Grade curving means adjusting a set of test scores upward using one of a handful of common methods — curve-to-target, flat/fixed-points, or square-root — and which method an instructor picks changes how much of a boost each student gets, sometimes dramatically. On the same set of scores, a square-root curve can lift a 58% into the mid-70s while barely touching a 92%, whereas a flat-points curve adds the exact same boost to everyone regardless of where they started.
Why curves exist
A curve is usually applied when a test's raw results come back lower than the instructor expected — maybe the questions were harder than intended, there wasn't enough time, or a concept wasn't covered as clearly in class as it should have been. Rather than leave the recorded grades where they landed, the instructor adjusts every score using a consistent rule. Whether to curve at all, and which method to use, is entirely at the instructor's discretion — there's no requirement or universal standard here.
The three common curving methods
Curve to target
This method finds the highest score in the set and adds enough points to every score to bring that highest score up to 100 (or another target, though 100 is the most common choice). Everyone gets the same flat point boost, calculated from the top performer.
New Score = Old Score + (100 − Highest Score)
If the top score in the class was a 92, every student gets a flat +8 points added to their score, capped at 100. It's simple, transparent, and guarantees at least one perfect score, which is likely why it's a commonly used approach.
Flat / fixed-points curve
Similar in mechanics to curve-to-target, but the instructor picks the number of bonus points directly instead of deriving it from the highest score. Everyone still gets the same flat boost — it's just a number the instructor sets rather than one calculated from the class results.
New Score = Old Score + Fixed Points (capped at 100)
Square-root curve
This method works completely differently: instead of adding a flat number of points, it multiplies the original score (as a decimal, i.e., a percentage divided by 100 and then re-multiplied by 100 inside the square root) by 100 and takes the square root, rounding to the nearest whole point.
New Score = round(√(Old Score × 100))
Because the square root function compresses larger numbers less than it expands smaller ones, this method boosts low scores by a lot more than it boosts high scores — a 100% stays at 100, but a 50% jumps to roughly 71%. It only makes sense applied to scores already expressed as a 0–100 percentage.
Worked example: the same five scores, three ways
Take five test scores — 58, 67, 72, 80, and 92 — and apply each method. For curve-to-target, the highest score (92) sets the boost at 100 − 92 = 8 points for everyone. For the flat curve, assume the instructor chose to add 5 points across the board.
| Original Score | Curve to Target (+8) | Flat Curve (+5) | Square-Root Curve |
|---|---|---|---|
| 58 | 66 | 63 | 76 |
| 67 | 75 | 72 | 82 |
| 72 | 80 | 77 | 85 |
| 80 | 88 | 85 | 89 |
| 92 | 100 | 97 | 96 |
| Class average | 81.8 | 78.8 | 85.6 |
The starting class average was (58 + 67 + 72 + 80 + 92) ÷ 5 = 73.8. All three methods raise it, but by different amounts and in very different patterns:
- Curve-to-target and the flat curve both add a constant amount to every score, just a different constant (8 points versus 5 points in this example). The gap between students stays exactly the same — the whole distribution just shifts upward.
- The square-root curve compresses the range instead of shifting it. The lowest score (58) gained 18 points, while the highest score (92) gained only 4. This narrows the spread between students significantly — the same five scores that spanned 34 points originally (58 to 92) span only 20 points after a square-root curve (76 to 96).
That last property is the real reason to pick a square-root curve over the other two: it's specifically useful when an instructor wants to give struggling students a bigger boost than students who were already doing well, rather than preserving the original ranking gaps.
Choosing between methods (as a student, not an instructor)
You generally won't get to choose how your own test is curved — that's the instructor's call. But understanding which method is in play tells you how to read a curved grade you're handed back:
- If your instructor mentions curving "everyone up to the top score," that's curve-to-target — expect the same point boost as the highest scorer in the class got.
- If they mention adding a specific number of points, that's a flat curve — the boost is fixed regardless of your original score.
- If lower scores in the class seem to have jumped disproportionately more than higher ones, a square-root curve (or something similar) is likely in play.
What curving doesn't change
None of these methods change how many points you actually earned on the exam itself — they're a post-hoc adjustment applied to the whole class, and results are capped at 100 regardless of the formula's raw output. A curve also isn't the same thing as grading on points relative to a total possible score, which happens before any curve is applied. And because curving is entirely at an instructor's discretion, none of the methods here should be treated as a guarantee of what will happen to your own grade — always confirm with your instructor's actual stated policy rather than assuming one of these formulas applies.
Try it yourself
Rather than recalculating a curve by hand for a whole class or score list, the Grade Curve Calculator applies any of these three methods to a full list of scores at once, showing the original score, curved score, and point boost for each entry side by side, along with the before-and-after class average.
