D&D 5e Point Buy vs Rolling for Ability Scores: Which Method Should You Use?

The Two Methods, Side by Side

D&D 5e offers two primary ways to generate a character's six ability scores. Point buy allocates a fixed 27-point budget across all six abilities using a defined cost table, guaranteeing every character at the table starts from an equally strong array. Rolling — almost always the 4d6-drop-lowest method, where you roll four six-sided dice and keep the highest three — introduces genuine randomness, capable of producing both stronger-than-point-buy arrays and weaker ones. Neither method is objectively "correct"; they optimize for different things, and the right choice depends on what your table values.

Point Buy Mechanics

Every ability starts at a base of 8. Raising a score costs points according to a fixed table, and the cost is not linear — the last two increments, from 13 to 14 and 14 to 15, each cost two points instead of one, which is why a single 15 consumes almost a third of the entire 27-point budget.

Rolling Mechanics: 4d6 Drop Lowest

For each of the six ability scores, roll four six-sided dice, discard the single lowest result, and sum the remaining three. Repeated across six abilities, this produces a full array with genuine variance — some characters will roll well above average, some will roll below it, and no two rolled arrays are guaranteed to be comparable in strength.

The exact probability distribution for a single 4d6-drop-lowest score can be computed by enumerating all 6^4 = 1,296 possible ordered dice outcomes. Doing so gives an exact mean of approximately 12.245 per individual score, and therefore an expected total of about 73.47 across six independent scores.

How Often Does Rolling Beat an Efficient Point-Buy Array?

This is the question point buy vs. rolling debates usually come down to in practice. An efficient point-buy spread (something like 15, 13, 13, 12, 10, 8, spending the full 27 points across a reasonably even distribution) produces a raw total in the low-to-mid 70s once you account for typical allocation choices. Because the average rolled total (73.47) sits close to that same range, a "typical" roll and a "typical" efficient point-buy array are, on average, roughly comparable in raw strength — the real difference is variance, not the central tendency.

The Full Percentile Breakdown of Rolled Totals

The mean (73.47) only tells part of the story, since it says nothing about how spread out actual results are. Computing the full six-score sum distribution by exhaustively combining all 1,296 possible outcomes per die roll across all six abilities gives an exact percentile breakdown, showing precisely how likely a given total actually is.

This percentile view is the clearest way to answer "how much am I actually risking by rolling instead of using point buy?" Half of all rolled arrays land at a total of 74 or below — essentially indistinguishable from a well-spent point-buy array's raw total — while only about one roll in ten reaches a total of 82 or higher, and only about one in a hundred reaches the kind of standout 89-plus total that makes for a memorable "the dice were kind to me" story.

Pros and Cons Framework

Does the Method Matter More for Certain Classes?

The choice of method interacts differently with different class designs. Single Ability Dependent (SAD) classes — a Sorcerer relying almost entirely on Charisma, for example — need only one very high score to function well, which point buy handles efficiently since you can dump every other score to fund a single 15 plus a background bonus. Multiple Ability Dependent (MAD) classes — a Paladin needing Strength, Constitution, and Charisma all to be reasonably strong, for instance — are structurally harder to build well under point buy's fixed 27-point budget, since spreading points across three or four abilities without any one being weak is expensive. Rolling, with its chance of producing several naturally high scores in one array, can occasionally favor MAD builds in a way point buy's fixed budget cannot replicate, though this cuts both ways — a weak roll hurts a MAD build far more than a SAD one, since a MAD build has less room to compensate by dumping unused stats.

Hybrid and Middle-Ground Approaches

Many tables use approaches that split the difference rather than picking a pure method. The "standard array" (15, 14, 13, 12, 10, 8) is essentially a pre-computed, maximally efficient point-buy allocation offered as a zero-math shortcut — it spends all 27 points and produces the identical result a careful point-buy allocation would reach on its own. "Roll and choose the better of two arrays," or "roll but allow swapping to the standard array if the roll is unusually weak," are common house rules that preserve some of rolling's excitement while capping the downside risk that pure rolling carries.

