Curving grades means grading against the class instead of against an absolute scale — and doing it by hand for forty students is an evening of arithmetic. This bell curve grade calculator takes the whole class’s scores at once and returns the mean, the standard deviation, every student’s z-score and curved grade, and the distribution counts per letter. The banding is the standard deviation scheme stated openly under the tool — A from 1.5 SD above the mean, F from 1.5 SD below — so students can reproduce every grade themselves.
How the curve assigns grades
Each score is converted to a z-score: z = (score − mean) ÷ standard deviation, which says how far above or below the class average the score sits, measured in standard deviations. The banding then maps z to letters: A at z ≥ 1.5, B at z ≥ 0.5, C for −0.5 ≤ z < 0.5 (the middle of the class), D at z ≥ −1.5, and F below that. On a truly normal distribution this yields roughly 7% A, 24% B, 38% C, 24% D and 7% F. The calculator uses the population standard deviation of the class itself — the class is the whole population being graded, not a sample of one.
Worked example
Ten scores: 55, 60, 65, 70, 75, 80, 85, 90, 95, 100. Mean = 77.5, standard deviation = 14.36. The A line sits at 77.5 + 1.5 × 14.36 = 99.0, so only the 100 earns an A. The B line at 84.7 catches 85, 90 and 95 (3 students); C spans 70.3–84.7 and takes 75 and 80 (2); D from 56.0 catches 60, 65 and 70 (3); and the 55, below 56.0, is the class’s one F. Distribution: 1–3–2–3–1 — the symmetric bell this scheme is named for. Note what the curve did: on an absolute scale the 55 and the 100 are 45 points apart; the curve preserves their order but re-expresses both relative to a class that clearly found the exam hard (a mean of 77.5 with this spread).
The honest limits of curving
A curve measures students against each other, not against the material: in a strong class an F can be a competent performance, and in a weak class an A can certify very little — which is why many institutions curve only within limits, or forbid failing a fixed share of a small class. Small classes are the statistical weak point: with 8 students the standard deviation is a blunt instrument and one outlier drags every boundary. Ties at band edges, and scores so clustered that the SD collapses toward zero, both need a teacher’s judgement rather than a formula’s. Use this tool to model a curve transparently and to check a curve you have been given — and read your institution’s grading policy before treating the model as a promise.
Bell curve grading — frequently asked questions
Does curving always raise grades? No — that is the “curve” of a single difficult exam, where scores are shifted up. A true bell curve redistributes in both directions: if the class mean is high, scores near the bottom are graded down relative to an absolute scale. Ask which kind your teacher means.
Why the population SD and not the sample SD? Because the class being graded is the entire group of interest, not a sample drawn to estimate a bigger population. The difference is small in large classes; this calculator states its choice so grades can be reproduced.
What if everyone scores nearly the same? Then the SD is tiny, z-scores explode, and a one-point gap can separate an A from a C. The maths is correct and the result is still unfair — a known failure mode of strict curving in homogeneous classes.
Can I see what a score needs for a B? Read it off the stats: the B threshold is mean + 0.5 × SD. With the example’s numbers that is 84.7 — so an 84 misses the B band by less than a point, which is exactly the kind of edge case to raise with the actual grading policy in hand.
Related tools
Versión en español: Calculadora de Curva de Campana para Notas: Media, Desviación Típica y Calificaciones Curvas · Class Pass/Fail Tracker · Weighted Grade Calculator.
Embed this tool on your website — free
Copy the code below to add this working tool to your own page. A small credit link stays with it — that’s all we ask.