On this concept: Formula

Not sure which lesson to do first? Find your level and get a plan →

Math · Statistics & Probability · Grade 6-8

Experimental vs. Theoretical Probability

Theoretical probability is calculated from known equally-likely outcomes (favorabletotal), while experimental probability is the observed fraction from actual trials (times it happenedtrials).

Ptheoretical=favorable outcomestotal possible outcomes Pexperimental=times event occurredtotal trials
Venn diagram of Event A and Event BA two-event view of experimental vs. theoretical probability.ABA ∩ B
A: Event A
B: Event B
A two-event view of experimental vs. theoretical probability.

Quick Answer

Theoretical probability is calculated from known equally-likely outcomes (favorabletotal), while experimental probability is the observed fraction from actual trials (times it happenedtrials). Use the theoretical kind when you can list equally likely outcomes, like the faces of a fair die. Use the experimental kind when you have results from real trials, or when no such list exists, like a thumbtack landing point up. As the number of trials grows, the experimental value usually settles close to the theoretical one.

Why This Matters

Many real chances cannot be worked out from a list of equal outcomes: whether a medicine works, whether a shot goes in, whether a seed sprouts. Experimental probability is how those are measured. Comparing the two kinds also tests a claim: if a 'fair' die shows a six 300 times in 1,000 rolls instead of about 167, the die is almost certainly not fair. Without the distinction, a short run of luck gets mistaken for proof.

Intuitive Explanation

A fair die has six faces, all equally likely, so before you roll it you can already say P(3)=16. That is theoretical probability: reasoning about how the object is built. Now roll it 60 times and suppose 3 comes up 14 times. Your experimental probability is 1460≈0.23, noticeably more than 16≈0.17. Nothing is wrong; 60 rolls is a small sample, and small samples wobble.

As the number of trials grows, the wobble shrinks compared with the total. In 60 rolls, 14 threes instead of the expected 10 is a difference of 4, a big share of 60. In 6,000 rolls the count of threes usually lands within about 60 of the expected 1,000, so the experimental fraction comes out within about one percentage point of 16. That settling down is the law of large numbers.

Some situations look theoretical but are not. A thumbtack can land point up or point down, two outcomes, yet the chance is not 12, because the two outcomes are not equally likely. A basketball shot either goes in or misses, but that does not make it 50%. Theoretical probability needs outcomes that are equally likely by symmetry or by design; without that, only trials can tell you the chance.

Think of it like this. Theoretical probability is like a recipe that says 'makes 24 cookies'; experimental probability is what your oven actually produced: 22 one time, 25 the next. Over many batches your average comes close to 24. If the recipe is wrong, like a loaded die, the batches settle somewhere else, and only your records show it.

Core idea. Theoretical probability is computed from equally-likely outcomes; experimental probability is the observed fraction from real trials, and they converge as trials grow.


When to Use

Use theoretical probability when the outcomes are equally likely and you can list them: a fair coin, a fair die, a well-shuffled deck. Use experimental probability when the problem reports trials or results ('out of 50 tries', 'in 200 spins') or when the outcomes have no symmetry. When both appear, compare them and think about how many trials there were. If the question is why the gap shrinks with more trials, that is the law of large numbers.

Tip. Ask: Does the probability come from reasoning about the possible outcomes (theoretical) or from counting what occurred in real trials (experimental)?

How to Recognize It

Before choosing which probability to give, look at what the problem hands you: a list of possible outcomes, a record of results, or both, and how many trials there were.

  1. Does the question ask what to expect in future trials, such as how many 6s will come up in 60 rolls?

    If yes, multiply the probability by the number of trials: theory predicts 16×60=10 sixes. If earlier results show the die is unfair, base the prediction on the experimental probability instead.

  2. Are all the possible outcomes equally likely, as on a fair die or a spinner with equal sections?

    If yes, theoretical probability is favorable over total. If the outcomes are not equally likely, like a thumbtack landing point up or a paper cup landing on its side, you need trials to estimate the chance.

  3. Were there only a few trials, say fewer than 50?

    If yes, expect the experimental value to wander: a fair coin gives 7 or more heads in 10 flips about 17% of the time. With hundreds of trials the two should be close, as the law of large numbers predicts.