Why 4d6-Drop-Lowest Became the Standard Rolling Method

Earlier editions of the game commonly used a flat 3d6 method — roll three six-sided dice, sum them directly, no dropping — which produces a lower average (10.5 per score) and a narrower, more centrally clustered distribution than 4d6-drop-lowest. The shift to dropping the lowest of four dice was a deliberate design choice to push the average score upward (from 10.5 to roughly 12.24) and widen the upper tail of the distribution, supporting the "heroic" power level 5e characters are generally designed around, where player characters are meant to feel more capable than an average person from the very first session. Point buy's fixed 27-point budget was calibrated separately to land in a comparable overall power range to a typical 4d6-drop-lowest roll, which is exactly why the two methods' central tendencies (73.47 rolled average vs. a realistic mid-70s efficient point-buy total) end up so close together despite using completely different generation mechanics.

Handling Reroll Requests and Table Etiquette

Because rolling has no guaranteed floor, Dungeon Masters running a rolled-stats table eventually face the question of what to do about a genuinely unplayable low roll. A common and reasonable house rule sets a minimum acceptable total or a minimum single highest score below which a full reroll is automatically permitted — for example, requiring at least one score of 15 or higher, or a six-score total above a set floor — which preserves rolling's excitement while removing the rare truly unplayable outcome from the table. Whatever threshold you choose, setting and announcing it before anyone rolls avoids the awkward situation of adjudicating a borderline case in the moment, which can feel arbitrary or unfair to the player involved even when the ruling itself is reasonable.

It is worth deciding this policy the same way for the whole table rather than case by case, for the same fairness reason point buy exists in the first place — a reroll threshold applied inconsistently between players reintroduces exactly the kind of perceived unfairness that a table-wide, announced-in-advance policy avoids.

Worked Example: A Real Point-Buy Array vs. a Real Rolled Array

Consider a point-buy allocation of Strength 15, Dexterity 14, Constitution 13, Intelligence 8, Wisdom 10, Charisma 8 — spending exactly 9 + 7 + 5 + 0 + 2 + 0 = 23 of the 27-point budget, with 4 points left over (enough to raise Wisdom from 10 to 12, which costs 2 more points, with 2 still spare for a further small increase). Compare that against a sample rolled array of 17, 17, 16, 15, 15, 9 (a genuinely strong roll, well above the 73.47 average total). The rolled array's raw total of 89 exceeds even a fully-spent point-buy array's realistic total in the mid-to-high 70s, illustrating exactly the kind of outlier a rolled array can (rarely) produce — and exactly why point buy exists as a way to avoid the possibility of the opposite outcome.

Guidance for Dungeon Masters Choosing a Table-Wide Method

• For competitive or tournament-style play where balance matters most, point buy (or the equivalent standard array) removes a fairness variable entirely.
• For narrative-driven home games where players enjoy the surprise and story of a random array, rolling adds genuine excitement many tables value highly.
• If mixing new and experienced players, point buy avoids a situation where a new player unluckily rolls a weak array and feels mechanically behind from session one.
• Consider a hybrid rule (roll but allow falling back to the standard array) if you want some of rolling's excitement without its full downside risk.
• Whatever method you choose, apply it consistently across the whole table — mixing methods between players is the most common source of perceived unfairness, more than the method itself.

There Is No Universally Correct Answer

Framing this as a search for the "better" method misses the actual decision. Point buy and rolling are solving for different things — guaranteed fairness versus genuine chance — and a table that values one over the other is not making a mistake by choosing it. What matters more than which method you pick is applying it consistently, understanding the real numbers behind it rather than relying on vague impressions of "rolling is stronger" or "point buy is weaker," and setting clear expectations (including reroll policy, if rolling) before character creation begins.

If your table is genuinely undecided, a reasonable default is to try point buy for your first campaign together, since it removes one more variable while everyone is still learning the system, and revisit the question once the group has enough shared experience to know whether they specifically miss the excitement rolling can bring.

Sources to Verify or Cite

• D&D Beyond 2024 Basic Rules, character creation: https://www.dndbeyond.com/sources/dnd/br-2024/creating-a-character
• D&D Beyond 2024 Basic Rules, ability scores and modifiers: https://www.dndbeyond.com/sources/dnd/br-2024/playing-the-game
• The 4d6-drop-lowest probability figures in this article (mean ~12.245 per score, ~73.47 expected six-score total) come from exact enumeration of all 6^4 = 1,296 ordered four-die outcomes.
• The standard 27-point cost table is unchanged between the 2014 and 2024 Player's Handbooks; ability bonuses come from species in 2014 rules and from background in 2024 rules.
• Use your campaign's chosen rulebook and Dungeon Master rulings for table-specific character creation decisions.

Point Buy vs Rolling FAQ