Grade Curve Calculator
Commas, spaces, or one per line. Names and other words are ignored, so you can paste a whole column. Scores are treated as percentages out of 100.
Every score becomes 100 × (score ÷ 100)k. k = 0.5 is the square root curve. Below 1 lifts marks, 1 changes nothing, above 1 pushes them down. A 100 stays 100 and a 0 stays 0 at any k.
Flat bump of +5 points, capped at 100. Class average goes from 52.3 to 57.3 across 10 scores.
- One dot per student
- The curve
- No change
- Before
- After
- 50
| # | Score | Curved | Change |
|---|---|---|---|
| 1 | 34 | 39 | +5 |
| 2 | 41 | 46 | +5 |
| 3 | 46 | 51 | +5 |
| 4 | 48 | 53 | +5 |
| 5 | 51 | 56 | +5 |
| 6 | 54 | 59 | +5 |
| 7 | 57 | 62 | +5 |
| 8 | 60 | 65 | +5 |
| 9 | 64 | 69 | +5 |
| 10 | 68 | 73 | +5 |
- Runs in your browser
- No scores are uploaded
- Free, no sign-up
- Built for Ontario marks
Four ways to curve, and what each one really does
Same class of ten, four very different answers. Each chart plots the mark you collected across the bottom against the mark after the curve up the side, so the dashed line is where nothing changes and the purple line is the curve itself. Every blue dot is one student.
1. Flat bump
Add the same number of points to every score. The gaps between students stay exactly the same and so does the order. It is the easiest curve to explain to a class and to a parent.
new score = score + points
Add 5 points and a 34 becomes 39, a 68 becomes 73. Two students cross 50, and the two lowest marks are still fails.
Fair enough when one question was flawed and was worth about that many marks to everybody.
Not a fix when the class simply has not learned the material yet. It moves the number, not the learning.
2. Scale so the top score becomes 100
Divide your target by the highest score in the class, then multiply every score by that factor. The strongest paper sets the ceiling for everyone else.
new score = score × (target ÷ highest score)
With a top score of 68 and a target of 100, the factor is about 1.471. A 34 becomes 50 (up 16) while the 68 becomes 100 (up 32). The students who were already ahead gain the most points.
Defensible when a test turned out harder than the work students had practised and the top paper shows the real ceiling.
One paper decides everyone's mark. If the top score is 62, the whole class gets multiplied by about 1.6.
3. Power / square root curve
One family of curves with a single dial. At k = 0.5 it is the classic square root curve: it lifts the lowest marks the most and barely touches the highest ones, so the class bunches up near the top. Below 1 lifts marks, 1 changes nothing, above 1 pushes them down.
new score = 100 × (score ÷ 100)k
At k = 0.5, a 34 becomes 58.3, a jump of 24.3 points, while a 68 becomes 82.5, a jump of 14.5. Every fail in the class disappears. A 100 stays 100 and a 0 stays 0.
Sometimes used when a whole assessment was badly pitched and you want to keep the order while squeezing the spread.
It changes the meaning of a mark more than any other method here. Work that showed Level 1 lands in Level 3, which is hard to defend to anyone who reads the evidence.
4. Rescale to a target class average
Pick the average you want, divide it by the average you have, and multiply every score by that factor. The new class average lands exactly on your target.
new score = score × (target average ÷ current average)
To move an average of 52.3 up to 75 the factor is about 1.434. A 34 becomes 48.8 and a 68 becomes 97.5. The student furthest behind is still below 50, and the gaps between students grow.
Handy for testing a hunch: if the average sat at 75, what would each student's mark have to be?
Starting from the average you want and working backwards is the wrong direction. The mark is supposed to come from the evidence, not from a target.
Curving and the Ontario report card
Why you will not find a bell curve anywhere in Growing Success.
Most grade curve calculators were built for universities, where a class can be ranked and a set share of students is expected to land in each band. Ontario schools do not work that way, and have not for a long time.
Growing Success, the province's assessment, evaluation and reporting policy, is direct about it. Ontario, like a number of other places, has:
"...moved from norm-referenced to criterion-referenced assessment and evaluation."
Growing Success: Assessment, Evaluation and Reporting in Ontario Schools, 2010, p. 19
In plain language: a mark describes a student's work against the curriculum expectations and the achievement chart, not against the student sitting beside them. The same section adds that there is no expectation that a certain number or percentage of students will land at any one level. There is no quota to fill, so there is no bell to force marks into.
A curve moves students across level lines
This is the part teachers underestimate. In Ontario, percentage marks line up with achievement levels:
In Grades 9 to 12, a final mark of 50 or higher earns the credit. So moving a 67 to a 72 is not just five points. It is a statement that the work shows Level 3 rather than Level 2. In the sample class in the calculator above, a flat bump of five points moved four of the ten students up a level, and two of those crossed 50 into a credit. That is the question worth sitting with before you keep a curve: does their work actually show that?
