The exact math behind the two most common grading curves, with worked examples for an entire class.
Published August 27, 2026 · Reviewed by the Ihsabha Editorial Team
"This exam will be curved" is one of the most common sentences in a classroom, and also one of the least precisely understood — students often treat it as a vague promise that things will somehow get better, without knowing the actual rule being applied to their score. In reality, curving usually follows one of a small number of well-defined mathematical methods. This guide covers the two most widely used — the square-root curve and the highest-score (bonus-point) curve — with the exact formula for each, full worked examples across a whole class, and how a curved score fits into your overall weighted course grade.
A grading curve is a single, consistent mathematical rule an instructor applies to every student's raw score at once, almost always because the class's raw average came out lower than the material or the students' preparation actually warranted — commonly because an exam turned out harder than intended. It is a correction applied to the whole class as a group, not a personal adjustment made to any one student's score in isolation, and it is never guaranteed by default. Some courses are never curved regardless of how an exam turns out; others are curved routinely as a stated course policy. Whether curving applies at all, and which method is used if it does, is set entirely at the instructor's discretion. You can find a fuller explanation of this specific point in How to Calculate Grade Percentage: Formula, Steps & Letter Grades. To save time, enter your values into the Test Curve Calculator and get an instant result.
The square-root curve is popular because of a useful mathematical property: square roots compress the distance between larger numbers while stretching the distance between smaller ones. In practice, this means struggling scores get boosted by more than already-strong scores.
Curved Score = 10 × √(Raw Percentage)
Take your raw percentage score, find its square root, and multiply the result by 10. A raw score of 64% has a square root of 8, so the curved score becomes 8 × 10 = 80% — a 16-point boost. A raw score of 81% has a square root of 9, giving a curved score of 90% — only a 9-point boost. A perfect 100% stays exactly 100%, since the square root of 100 is 10. This shrinking boost as scores rise is precisely why many instructors consider the square-root method fairer than a flat point addition: it helps the students who most needed help without inflating already-strong scores by very much.
The highest-score curve takes a more literal approach: it treats the top scorer in the class as having effectively earned a perfect 100%, and gives every other student in the class that exact same flat bonus.
Curved Score = Raw Score + (100 − Highest Score in Class), capped at 100
First, find how many percentage points below 100% the highest raw score in the class fell. If the top score was 90%, that's a 10-point gap. Add that same flat 10-point bonus to every student's own raw score, including the top scorer (whose new score is capped at 100% rather than exceeding it). Unlike the square-root method, this approach adds an identical number of points to every student regardless of where their score started — a student who scored 50% and a student who scored 85% both receive the same +10 points.
Suppose five students took an exam with the following raw scores, and the instructor decides to curve the results. The highest raw score in the class was 88%.
| Student | Raw Score | Square-Root Curve | Highest-Score Curve |
|---|---|---|---|
| Student A | 88% | 10×√88 = 93.8% | 88 + 12 = 100% |
| Student B | 72% | 10×√72 = 84.9% | 72 + 12 = 84% |
| Student C | 60% | 10×√60 = 77.5% | 60 + 12 = 72% |
| Student D | 45% | 10×√45 = 67.1% | 45 + 12 = 57% |
| Student E | 30% | 10×√30 = 54.8% | 30 + 12 = 42% |
Notice how the two methods diverge as raw scores drop. For Student A (the top scorer), the highest-score curve delivers a full 100%, while the square-root method gives a smaller boost to 93.8%. But for Student E, who started at just 30%, the square-root curve delivers a much larger relative rescue — pushing the score up to 54.8%, compared to only 42% under the flat +12-point highest-score method. This is the practical, visible effect of the square-root method's shrinking-boost property: it concentrates its help on the students who most needed it, at the cost of a smaller boost for the students who were already doing well.
| Square-Root Curve | Highest-Score Curve | |
|---|---|---|
| Boost for low raw scores | Larger | Smaller (same flat amount as everyone else) |
| Boost for high raw scores | Smaller | Same flat amount as everyone else |
| Complexity to explain to a class | Requires understanding square roots | Simple, transparent, easy to verbalize |
| Guarantees a perfect score for the top student | Not necessarily | Always, by definition |
Curving exists to solve a specific measurement problem, not to hand out free points. When an exam turns out to be harder than intended — a set of questions that relied on a concept covered too briefly in lecture, an unusually tight time limit, or simply a level of difficulty that didn't match what students had been prepared for — the raw scores that result no longer cleanly reflect how well students actually understand the material. A class where the average raw score comes back at 58% doesn't necessarily mean the students learned less than a class that averaged 78% on an easier exam; it may simply mean the exam itself measured poorly. Curving is the instructor's way of correcting for that measurement problem after the fact, restoring the exam's scores to a range that better reflects actual understanding, without having to rewrite grades from scratch using subjective judgment. It's also, in some courses, a planned and expected part of the grading design from the start — an instructor may intentionally set a challenging exam specifically to differentiate top performers, with a curve already built into their grading plan regardless of how the raw scores land. For an instant, practical check, the Test Curve Calculator is ready to go.
