Dice Probability Calculator

Background.

The Quanta dice probability calculator computes the exact probability of a target sum across any number of dice with any number of sides — the specific, high-volume question behind searches like 'probability of rolling a 7 with two dice' or 'chance of getting at least 10 on 3d6'. Enter the defaults, two standard six-sided dice and a target sum of exactly 7, and the calculator returns a probability of 1/6 (about 16.67%), the single most quoted fact in introductory probability, alongside the exact count of favorable outcomes (6 out of 36 total rolls) and the mean and standard deviation of the sum.

The two-dice sum problem is not a modern textbook invention — it is the literal origin of formal probability theory. In 1654, the French gambler Chevalier de Méré asked Blaise Pascal a related dice question (about the fair division of stakes in an interrupted game), and the resulting correspondence between Pascal and Pierre de Fermat is generally credited as the founding moment of the mathematical theory of probability. Christiaan Huygens formalized the treatment shortly after in 1657. The reason dice problems drove the birth of the field is instructive: with a sample space small enough to enumerate completely by hand (36 outcomes for two six-sided dice, 216 for three), dice are the simplest possible laboratory for testing whether a probability theory's predictions match reality, which is exactly why every statistics course still opens with them centuries later.

The calculator supports three distinct questions because 'the probability of rolling a 7' is genuinely ambiguous without specifying which one you mean. 'Exactly' asks for the probability of that one specific sum and nothing else — the question behind almost every single-roll tabletop-game rule ('you need to roll exactly a 7 to hit'). 'At least' asks for the probability of that sum or higher — the question behind save-or-suffer mechanics ('you need 10 or higher to succeed') and behind casino craps, where a 'pass line' bet on the come-out roll wins outright on a natural 7 or 11 (an at-least-type union, not a single exact value). 'At most' asks for the sum or lower — useful for 'bust' mechanics where exceeding a threshold is bad. Confusing these three is one of the most common real-world dice-probability mistakes, and this calculator's explicit mode selector exists specifically to prevent it.

The underlying combinatorics scales quickly. With two six-sided dice there are only 36 equally likely outcomes, small enough to enumerate on a single sheet of paper; the sum of 7 is the most likely single value precisely because more ordered pairs of faces (six of them) add up to 7 than to any other sum, forming the familiar triangular distribution that peaks in the middle and tapers toward the extremes 2 and 12, which can each only be made one way (1+1 or 6+6). Adding a third die multiplies the sample space to 216 outcomes and reshapes the distribution into a taller, more sharply peaked curve centered on 10 and 11 — a visual preview of the central limit theorem, which guarantees that the sum of enough independent dice approaches a bell-shaped normal distribution no matter how simple each individual die is.

This calculator handles the general case — any n from 1 to 50 dice, any s from 2 to 100 sides — using an efficient dynamic-programming convolution rather than brute-force enumeration, so it stays fast even for combinations (say, 20 ten-sided dice) where literal enumeration of every outcome would be computationally impossible.

What is dice probability calculator?

For n dice, each with s equally likely faces numbered 1 through s, the total number of possible outcomes is sⁿ, since each die's face is chosen independently of every other die's (Feller, 1968, An Introduction to Probability Theory and Its Applications, Vol. 1, 3rd ed.). The probability of any specific event — a target sum being hit exactly, met or exceeded, or met or undershot — is the count of outcomes satisfying that event divided by sⁿ, following the classical (equally-likely-outcomes) definition of probability. The distribution of the SUM across n dice is not itself uniform, even though each individual die is: it is the n-fold discrete convolution of a uniform distribution on {1, ..., s}, which produces a distribution shaped like a widening, increasingly bell-like curve as n grows — narrow and triangular for two dice, progressively smoother and more symmetric for larger n, converging toward a normal distribution by the central limit theorem. The mean of that sum distribution is n(s+1)/2, since each individual die has mean (s+1)/2 and expectations add across independent dice; the variance is n(s²−1)/12, since each die's variance is (s²−1)/12 and variances also add across independent random variables, giving a standard deviation of √(n(s²−1)/12). 'Exactly k', 'at least k', and 'at most k' are three genuinely distinct questions about the same underlying sum distribution — the count of favorable outcomes for 'at least' and 'at most' is a cumulative sum across a range of possible totals, while 'exactly' isolates a single value.

How to use this calculator.

1. 01Enter the number of dice (n) you are rolling together.
2. 02Enter the number of sides (s) on each die — 6 for standard dice, 20 for a d20, or any value from 2 to 100.
3. 03Enter your target sum. It must fall between n (if every die shows its minimum face, 1) and n×s (if every die shows its maximum face, s) — anything outside that range is impossible and the calculator will tell you so.
4. 04Choose your mode: 'exactly' for that one specific sum only, 'at least' for that sum or higher, or 'at most' for that sum or lower. These are three different questions, and mixing them up is the single most common dice-probability mistake.
5. 05Read the primary result, the probability (and its percentage form), plus the exact favorable-outcomes count and the full size of the sample space (sⁿ) underneath it.
6. 06Use the mean sum and standard deviation, shown regardless of your chosen mode, as a quick sanity check — a target sum far from the mean, in units of standard deviation, should correspond to a small probability, and vice versa.

The formula.

Every outcome of rolling n dice, each with s sides, is an ordered n-tuple of face values, and because each die is independent and equally likely to show any of its s faces, all sⁿ such tuples are equally likely (Feller, 1968; Ross, 2019, A First Course in Probability, 10th ed., Ch. 2). The probability of any event is therefore just the count of tuples satisfying it, divided by sⁿ — the classical definition of probability applied directly.

Counting how many of those sⁿ tuples sum to a particular value k is the combinatorial heart of the calculator. For a single die (n=1) exactly one face gives any particular sum. For two or more dice, the count is found by convolution: the number of ways n dice sum to k equals the sum, over every possible value v the last die could show (1 through s), of the number of ways the remaining n−1 dice sum to k−v. Applied recursively from n dice down to a base case of 0 dice (which trivially has exactly one way to sum to 0), this convolution builds up the full distribution of the sum for any n and s. This calculator computes it with a prefix-sum optimization that turns each die's convolution step into a single linear pass rather than a full nested loop, keeping the computation fast even at the tool's upper bound of 50 dice with 100 sides each.

'Exactly k' reads a single entry from that computed distribution. 'At least k' sums every entry from k up to the maximum possible sum n×s. 'At most k' sums every entry from the minimum possible sum n down to k. Because the full distribution always sums to sⁿ (every outcome falls somewhere), P(at least k) + P(at most k−1) always equals exactly 1, a useful sanity check available whenever you compute both.

The mean and standard deviation of the sum follow from the additivity of expectation and variance across independent random variables. A single fair s-sided die has mean (s+1)/2 (the average of 1 through s) and variance (s²−1)/12 (a standard result for a discrete uniform distribution). Summing n independent, identically distributed dice adds their means directly, giving n(s+1)/2, and adds their variances directly (independence is what licenses adding variances rather than needing covariance terms), giving n(s²−1)/12 and a standard deviation equal to its square root. These two summary values are mode-independent — they describe the underlying sum distribution itself, not any particular exactly/at-least/at-most question about it — which is why they are always shown alongside whichever specific probability you asked for.

A worked example.

Frequently asked questions.

References& sources.

1. [1]Feller, W. (1968). An Introduction to Probability Theory and Its Applications, Volume 1, 3rd edition. New York: John Wiley & Sons.
2. [2]Ross, S.M. (2019). A First Course in Probability, 10th edition, Ch. 2. New York: Pearson.
3. [3]NIST/SEMATECH e-Handbook of Statistical Methods.