Probability Calculator

A probability calculator finds the chance an event happens by dividing favorable outcomes by total outcomes: P = favorable ÷ total. Rolling a 4 on a six-sided die gives P = 1/6 = 0.1667 = 16.67%. This tool also computes combined events (A and B, A or B) and the odds of something happening at least once over many trials, returning every answer as a fraction, a decimal and a percent.

To count the outcomes behind a probability (how many poker hands or lottery tickets there are), use the Permutation & Combination Calculator; to weigh each outcome by what it pays, the Expected Value Calculator; and for the chance of a value under a bell curve, the Z-Score Calculator.

The core formulas

Probability is a number between 0 and 1. A 0 means the event is impossible; a 1 means it is certain. The four rules this tool uses are:

When the events are independent (one does not change the other), P(B given A) is simply P(B), and the multiplication rule becomes P(A) × P(B). When they are mutually exclusive, P(A and B) is 0 and the addition rule reduces to P(A) + P(B). The "at least once" rule uses the complement: instead of adding every way the event could happen, it works out the single probability that it never happens and subtracts that from 1.

How to use this probability calculator

Pick the mode that matches your question, then fill in the inputs:

1. Choose a mode: "Single event," "Two events," or "Repeated trials" at the top.
2. Single event: enter how many outcomes count as a success (favorable) and how many outcomes are possible in total. The quick buttons load common setups like a coin or a die.
3. Two events: enter P(A) and P(B) as a decimal (0.25), a percent (25%) or a fraction (1/4), then choose whether the events are independent, mutually exclusive or dependent. For dependent events, also enter P(B given A). Each field shows its percent equivalent so you can double-check.
4. Repeated trials: enter the probability per try and how many tries (n), and read the chance it happens at least once.

The result updates instantly. The large number at the top is the headline answer, and the cards beneath it give the fraction (or a "1 in N" frequency), the decimal and a full breakdown of the related probabilities.

Worked example 1: a single die roll

What is the probability of rolling an even number on a die? The even faces are 2, 4, and 6: 3 favorable outcomes out of 6. So P = 3/6, which reduces to 1/2 = 0.5 = 50%. The complement (rolling an odd number) is also 50%, because the two possibilities cover every outcome and must add to 1.

Worked example 2: two coins (and / or)

Flip two fair coins. Each has a 0.5 chance of heads. The probability that both land heads is P(A and B) = 0.5 × 0.5 = 0.25 (25%). The probability that at least one is heads is P(A or B) = 0.5 + 0.5 − 0.25 = 0.75 (75%). Notice we subtracted the 0.25 overlap; without that subtraction you would wrongly get 100%, which would mean heads is guaranteed. It is not: there is a 25% chance both coins come up tails.

Worked example 3: at least once over n tries

Suppose a game has a 10% (p = 0.1) drop chance each time you play. Over 10 plays, your intuition might say 100%, but that is wrong. The chance of never getting the drop is 0.9¹⁰ = 0.349, so the chance of getting it at least once is 1 − 0.349 = 0.651, about 65%. You would actually need around 22 plays to cross 90%, and rare events never truly reach 100% no matter how many tries you make.

At-least-once probability table: chance per try vs. number of tries

Read across a row for a fixed per-try chance and down a column for the number of attempts. Every cell is 1 − (1 − p)ⁿ, the probability the event happens at least once, computed and rounded to one decimal.

Notice how a 1% per-try chance is still only about 22% after 25 tries, while a 50% chance is already near-certain by 5 tries. This is the "at least once" pattern that also answers or questions across several equal events.

Quick reference table

Common single-event probabilities, shown three ways:

Who this calculator is for

• Students checking homework on basic probability, the and/or rules, and the complement.
• Gamers and hobbyists working out drop rates, loot odds, or the chance of a result over many attempts.
• Anyone making decisions under uncertainty: estimating the chance that at least one of several things goes wrong (or right).
• Teachers who want a clean way to show the same probability as a fraction, decimal, and percent.

Key probability terms

• Outcome: a single possible result, such as rolling a 4.
• Sample space: the full set of possible outcomes, such as {1, 2, 3, 4, 5, 6} for a die.
• Favorable outcome: an outcome that counts as a success for the event you are measuring.
• Independent events: events where one result has no effect on the other.
• Complement: the event "this does not happen," with probability 1 − P.
• Conditional probability: the chance of one event given that another has happened, P(A given B) = P(A and B) ÷ P(B).
• Mutually exclusive: events that cannot both occur at once; for those, P(A or B) = P(A) + P(B) with nothing to subtract.

