How Many Outcomes Are There When Rolling Two Dice?
The sample space is the complete set of outcomes under a stated experiment; for two distinguishable fair d6 dice it contains 36 ordered pairs.
A worked example with numbers
Make a grid with first-die values 1 through 6 down the side and second-die values 1 through 6 along the top. Every cell is one ordered pair. The cell in row 2 and column 5 is (2,5); row 5 and column 2 is (5,2). They are two cells even though both yield total 7. The probability of one named pair such as (2,5) is 1/36, assuming each die has six equally likely faces and their results are independent.
How to set the rule before the result
For dice with s sides each, two ordered dice have s² outcomes: two d8 dice have 8 × 8 = 64, and two d20 dice have 20 × 20 = 400. For r dice with s faces each, the ordered count is s^r; three d6 dice have 6³ = 216 face triples. Use the two-dice tool for a concrete pair and the custom roller when the face count differs from six.
Common mistakes that change the odds or the process
Do not say two dice have 21 outcomes merely because unordered pairs such as {2,5} can be grouped. Grouping is legitimate only if you retain that doubles have one ordering and non-doubles have two; it is not a list of equally likely groups. Do not replace 36 with 11 when calculating a total. Do not assume s² applies when dice have different sizes: a d6 and d8 have 6 × 8 = 48 ordered outcomes.
Where this method stops being appropriate
The multiplication count applies before modifiers, rerolls, exploding dice, advantage rules, or a rule that discards one die. Those game rules create a transformed outcome space. Physical dice may also fail the fair-face assumption. The count describes the model needed for a probability calculation; it does not verify equipment or determine whether a game’s stated procedure is acceptable.
How the random source fits into the rule
Wolfram MathWorld’s dice reference documents distributions for standard dice and illustrates why face combinations must be counted before totals. Strong browser integers described by MDN’s Crypto.getRandomValues reference can make individual digital faces unpredictable, but 36 follows from the declared 6-by-6 sample space. State that space before interpreting a frequency or a total.
Build the grid rather than guessing
Draw six rows labelled 1–6 for die A and six columns labelled 1–6 for die B. Each intersection represents one outcome and each row contains six possibilities. The grid therefore has 36 cells before any totals are written. This visual method scales: an 8-row by 10-column grid for a d8 and d10 has 80 cells. It also exposes why a two-die event must be counted by pairs, not by the number of distinct totals.
Ordered and unordered descriptions serve different jobs
If only the sum matters, (1,6) and (6,1) can be grouped but their group has multiplicity two. Doubles such as (4,4) have multiplicity one. There are 21 unordered pairs from two d6 dice, but those 21 groups do not have equal probability. The 36 ordered pairs are the equally likely elementary outcomes, which is why they are the correct denominator for ordinary fair-dice probability.
Generalise to unequal dice
| Dice | Ordered outcomes | One named pair probability |
|---|---|---|
| d6 + d6 | 36 | 1/36 |
| d6 + d8 | 48 | 1/48 |
| d10 + d10 | 100 | 1/100 |
Multiply the face counts because each face of one die can occur with every face of the other.
Check the game’s actual output
A roller may show two dice but report a kept high die, a sum plus modifier, or a success total. Those outputs do not all use the same sample-space calculation. Read the rule at the point where it transforms the face pair. The 36 count remains the starting space for two d6 faces, while the reported value may have fewer categories and unequal probabilities.
Use the count as a probability denominator
Once the 36 elementary pairs are visible, every event becomes a count. “At least one die shows 6” contains 11 pairs: six pairs with die A equal to 6 plus five more with die B equal to 6 and die A below 6. Its probability is 11/36. Adding six and six without removing the shared (6,6) pair would double-count it.
This example shows why listing outcomes is stronger than a verbal shortcut. The same approach handles doubles, a particular face, a sum range, or a condition on both dice as long as the event is stated precisely.
Keep elementary outcomes visible
For a fair pair of dice, every one of the 36 cells has probability 1/36 before a game groups them. The cell model avoids accidental double counting and remains valid even if the later question concerns a sum, a double, a high face, or a condition on one named die.