Dice Probability Distributions: Totals, Averages, and Outcomes

A dice probability distribution describes how likely each possible result is under a stated mathematical model. A single fair d6 has six equally likely face outcomes. Rolling several dice and adding them produces a different distribution: middle totals can have more combinations than extreme totals. The rules that select, discard, reroll, or modify results matter just as much as the dice expression.

One die and several dice

For one fair d6, each face from 1 through 6 has probability 1/6. For 2d6 summed together, the total 2 has one combination, while 7 has six combinations: 1+6, 2+5, 3+4, 4+3, 5+2, and 6+1. The center is therefore more likely than either extreme. A list of possible totals without their probabilities is incomplete.

Average is not the most likely result

The mean or expected value summarizes the long-run average of a model. It does not necessarily identify the single most likely outcome. For 2d6, the expected total is 7 and 7 is also the most likely single total, but in other systems the mean may fall between integer results or the distribution may have several peaks.

Rules change the distribution

Keep-highest, keep-lowest, advantage, disadvantage, rerolls, exploding dice, success thresholds, and modifiers all create new models. A fixed modifier shifts totals, while selecting the highest of multiple dice changes the relative likelihood of outcomes. A probability explanation should show which operation is being modeled and in what order.

Dice pools that count successes are distributions over counts rather than sums. Symbol dice may require a table of outcomes or a joint probability model. Avoid comparing two systems solely by their number of dice when their interpretation rules differ.

Model versus physical object

A theoretical distribution assumes conditions such as equally likely faces and independent rolls. A physical die may depart from the model because of construction, wear, surface, rolling method, or measurement error. The model remains useful, but it should be labelled as theoretical. Observed frequencies should be reported separately with the number of rolls and the test conditions.

Comparing two systems

Use the same question when comparing systems. “What is the chance of at least one success?” is different from “What is the expected total?” and both differ from “How often does a result exceed a target?” State the metric, the dice expression, and the rule operations. A chart or table should identify its assumptions so readers can reproduce the calculation and see where two systems genuinely differ.

Common interpretation errors

• Confusing the average with the most frequent outcome.
• Adding probabilities for outcomes that are not mutually exclusive.
• Comparing a summed pool with a success-count pool as if they measured the same result.
• Calling a model empirical without recording physical observations.

How to present a distribution

1. Write the dice expression and interpretation rule.
2. List the possible outcomes and whether they are sums, successes, symbols, or table ranges.
3. Explain the assumptions, including fairness and independence where relevant.
4. Show combinations, probabilities, or a clearly labelled simulation method.
5. Separate theoretical results from measurements of a physical die.

Related DICEWIKI pages

Continue with dice probability basics, dice modifiers, dice pools, and percentile dice.

Sources and editorial notes

1. OpenStax Algebra and Trigonometry: Probability, for probability models, expected values, and the distinction between possible outcomes and their likelihoods.
2. D&D Beyond Basic Rules: Playing the Game, for attributed dice expressions and rule operations.
3. DICEWIKI editorial policy, for separating theoretical probability from physical-die observations.