Independent vs. dependent events

The simple product P(A) × P(B) assumes independence. That holds for coins, dice, and spinners. It breaks down when one event changes the next, for example drawing two cards without replacing the first. There, the probability of the second draw depends on what you drew first, so you need conditional probability, P(B given A). For two aces in a row the chance is 4/52 × 3/51 = 1/221, about 0.45%, slightly less than the 1/169 (0.59%) you would get by wrongly treating the draws as independent. Choose Dependent in the two-event mode, or load the "Two aces, no replacement" example, to get this directly.

Why "at least once" surprises people

People routinely overestimate how fast a rare event becomes near-certain. A 1% chance per attempt does not reach 50% until about 69 attempts, and never quite hits 100%. The complement rule, 1 − (1 − p)ⁿ, is the cleanest way to see this. Whenever a question contains the phrase "at least one," reach for the complement rather than trying to add up every individual case; it is faster and far less error-prone.

Tips for getting it right

• Always sanity-check that your answer sits between 0% and 100%.
• For "and" think "smaller": combining requirements makes an event rarer, so P(A and B) is never larger than either piece.
• For "or" think "bigger": P(A or B) is at least as large as the larger of the two.
• When multiplying by hand, convert percentages to decimals (divide by 100) first; the calculator accepts 25% as typed.
• Use the complement for any "at least one" question.

Related concepts and tools

Probability sits next to several other math topics. For exactly k successes in n trials (like "exactly 3 heads in 5 flips"), you need the binomial distribution, which builds on the counting methods in our Permutation & Combination Calculator. To turn a probability into a percentage of a quantity, the Percentage Calculator helps; to simplify the fraction form of a result, the Fraction Calculator does the reducing; and powers like (1 − p)ⁿ are exactly what the Exponent Calculator evaluates. If you want to summarise the spread of many outcomes rather than the chance of one, the Standard Deviation Calculator measures how far results stray from the average. This page focuses on the everyday rules (single, combined, and "at least once") that cover the large majority of probability questions.

Theoretical vs. experimental probability

There are two ways to arrive at a probability. Theoretical probability is what you get from reasoning about a fair setup: a coin "should" land heads half the time, so P = 1/2. That is what this calculator computes. Experimental (empirical) probability comes from actually running trials and dividing the number of successes by the number of attempts. If you flip a real coin 100 times and get 54 heads, the experimental probability is 0.54. The two converge as the number of trials grows, a result known as the law of large numbers. Small samples can stray a long way from the theoretical value, which is exactly why a coin landing heads five times in a row does not mean the coin is "due" for tails. Each flip is still 50/50; the past results carry no memory.

This distinction matters when you read claims in the real world. A basketball player's free-throw percentage is counted from shots actually taken (experimental), while the chance of drawing a red card from a full deck (26/52 = 50%) is pure theory. When a stated probability does not match what you observe over many repetitions, the underlying assumption (usually "fair" or "equally likely") is what is wrong, not the arithmetic.

Worked example: estimated probability from a frequency table

Suppose a number cube is rolled 250 times with the tally below. The estimated (experimental) probability of each outcome is its count ÷ 250. To find the estimated probability of an even number, add the counts for 2, 4 and 6 and divide by 250.

The even faces total 58 + 36 + 57 = 151 rolls, so the estimated probability of an even number is 151/250 = 0.604 = 60.4%. The odd faces make up the remaining 99/250 = 39.6%. This is experimental probability: with a truly fair cube the theoretical value is exactly 50%. A gap this large is not ordinary noise, though: a fair cube lands 26 or more rolls away from the expected 125 even rolls in only about 0.12% of 250-roll experiments (roughly 1 in 800), so a real tally like this would be good evidence that the cube is not fair.

The gambler's fallacy and independence

The single most common probability error in everyday life is the gambler's fallacy: believing that a run of one outcome makes the opposite outcome more likely. After five reds on a roulette wheel, red and black are still equally likely on the next spin, because the spins are independent. The wheel has no memory. The "at least once" formula in this tool quietly relies on the same independence assumption: each trial is treated as a fresh event with the same probability p. If your trials are not independent (cards drawn without replacement, or a machine that wears out and changes its failure rate), the simple powers of (1 − p) no longer hold, and you have to track how the probability shifts from one trial to the next.