  4. Is the question asking you to compare the two values and explain a gap?

    If yes, report both, then judge the gap by the number of trials. A small gap after many trials fits a fair object, while a large gap that lasts over hundreds of trials suggests the die or spinner is unfair.

Experimental vs. Theoretical Probability compared with Theoretical probability (basic), Relative frequency (statistics), and Law of large numbers

All four rows involve a fraction about chance. Ask where the fraction comes from: the possible outcomes (theoretical), counted results used to estimate a chance (experimental), counted results used only to describe data (relative frequency), or the rule linking them (law of large numbers).

Experimental vs. Theoretical Probability

Meaning
Use this when you must decide whether a probability comes from known equally-likely outcomes (theoretical) or from counting actual trial results (experimental).
Key test
Does the probability come from reasoning about the possible outcomes (theoretical) or from counting what occurred in real trials (experimental)?
Formula
Ptheoretical=favorable outcomestotal possible outcomes Pexperimental=times event occurredtotal trials
Example
You flip a coin 20 times and get 12 heads. Give the experimental probability of heads and compare it to the theoretical.

Theoretical probability (basic)

Meaning
The favorable-over-total calculation alone, without the data-vs-theory comparison.
Key test
Use when only the known setup is given.
Formula
favorabletotal
Example
Rolling an even number is 36

Relative frequency (statistics)

Meaning
The observed proportion treated as a data summary; experimental probability is the same number used to estimate chance.
Key test
Use when summarizing data rather than estimating likelihood.
Formula
counttotal
Example
60% of surveyed people said yes

Law of large numbers

Meaning
The principle that experimental APPROACHES theoretical, not a probability you compute.
Key test
Use when explaining WHY the two converge.
Example
Heads ratio nears 0.5 over many flips

Formula & Notation

Ptheoretical=favorable outcomestotal possible outcomes Pexperimental=times event occurredtotal trials
Ptheo(A)=∣A∣∣S∣; Pexp(A)=count of An; by LLN, Pexp(A)→Ptheo(A) as n→∞

How to read it: Ptheo for theoretical probability; Pexp or p^ for experimental (observed) probability

Worked Examples

Example 1 — Coin flips

Easy

You flip a coin 20 times and get 12 heads. Give the experimental probability of heads and compare it to the theoretical.
Show the steps and the answer
  1. Experimental: divide what happened by the number of trials, 1220=0.6.

    The 12 heads were actually observed, so this probability comes from data.

  2. Theoretical: a fair coin has 2 equally likely faces, 1 of them heads, so P(heads)=12=0.5.

    This value comes from reasoning about the coin; no flipping is needed.

  3. Compare: 0.6−0.5=0.1, which is 2 more heads than the expected 10.

    In 20 flips the number of heads typically varies by about 2 from 10, so this gap is ordinary.

Answer: Experimental =0.6, theoretical =0.5

Takeaway: Experimental probability counts what happened; with many more flips it would tend to settle near the theoretical 0.5.

Example 2 — Known fair die

Standard

Instead you're asked the probability of rolling a 4 on a fair die, with no trials done. Is that experimental?
Show the steps and the answer
  1. Look for data: the problem reports no rolls at all.

    Without trial results there is nothing to count, so an experimental probability cannot exist yet.

  2. Reason from the die: 6 equally likely faces, 1 of them a 4, so P(4)=16≈0.167.

    'Fair' tells you the faces are equally likely, which is exactly what theoretical probability needs.

  3. Predict what an experiment would show: in 600 rolls, about 600×16=100 fours, though rarely exactly 100.

    A theoretical probability predicts the long-run share of results; an experiment would only estimate it.

Answer: No — it's theoretical, 16

Takeaway: No trials means the probability is theoretical; experimental probability needs actual results to count.

Example 3 — Find the mistake: two outcomes, so one half?

Application

A thumbtack is dropped 50 times and lands point up 32 times. Noah writes: "There are two ways it can land, so P(point up)=12." Find the mistake and correct it.
Show the steps and the answer
  1. Test Noah's reasoning: 'two outcomes, so 12' works only when the outcomes are equally likely.

    A thumbtack is not symmetric like a coin; its shape can favour one way of landing.

  2. Use the data given: P(point up)≈3250=0.64.

    With no symmetry to justify 12, the trials are the best evidence available.

  3. Use it to predict: in 200 more drops, expect about 200×0.64=128 point-up landings.

    An experimental probability is an estimate; more drops would make it more reliable.

