Grade Curve Calculator four curves, four answers
Enter a raw score and pick a curve. You get the curved percentage, the points gained, and what all four common curves would do to the same score.
Curving Is Not One Thing
The same 68 % becomes 73, 82.5, 73.9 or 74.8 depending on which method the instructor uses.
Your Instructor Decides
This shows what each method produces, not which one applies to your class.
What is a grade curve?
At a Glance
A grade curve is an adjustment applied to raw marks after an assessment, usually because the test turned out harder than intended. "Curving" is not a single procedure — at least four methods are in common use, and they can differ by more than ten percentage points on the same score.
Each method makes a different assumption about what went wrong with the assessment.
curved = raw + bump · curved = raw × factorcurved = 10 × √raw · curved = raw × 100 ÷ topThe square-root curve is the interesting one. Because the square root grows more slowly than its input, low scores gain far more than high ones: a 30 % becomes 54.8 %, a gain of 24.8 points, while an 88 % becomes 93.8 %, a gain of only 5.8. It also cannot push anyone above 100 %, which the other three can — so every result here is capped.
These are the values the calculator loads by default.
Convert to a percentage
68 ÷ 100 = 68 % raw.
Apply the square-root curve
10 × √68 = 10 × 8.246 = 82.46 %, a gain of 14.46 points.
Compare a flat 5-point bump
68 + 5 = 73 %. Simple, and it treats everyone identically.
Compare scaling to a 92 % top score
68 × 100 ÷ 92 = 73.91 %. This one depends entirely on how well the best student did.
See where the methods diverge
At 30 % raw, the square root gives 54.8 % while the flat bump gives 35 % — nearly twenty points apart. At 88 % they land within a point of each other.
The arithmetic is exact for each named method. What it cannot tell you is which method your instructor is using.
The choice matters most at the bottom. All four curves agree roughly at the top of the scale and diverge sharply at the bottom. If you are near a pass boundary, the method matters more to you than to anyone else in the class.
Scaling to the top score is contingent on one person. If the best paper scored 92, everyone divides by 0.92. If somebody scored 100, the curve does nothing at all. It is the least predictable of the four.
Nothing here models a bell curve. Some institutions curve to a target distribution — a fixed share of A grades — rather than by transforming each score. That is rank-based and cannot be computed from your score alone.
Curves can be applied per question, not per paper. Dropping a bad question or awarding it to everyone changes the raw score first; enter the adjusted raw score in that case.
Everything is capped at 100 %. Three of the four methods can push a high score past full marks, which almost no institution allows.
For the course grade the curved mark feeds into, use the Weighted Grade Average Calculator. To see what you need on a final, use the Final Grade Calculator.