D&D Point Buy Calculator

About this tool

Point buy looks like simple addition and is not. The cost of a score is not proportional to the score: from 8 to 13 every step costs one point, and then the last two steps cost two each. A 15 therefore costs nine points, not seven, which is the single most common arithmetic error on a 5e character sheet — and one that is easy to miss because the character still works, just with fewer points spent than the player thinks.

The second thing the table makes clear is that 27 points cannot buy a good score in everything. Six scores at 15 would cost 54. Even the flattest legal build, six 13s, costs 30 and is over budget. Any 27-point character is therefore a set of deliberate compromises, and seeing the remaining points tick from positive to negative as a score is raised is the fastest way to understand which compromise a build is making.

The leftovers are not always spendable either. Every step from 8 to 13 costs one point, but 13 to 14 costs two. A build that ends with a single point left and every score at 13 or above cannot use it: there is no one-point step available. That is not a mistake in the arithmetic, it is the shape of the table, and it is why some builds finish on 26 or 25 and are still legal.

The point-buy cost table, and why it bends at 13

A score of 8 is free, 9 to 13 cost one point per step, and 14 and 15 cost two points per step. The bend at 13 is deliberate: it makes very high scores expensive so that a starting character cannot be good at everything.

Reading the table as a straight line is the error to avoid. Five steps from 8 to 13 cost five points. The step to 14 costs two, and the step to 15 costs two more, so a 15 is nine. If it were linear it would be seven, and every character built that way would be two points over budget without anything on the sheet looking wrong.

The bend does real work in the design. Without it, 27 points would buy three scores of 15 and change, which is exactly what it does buy with the bend in place — but it would also make 16 and 17 reachable, and those are meant to be racial-bonus territory. The price step is what keeps point buy inside the intended range.

One consequence worth memorising: the marginal value of the last two points is terrible. Turning a 13 into a 15 costs four points and buys a +1 modifier, the same +1 you would get from turning a 9 into an 11 for two points. That is not a bug, it is a cost curve, and it is why balanced spreads are competitive with specialised ones.

Why 27 points never fill six scores

Six scores at 15 would cost 54 points. Even six 13s costs 30. Every legal 27-point build therefore has at least one score at 8, and usually two or three.

The arithmetic is worth doing once. A 15 is nine points, so three of them is 27 and the other three scores have to sit at 8, where they cost nothing. Six 13s at five points each is 30, already over. Six 12s is 24, which leaves three points and lands you at 12/12/12/12/12/15 or similar.

This is the design working as intended. A character who is excellent at everything has no decisions to make and no weakness for the party to cover. The budget forces specialisation, and the most common shapes — three high scores with three dump stats, or a broad spread of 12s and 13s — are exactly the two ways of answering that pressure.

It also means the "best" build does not exist in the abstract. A 15/15/15/8/8/8 spread and a 13/13/13/12/12/12 spread both cost exactly 27 and produce completely different characters. The calculator reports the cost; which one is right depends on the class and the table.

The leftover point that cannot be spent

Because steps cost one point below 13 and two above it, a build that ends with one point left and every score at 13 or higher has nothing to buy. It is still legal — it just finishes under budget.

This surprises people who expect the total to always be exactly 27. The budget is a ceiling, not a target. Nothing in the rules requires you to spend every point, and some spreads genuinely cannot absorb the last one. A character at 13/13/13/13/13/13 has spent 30, so that particular shape is over; a character at 13/13/13/13/12/12 has spent 28.

The cheapest available step is what matters. If any score is at 12 or below, one point buys a step. If every score is at 13 or above, the next step costs two, and a single leftover point is stranded. The tool flags exactly this case rather than silently reporting "1 point left" as if it were spendable.

In practice a stranded point is a signal to reconsider the spread, not a problem to fix. Moving one point from a 13 to a 14 frees a point elsewhere and often produces a better character than leaving it unspent.

It gives the arithmetic, not a recommendation

Twenty-seven points can go anywhere: pile them into two scores, or spread them to 13.5 across six. This tool costs the build and stops there, because which spread is stronger depends on your party and your campaign, not on a table.

The point-buy cost table is rules data, not opinion: 8 costs 0, 9 costs 1, 10 costs 2, 11 costs 3, 12 costs 4, 13 costs 5, 14 costs 7, 15 costs 9. Each step past 13 gets more expensive, rising by two instead of one, which is the bend at thirteen.

Six scores of 15 would cost 54, twice the budget, so six high scores simply cannot be bought. Six scores of 8 cost nothing, which means the remaining 27 points must be spent — hence the whole notion of leftover points.

One point left over is normal: the smallest increment in the table is one point, so when the balance lands where nothing can be bought, the page says how many points are unspent rather than quietly dropping them or forcing them into a score.

Withholding a recommendation is deliberate. With the same 27 points, 15 / 15 / 14 / 10 / 8 / 8 and six 13s are both legal, and which is better depends on the checks you expect and who covers what in the party. That is a player judgement, not something this table can answer.