Build and test a custom grading scale
Designing a fair grading rubric means picking cutoffs and then seeing how real scores land under them. This builder lets you define your own percentage-to-letter thresholds and instantly batch-convert a list of scores, with a distribution summary so you can tell at a glance whether your scale is too strict or too lenient before you commit to it.
How it works
You set a minimum percentage for each letter grade. Every score is checked against the cutoffs from highest to lowest — the first one the score meets or exceeds wins that letter:
if score >= cutoffA -> A
else if score >= cutoffB -> B
else if score >= cutoffC -> C
else if score >= cutoffD -> D
else -> F
Cutoffs must descend (A ≥ B ≥ C ≥ D); the tool warns you if they are out of order. The distribution tallies how many scores fall into each grade across your entire input list.
Why the distribution matters more than the cutoffs
The cutoffs are just rules on paper. The distribution is what your students will actually experience. A “90 for an A” policy sounds rigorous, but if the assessment was hard and most scores cluster in the 70s, the distribution will tell you whether a curve is warranted before grades go out. Conversely, if you intended a challenging scale and the distribution shows 80% As, the assessment was probably easier than intended.
The distribution also helps defend grading decisions. If a student disputes a grade, a documented scale and a clear distribution are the two strongest pieces of evidence that the assessment was fair and consistently applied.
Common scale starting points
Most educational institutions use one of a few reference scales as a baseline:
| Scale type | A | B | C | D |
|---|---|---|---|---|
| Standard (common US) | 90 | 80 | 70 | 60 |
| Strict academic | 93 | 83 | 73 | 63 |
| Lenient / generous | 85 | 75 | 65 | 55 |
| 10-point spread | 90 | 80 | 70 | 60 |
Start from the standard 90/80/70/60 scale, paste in your actual scores, and see where the distribution lands. Then nudge the cutoffs — for example, lower A to 88 or raise it to 92 — and observe the effect. The right scale is the one where the grade distribution reflects what you believe about student performance, not one that produces a predetermined curve by design.
Practical tips
- Run the distribution before you return grades. It is much easier to adjust a scale before students see results than to explain changes afterward.
- A heavily F-skewed distribution under reasonable cutoffs usually signals an assessment that was too difficult, not students who underperformed — review the question difficulty first.
- Keep your cutoffs in the syllabus and use the same scale across assessments in a course. Students should not be surprised by different rules on different assignments.
- If your institution requires a specific scale, enter it here to confirm your batch of scores maps correctly before submitting them to the grade system.