What the policy does ask for
Growing Success is not anti-arithmetic. For Grades 7 to 12 it expects percentage marks to be informed by both mathematical calculation and the teacher's professional judgement. When you set a report card grade, it asks you to weigh the whole body of evidence. The grade should reflect the level the student most consistently shows, with more recent evidence counting for more. Professional judgement is not a loophole in the policy. It is the policy.
That matters here, because it means the honest tool for a disappointing set of results is usually not a curve at all. It is newer, better evidence.
So when is curving fair?
When the marks you collected are not a fair measure of what students can do. A question was unclear or broken. The test covered something you had not taught yet. The photocopy cut off a diagram. The period was too short to finish. In those cases the number in your gradebook is a bad measurement, and fixing it is accuracy, not generosity.
And when is it not? When the assessment was sound and the results are simply low. A curve then paints over the evidence, and the report card ends up describing a level the student never demonstrated. It also quietly hurts the students who did learn the material, because their work no longer stands apart.
One more line worth holding: never curve marks down. If a curve you are considering lowers anyone's mark, the tool above will warn you. Stop and pick a different fix.
Six things to try before you curve
Most of these repair the mark at its source, which is far easier to explain than any curve.
- Look at the item, not the class. Go question by question and count who missed what. If eight of ten students missed the same question, the question is usually the problem, not the class.
- Drop or re-mark the broken question. Take it out of the total, or re-score it against a fairer key, then recalculate. That is a repair, not a curve, and every student and parent understands it.
- Reteach and reassess. If the learning is the gap, teach it again and give a second chance to show it. Growing Success asks you to give special consideration to more recent evidence, so the newer work counts.
- Ask whether the task was pitched wrong. If the test asked for Level 4 thinking on material you taught at Level 3, no curve fixes that. Replace the task with one that matches the expectations.
- Use your professional judgement on the report card grade. The term mark is not a calculator output. Look at the full body of evidence and decide the level the student most consistently shows.
- Tell students exactly what you changed and why. "I threw out question 7 because it was unclear" builds trust. A silent bump teaches students that marks are negotiable.
When you write the comment that sits beside the mark, the same honesty helps: name the strength, name the next step, and keep it specific to that student. Our guide to Ontario report card comments walks through how.
Questions teachers ask about curving
How do I curve grades?
Pick a method, apply it to every score the same way, then check what it did to each student before you keep it. The four common methods are a flat bump, scaling so the top score becomes 100, a square root curve, and rescaling the class to a target average. Paste your scores into the calculator on this page and the chart shows you every mark before and after.
Do Ontario teachers curve grades?
It is not standard practice here. Growing Success does not describe curving as part of determining a grade. Ontario evaluation is criterion-referenced, which means a mark should describe what a student can do against the curriculum expectations, not how they placed against classmates. The policy also says there is no expectation that a set number or percentage of students will land at any one level. Repairing a mark because a question was flawed is a different thing, and that is a normal part of fair assessment.
What is a square root curve?
A square root curve replaces each score with 10 times the square root of that score, which is the same as 100 times (score divided by 100) to the power of 0.5. A 34 becomes 58.3 and a 68 becomes 82.5. It lifts the lowest marks the most and barely moves the highest ones, so it pulls the whole class together. A score of 100 stays 100 and a score of 0 stays 0. The calculator on this page has a slider for that power, so you can make the curve gentler or stronger: below 1 lifts marks, 1 changes nothing, and above 1 pushes them down.
Which curve raises the class average the most?
It depends on your numbers. On the sample class here, which averages 52.3, scaling the top score to 100 moved the average furthest, to 76.9. Rescaling to a target average of 75 gave exactly 75, the square root curve gave 72, and a flat five point bump gave 57.3. Bigger is not better. The only question that matters is whether the new number is a truer description of what students can do.
Does curving change a student's level?
It can, and that is the part to watch. In Ontario, percentage marks line up with achievement levels: 50 to 59 is Level 1, 60 to 69 is Level 2, 70 to 79 is Level 3, and 80 to 100 is Level 4. In the sample class here, a flat five point bump moved four of the ten students up a level, and two of them crossed the 50 that earns a credit. Ask whether their work really shows that level before you keep the change.
What should I do instead of curving when a whole class does badly?
Look at the assessment first. If one question was unclear, broken, or covered material you had not taught yet, drop it or re-mark it and recalculate. That is a repair, not a curve. If the learning is the gap, reteach and reassess, then let the newer evidence carry more weight, which is what Growing Success asks for when you decide a report card grade.
Where the claims on this page come from
The documents behind the claims on this page. Page numbers are the printed page.
Links checked July 2026. If a source here does not say what we say it says, tell us at /contact/ and we will fix it.
Keep the marks and the comments in one place
Gradefully is the gradebook this calculator was built beside: your marks, your class lists, and the report card comments in one app.