To see how a curve moves your actual course standing, not just a single exam score, consider a student whose course has two categories: Homework, worth 30% of the grade, where they're currently holding a steady 92%, and a Midterm Exam, worth 70% of the grade, where they scored a raw 64%. Calculated with the raw midterm score, their overall course grade so far would be (92 × 0.30) + (64 × 0.70) = 27.6 + 44.8 = 72.4%. Now suppose the instructor announces a square-root curve on the midterm. The curved midterm score becomes 10 × √64 = 80%. Recalculating the overall weighted grade with the curved score in place of the raw one: (92 × 0.30) + (80 × 0.70) = 27.6 + 56.0 = 83.6% — an 11.2 percentage-point jump in the student's overall course standing, driven entirely by the curve applied to one heavily-weighted category. This example illustrates why a curve on a lightly-weighted quiz barely moves an overall grade, while the identical curve applied to a 70%-weighted midterm or final can transform a borderline standing into a comfortably strong one.
The square-root and highest-score methods are common, but they are not the only approaches instructors use. A true statistical bell-curve adjustment redistributes scores based on the class's mean and standard deviation, forcing grades into a predetermined distribution (a set percentage of As, Bs, and so on) regardless of the raw numbers — a method that, unlike the two covered here, can in principle lower some students' relative standing even while raising the class average. Some instructors instead apply a flat fixed-point addition to every score regardless of any student's starting point, functionally similar to the highest-score method but with an arbitrary bonus chosen by the instructor rather than one derived from the top score. Others simply drop the lowest-scoring question from the exam for the entire class. Because curving policy varies so widely, it is always worth confirming your instructor's specific stated method rather than assuming either of the two formulas in this guide applies.
Both formulas in this guide are written in terms of a percentage score, but many exams are graded in raw points instead — say, 57 out of 80. Before either curve formula can be applied, convert the raw point score to a percentage first: 57 ÷ 80 × 100 = 71.25%. From there, both methods work exactly as described — a square-root curve would take 10 × √71.25 ≈ 84.4%, and a highest-score curve would add whatever flat bonus the class's top percentage-scorer generated. If your instructor prefers to report the final curved result back in points rather than as a percentage, simply reverse the conversion once the curved percentage is known: multiply the curved percentage by the total possible points and divide by 100. Using the example above, an 84.4% curved score out of 80 total points converts back to 84.4 ÷ 100 × 80 ≈ 67.5 points. Working in percentages throughout the calculation, then converting back to points only at the very end if needed, avoids the rounding errors that can creep in from repeatedly converting back and forth mid-calculation.
Because curving is discretionary and varies so much between courses, the most reliable source of information is always the instructor's own stated policy rather than guesswork or classroom rumor. A syllabus will sometimes address curving explicitly, often under a "grading policy" or "exam policy" heading, stating whether curves are used at all and, if so, which method. If the syllabus is silent on the topic, asking directly — ideally in a neutral, informational way rather than immediately after receiving a low score, when the question can come across as a grade-negotiation attempt — is entirely reasonable and common. Useful, specific questions include whether curving is standard practice in the course or decided case-by-case per exam, which method is typically used if a curve is applied, and whether curves apply only to a single exam or to the course's cumulative average at the end of the term. Getting a clear answer well before an exam, rather than after seeing a disappointing raw score, makes it much easier to interpret your own results accurately as they come in throughout the semester. Use the Test Curve Calculator below to check your own numbers quickly and accurately.
Skip the manual square roots. Ihsabha's Test Curve Calculator lets you enter your raw score and choose the square-root or highest-score method to see your exact curved percentage and approximate letter grade instantly.
A curved exam score doesn't exist in isolation — like any other graded item, it typically slots into a specific weighted category (such as "Exams" or "Final Exam") within your course's overall grading structure, exactly as covered in Ihsabha's weighted grade calculator guide. Once you know your curved percentage for a given exam, that number — not the original raw score — is what should be entered into the weighted-grade formula for that category, since the curve is intended to represent your actual, final standing on that assessment. If an exam's curve is announced after other categories have already been calculated, it's worth recalculating your overall weighted course grade using the new curved figure, since even a modest curve on a heavily-weighted final exam can shift your overall standing by a meaningful amount.
Take your raw percentage score, find its square root, and multiply by 10: Curved Score = 10 × √(Raw Percentage). A raw score of 64% becomes 80%, since the square root of 64 is 8.
Subtract the top score in the class from 100 to find the bonus, then add that same bonus to every student's raw score, capped at 100: Curved Score = Raw Score + (100 − Highest Score in Class).
No. Both the square-root and highest-score methods only ever raise a score or leave it unchanged — neither method can produce a curved score lower than the original raw score.
Curving is entirely at the instructor's discretion and is not guaranteed by default, so always check your syllabus or ask directly rather than assuming any particular method — some instructors use neither of the two methods covered here, applying a statistical bell curve or a flat point addition instead.
The two most common grading curves each follow a short, precise formula — a square root multiplied by 10, or a flat bonus derived from the class's top score — and both exist to solve the same underlying problem: an exam that turned out harder than the material or the students' preparation actually warranted. Knowing which method your instructor uses, and being able to calculate the resulting curved score yourself, turns "this will be curved" from a vague reassurance into a number you can verify. And because that curved score ultimately feeds into your overall weighted course grade like any other assessment, understanding the curve is really just one more piece of accurately tracking exactly where you stand in a course, rather than waiting to find out once final grades are posted.
Run the numbers instantly with Ihsabha's GPA Calculator and CGPA Calculator. For related reading on Ihsabha's blog, see How to Calculate GPA: The Complete Step-by-Step Guide, GPA vs. CGPA: What's the Difference and How to Calculate Each and What Score Do You Need on Your Final Exam? Final Grade Calculator Guide.