Answer: P(point up)≈3250=0.64, an experimental probability

Takeaway: Two possible outcomes do not mean a 50% chance each; when outcomes are not equally likely, estimate the probability from trials.

Where You'll See It

  • A basketball player's free-throw percentage

    A player who has made 168 of 200 free throws this season has an experimental probability of 168200=0.84 of making the next one. There is no theoretical value to calculate; the record of real shots is the only evidence.

  • Checking a game company's new dice

    A board-game company rolls a new die 1,200 times. A fair die should show about 1,200×16=200 sixes. Getting 205 is normal wobble; getting 290 means the die is badly made and the batch should be rejected.

Common Mistakes

  1. The slip: Reporting 12 when the problem gives trial results

    Instead: if data is provided, use observed-over-trials, not the theoretical value.

  2. The slip: Expecting experimental to equal theoretical in a few trials

    Instead: they only converge over MANY trials (law of large numbers).

  3. The slip: Thinking a deviation in 20 flips means the coin is unfair

    Instead: small samples wander; only large, persistent deviations suggest bias.


Mini Practice

Try these on your own. Tap Reveal when you want to check.

  1. A spinner has 4 equal sections: red, blue, green, yellow. In 40 spins it lands on red 13 times. Find the experimental and theoretical probability of red.

    Hint: Experimental uses the spin results; theoretical uses the equal sections.

  2. A fair die is rolled 60 times and shows a 3 eight times. Find the experimental probability of a 3 and the number of 3s you would expect.

    Hint: Expected count is the theoretical probability times the number of rolls.

  3. Ava made 18 of her last 25 free throws. Use her experimental probability to predict how many of her next 50 free throws she will make.

    Hint: Find her success rate, then apply it to 50 shots.

  4. Mia flips a coin 10 times and Leo flips one 1,000 times. Whose experimental probability of heads is more likely to be close to 0.5?

    Hint: Think about what happens to experimental results over many trials.

  5. You want the probability that a paper cup lands on its side when tossed. Can you find it theoretically, or do you need experimental probability?

    Hint: Are the cup's landing positions equally likely?

  6. A coin lands heads 7 times in 10 flips. Sam says, "The coin must be unfair, because 7 out of 10 is not one half." Correct Sam.

    Hint: How much can results wander in only 10 flips?

Want the full set?

50 practice questions for this concept — free to try, every one with a complete worked solution showing the why, not just the answer.

Frequently Asked Questions

If a coin is fair, why don't 10 flips give exactly 5 heads?

Because each flip is random, and 10 flips is a short run. Exactly 5 heads is the single most likely result, yet it happens only about 25% of the time; the other 75% of the time you get some other number. Getting 3, 4, 6 or 7 heads is common. The theoretical 12\frac{1}{2} describes the long run, not any one short set of flips.

After many flips, does the number of heads get closer to exactly half?

The fraction of heads gets closer to 12\frac{1}{2}, but the count can drift further from exactly half. In 100 flips you might be 4 heads away from 50, a fraction of 0.54. In 10,000 flips you might be 40 heads away from 5,000, a bigger gap in the count but a fraction of only 0.504. The law of large numbers is about the fraction, not the count.

Is experimental probability ever the only way to find a probability?

Yes, whenever the outcomes are not equally likely and there is no way to work out the chances by reasoning. Nobody can calculate the chance that a dropped slice of buttered toast lands butter side down, or that a particular player makes a free throw. Weather forecasts, insurance prices and sports statistics all rest on experimental probability: records of many past trials used to estimate what happens next.

What if the experimental result is very far from the theoretical one?

First check whether the setup really matches the theory. A die with worn edges, a spinner that sticks, or a coin tossed the same way every time may not be fair. If 600 rolls of a die give 160 sixes instead of about 100, the gap is far too big to blame on chance, and the die is probably loaded. Then the experimental value is the better estimate of the true chance.

Learning Path

Experimental vs. Theoretical Probability

You are here

You need Probability (a number from 0 to 1) and Sample Space (the list of possible outcomes). This concept adds the difference between a chance you reason out and a chance you measure. Next, the Law of Large Numbers explains why the two get closer, Sampling Distribution describes how much experimental results wobble, and Hypothesis Testing decides when a gap is too big to be chance.

See Also