The flip side is just as important. A long run of the same result does not prove the events are dependent or the setup is rigged; streaks are completely normal in random data. Getting ten heads in a row has a probability of (1/2)¹⁰ = 1/1024, which is small for a specific person on a specific evening but is expected to happen now and then across thousands of people flipping coins. Rare-but-not-impossible is the default state of randomness.

Combining more than two events

The "and" and "or" rules extend naturally to three or more independent events. For "all of them happen," multiply every probability: P(A and B and C) = P(A) × P(B) × P(C). For "at least one of them happens," the cleanest route is again the complement: work out the chance that none happen and subtract from 1: P(at least one) = 1 − (1 − P(A)) × (1 − P(B)) × (1 − P(C)). Notice that when all the probabilities are equal to p and there are n of them, this collapses straight into the familiar 1 − (1 − p)ⁿ formula. That is why the "repeated trials" mode is really just the multi-event "or" rule in disguise: rolling at least one six in four dice throws is the same calculation as "six on throw 1 or throw 2 or throw 3 or throw 4."

Worked example 4: the birthday-style surprise

Imagine a small lottery ticket with a 1 in 1,000 (p = 0.001) chance of winning, and you buy 500 tickets over a year. It feels like you should be roughly halfway to a guaranteed win, but the complement rule tells the real story. The chance of never winning is 0.999⁵⁰⁰ ≈ 0.606, so the chance of winning at least once is only about 1 − 0.606 = 0.394, or 39%. To get past a coin-flip 50% chance you would need about 693 tickets. This same math explains why "one in a million" events still happen to someone every day once you multiply by a large enough population, and why repeated small risks add up faster than intuition suggests.

How to read a probability as a real-world frequency

Decimals and percentages can feel abstract, so it often helps to restate a probability as a frequency. A probability of 0.02 is "about 1 in 50," 0.001 is "about 1 in 1,000," and 0.0001 is "about 1 in 10,000." To convert any decimal to a "1 in N" form, divide 1 by the probability: 1 ÷ 0.04 = 25, so 0.04 is roughly 1 in 25. Many people find a natural frequency ("4 out of 100 patients") easier to picture than a bare percentage ("4%"), even though the two are identical. When you read off a result from this calculator, try both framings and use the one that makes the size of the chance click for you.

Convert a probability to percent, odds and "1 in N"

The same probability written five ways. Odds in favor compare favorable to unfavorable outcomes (reduced), and 1 in N is 1 divided by the decimal, handy for turning a fraction or an odds statement into a percentage.

Frequently confused: probability vs. odds vs. expected value

Three ideas get mixed up constantly, so it is worth keeping them separate. Probability is favorable outcomes over total outcomes, always between 0 and 1. Odds compare favorable to unfavorable: a 1/4 probability is "1 to 3" odds (one win for every three losses), which is why a result that "sounds" like 1:4 is actually a probability of 1/5. Expected value takes it one step further by weighting each outcome by its payoff: if a $1 bet pays $5 on a 1/6 event, the expected value is (1/6 × $5) + (5/6 × −$1) = $0.83 − $0.83 = $0, a fair bet. Most casino and lottery games carry a negative expected value by design, which is the mathematical reason the house wins over time even though any single player can get lucky. This calculator gives you the probability; converting that to odds or feeding it into an expected-value sum is the natural next step for any decision involving stakes.

Sources and further reading

• Common Core State Standards for Mathematics, Conditional Probability and the Rules of Probability: independence (HSS-CP.A.2), conditional probability as P(A and B)/P(B) (HSS-CP.A.3), the addition rule (HSS-CP.B.7) and the general multiplication rule P(A and B) = P(A)P(B|A) (HSS-CP.B.8).
• Common Core State Standards for Mathematics, Grade 7 Statistics and Probability: probability as a number between 0 and 1 (7.SP.C.5), estimating probability from long-run relative frequency (7.SP.C.6) and compound events (7.SP.C.8).
• NIST/SEMATECH e-Handbook of Statistical Methods, §1.3.6 Probability Distributions.
• Standard results in elementary probability: the multiplication rule for independent events, the addition rule with the inclusion-exclusion correction, and the law of large numbers.

This calculator is an educational tool. It assumes events are independent and equally likely unless stated otherwise, and it does not constitute financial or gambling advice.