numvana

Binomial Probability Calculator

Find the exact, at-least, and at-most probabilities for a fixed number of successes across repeated independent trials, plus the expected value and standard deviation.

P(exactly 5 successes)
24.6094%
P(at least 5 successes)
62.3047%
P(at most 5 successes)
62.3047%
Expected number of successes
5
Standard deviation
1.581

How it works

The binomial distribution gives the probability of getting exactly k successes in n independent trials that each succeed with the same probability p: P(X = k) = C(n,k) × pᵏ × (1−p)ⁿ⁻ᵏ, where C(n,k) is the number of ways to choose which k of the n trials succeed. This calculator evaluates that formula in log space (summing log-factorials rather than multiplying raw factorials) so it stays numerically accurate even for a large number of trials, rather than overflowing.

"At least k" and "at most k" sum that exact-probability formula across every outcome from k to n, or from 0 to k respectively — the standard cumulative binomial probability. The expected number of successes is n × p, and the standard deviation is √(n × p × (1−p)), both closed-form properties of the distribution rather than separate calculations.

This applies to any fixed-probability repeated trial with only two outcomes per trial — coin flips, pass/fail quality checks, free-throw attempts — as long as each trial is independent and shares the same success probability. It's a different question from a dice-sum calculator, which convolves multiple dice faces rather than counting successes across repeated identical trials.

FAQ

What counts as a valid 'trial' for this calculator?

Any repeated event with exactly two outcomes (success/failure) where every trial has the same success probability and trials don't influence each other — like independent coin flips or manufacturing units off the same production line. If the probability changes trial to trial, or trials aren't independent, the binomial model no longer applies exactly.

Why do the exact and at-least/at-most probabilities use log-space math internally?

For a large number of trials, C(n,k) and pᵏ can individually become far too large or far too small for standard floating-point numbers to represent, even though the final probability is a normal, well-behaved value. Working in log space avoids that overflow/underflow before converting back to a plain probability at the end.

Related calculators