Dice Rolling and Probability in D&D
Dungeons & Dragons runs on a surprisingly elegant mathematical engine: a set of polyhedral dice whose results, modified by character statistics, determine whether swords connect, spells land, and lies succeed. Probability isn't just a background mechanic — it shapes every tactical decision at the table. Understanding how dice math actually works makes both players and Dungeon Masters sharper, more intentional, and occasionally less surprised when a paladin fumbles a persuasion check.
Definition and scope
The core randomizer in D&D 5th Edition is the d20 — a 20-sided die that produces a flat, uniform distribution across integers 1 through 20. Every number has exactly a 5% probability of appearing on any given roll. That flatness is the defining feature of the system and the source of its drama: a seasoned fighter and a first-level apprentice share the same 5% chance of rolling a natural 20.
The full dice set — d4, d6, d8, d10, d12, and d20 — each serves a distinct mechanical purpose. Damage rolls use smaller dice (a shortsword deals 1d6 piercing damage; a greataxe deals 1d12). Hit point pools at character creation often involve rolling your class's Hit Die, as defined in the fifth edition Player's Handbook published by Wizards of the Coast. The d100, simulated by rolling two d10s, appears mainly in random tables and wild magic surges.
Modifiers shift the effective probability range without changing the distribution shape. Adding a +5 bonus to a d20 roll shifts the outcome window from 1–20 to 6–25, but the spacing between outcomes stays identical. This matters enormously when a Dungeon Master sets a Difficulty Class — see Saving Throws and Skill Checks for how DC thresholds interact with modifier stacks.
How it works
The standard resolution mechanic is: roll d20, add the relevant modifier, compare to a target number. Beat or meet the target — success. Fall short — failure. The elegance is in that simplicity; the complexity lives in how modifiers accumulate.
Advantage and Disadvantage are the system's most impactful probability tools, introduced in 5th Edition as a clean replacement for situational +2 or −2 modifiers. Rolling with Advantage means rolling two d20s and taking the higher result. Disadvantage means taking the lower.
The probability shift this creates is substantial:
- Straight roll: 55% chance to meet or beat DC 10 (rolling 10 or higher on a d20).
- Advantage: approximately 79.75% chance to meet or beat DC 10 — a swing of nearly 25 percentage points.
- Disadvantage: approximately 30.25% chance to meet or beat DC 10 — a drop of nearly 25 points in the other direction.
These figures follow directly from basic probability: the chance of failing both dice on Advantage equals 0.45 × 0.45 = 0.2025, so success probability is 1 − 0.2025 = 79.75%. The math is verifiable with any dice probability calculator.
Multiple sources of Advantage do not stack — a rule the Player's Handbook states explicitly. A character with three separate sources of Advantage still rolls only two dice. This cap on Advantage is a deliberate design constraint that keeps the system from inflating beyond its probability bounds.
Common scenarios
The d20 roll appears across the four major action categories in D&D 5e:
- Attack rolls: Roll d20, add attack bonus (proficiency + ability modifier), compare to the target's Armor Class. A natural 20 is a critical hit regardless of AC — damage dice are doubled.
- Ability checks: Roll d20, add the ability modifier (and proficiency if the character has it), compare to a DM-set DC. DCs range from 5 (Very Easy) to 30 (Nearly Impossible) per the Dungeon Master's Guide.
- Saving throws: Roll d20, add saving throw bonus, compare to the spell or effect's save DC. A rogue's Evasion feature illustrates a hard boundary — on a successful Dexterity save against a damaging effect, they take zero damage instead of half.
- Death saving throws: Three successes before three failures determines survival; a natural 20 immediately stabilizes the character with 1 hit point.
Damage rolls, by contrast, use non-d20 dice and produce bell-curve-like distributions when multiple dice are combined. Rolling 2d6 (the damage for a greatsword) produces results from 2 to 12, but the middle values — 6, 7, 8 — appear far more often than the extremes. The most probable result of 2d6 is 7, appearing 6 out of 36 possible combinations (approximately 16.7%).
Decision boundaries
Where probability becomes genuinely strategic is at the threshold of meaningful outcomes. A character with a +8 bonus to Athletics faces a DC 15 climb check with a 70% success rate — high enough to attempt without much hesitation. That same character attempting DC 25 has only a 20% success. The decision to try, to seek Advantage, or to find another path changes based on that math.
The Ability Scores and Modifiers system is the upstream input: ability scores from 1 to 20 translate to modifiers from −5 to +5, which directly sets the floor and ceiling of what a character can reliably accomplish. A wizard with Strength 8 (modifier −1) has a 30% chance of meeting DC 10 on a Strength check — barely better than even odds, and well below a fighter with Strength 20 (+5 modifier) who hits DC 10 on a 5 or higher (80% success).
The relationship between modifier magnitude and Advantage is asymmetric at the extremes. Advantage helps most at the midpoints of the probability curve (around DC 11 on a straight roll) and helps least when success or failure is already nearly certain. A character who needs only a 2 to succeed gains almost nothing from Advantage; one who needs a 20 gains only a 9.75% improvement. This asymmetry is what makes stacking modifiers more reliable than stacking situational bonuses for consistently high-DC tasks — a detail that becomes central to character creation basics and long-term build planning.
For a broader orientation to the game's interlocking systems, the D&D Authority home provides structured entry points into every major topic area.