can count on profit even though every single bet is a gamble.

Theoretical Probability Rules like the Law of Total Probability (Unit 4)

Rules like the law of total probability and Bayes' theorem give you exact, calculated probabilities. The Law of Large Numbers explains why a well-run simulation should land close to those calculated answers, so you can use one method to check the other.

Is the Law of Large Numbers on the AP® Statistics exam?

Expect this mostly in multiple-choice questions about simulation, and as a justification sentence in free-response answers. Common moves you'll need to make:

• Justify a simulated estimate. If 1,000 simulations produce event A 237 times, estimating P(A) ≈ 0.237 relies on the Law of Large Numbers. A large number of independent trials makes the relative frequency close to the true probability.
• Explain why estimates change with more trials. A simulation of getting at least one 6 on three fair dice might give 0.38 after 50 trials and 0.421 after 5,000. The true probability is 1 − (5/6)³ ≈ 0.421. The bigger simulation lands closer because relative frequencies settle down in the long run.
• Read a sequence of estimates. Estimates of 0.23, 0.252, 0.261, and 0.267 at 100, 500, 1,000, and 10,000 trials show the classic pattern. Early values bounce, later values stabilize, and the 10,000-trial value is the best estimate.
• Compare two simulations. If Student A runs 100 trials and gets 0.43 while Student B runs 10,000 trials and gets 0.412, trust Student B's estimate more, even if the true probability is unknown.

No released FRQ has used the term verbatim. Still, naming it ("by the Law of Large Numbers, with many trials the relative frequency approximates the probability") is a clean way to justify a simulation conclusion on a free-response question.

The Law of Large Numbers vs Central Limit Theorem

Both involve large samples, so they get mixed up. The Law of Large Numbers is about where the estimate ends up. A relative frequency or sample mean gets close to the true value as the number of trials grows. The Central Limit Theorem is about shape. The sampling distribution of a sample mean becomes approximately normal when the sample size is large. In short, LLN tells you the estimate is accurate, and the CLT tells you what the distribution of estimates looks like.

Key things to remember about the Law of Large Numbers

• The Law of Large Numbers says that as you repeat a random process more times, the relative frequency of an event gets closer to its true probability.

• Simulation works because of the Law of Large Numbers, so a simulated proportion from many trials is a reasonable estimate of a probability you can't easily calculate.

• A simulation with more trials gives a more reliable estimate than one with fewer trials, so 10,000 trials beat 100 trials.

• The law describes long-run behavior only and says nothing about the outcome of any single trial or short streak.

• Results from a small number of trials can bounce around a lot, and that short-run variability is normal, not evidence that something is wrong.

• The same idea also applies to averages, which settle toward the expected value over many repetitions.

Frequently asked questions about the Law of Large Numbers

What is the Law of Large Numbers in AP Stats?

It's the principle that as a random process is repeated many times, the proportion of times an event occurs gets closer to the event's true probability. In AP Stats, it's the justification for using simulation (Topic 2.3) to estimate probabilities.

Does the Law of Large Numbers mean a coin is 'due' for heads after several tails?

No. That's the gambler's fallacy. Each flip is independent, so the chance of heads is still 0.5 every time, and the law only says the overall proportion evens out over thousands of flips.

How is the Law of Large Numbers different from the Central Limit Theorem?

The Law of Large Numbers says a sample proportion or mean gets close to the true value with many trials. The Central Limit Theorem says the sampling distribution of the sample mean becomes approximately normal for large samples. One is about accuracy, and the other is about shape.

How many trials do you need for the Law of Large Numbers to work?

There's no magic number, but more is always better. For example, a three-dice simulation for "at least one 6" gave 0.38 after 50 trials and 0.421 after 5,000, and the true value is about 0.421.

Is the Law of Large Numbers on the AP Stats exam?

Yes. It shows up most in multiple-choice questions about simulation, where you identify why an estimate is valid or which of two simulations to trust. It's also a strong way to justify a simulated probability in a free-response answer.