How grading curves actually work (and when there isn't one)

The exam average was a 58. Walking out, the whole room said the same sentence to each other: "they'll have to curve it."

They did not have to curve it. They didn't curve it.

That's the part American students get wrong most often, and it costs real points, because a semester where you assume a safety net exists is a semester budgeted differently than one where you know it doesn't.

There is usually no rule forcing a curve

In some countries, universities cap the share of students who can receive an A. That's a school-level rule and everyone plans around it.

At most US schools there is no such cap. Grading is largely the instructor's call within whatever their department allows, and "the curve" is not one thing — it's a family of different adjustments that a professor may or may not apply.

Four common outcomes:

Flat points added. Everyone gets plus eight. Simple, and it doesn't change anyone's rank — it just moves the whole distribution up so the letter cutoffs land somewhere reasonable.

Scaled to a target average. Scores get stretched or shifted so the class mean lands on a number the instructor has in mind. More aggressive at the tails.

A question thrown out. If everyone missed number 14, sometimes number 14 stops counting. This is common and it's often the only "curve" that happens.

Nothing. The syllabus said 90 and up is an A, and 90 and up is an A. This is extremely common and it's the case people don't plan for.

There are also programs that go the other direction — some law schools and business programs mandate a target mean for required courses, which means the curve is enforced downward too. If you're in one of those, you already know.

What a curve is not

It's not a promise. Nothing in most syllabi obligates a professor to curve, and a hard exam is not by itself grounds for one. Instructors sometimes intend a low raw mean because the exam was written to spread people out.

It's not always to your benefit. In a course where grades are set relative to the class, doing better than you did last time doesn't help if everyone else improved more. Your score stops being the number that matters.

It's not something you can find out by asking classmates. Everyone in that hallway saying "they'll curve it" is guessing, and the guess propagates because it's comforting.

So ask. Early, and in the right words

This is a completely normal question and most instructors will answer it plainly.

The bad version is "are you going to curve the exam," asked after a bad exam. That sounds like a request, and the honest answer at that point is usually "I haven't decided."

The good version is asked in the first two weeks, and it's about how the course works:

"How are final letter grades determined in this course — fixed cutoffs from the syllabus, or adjusted based on the class distribution?"

That's a question about policy, not a plea. And the answer changes what you should do all semester.

If it's fixed cutoffs, you're playing a game against a number. You can compute exactly what you need on the final and plan around it. That's actually the easier situation, because it's knowable.

If it's adjusted relative to the class, your score alone tells you very little. The useful question after an exam stops being "what did you get" and becomes "how far did you get through it" — a 62 means one thing when the mean is 78 and something completely different when the mean is 55.

The number that actually matters in week nine

Whatever the curve situation is, the most useful thing you can do after a bad exam is arithmetic, not worry.

Open the syllabus weights. Add up what's already recorded. Add up what's still available. A midterm worth 20 percent, in a course where 60 percent of the grade hasn't happened yet, is genuinely recoverable — the panic is out of proportion to the damage.

The same midterm in a course where 80 percent is already on the books is a different situation entirely, and the honest move there might be to accept the grade and defend your other courses instead of pouring hours into one that can't move much.

People do this backwards constantly. The course you're most upset about absorbs the most time, and it's frequently the course where the time buys the least.

Small things in the syllabus that function like curves

Worth reading for specifically, because they're real points and nobody reminds you they exist:

• Lowest quiz or exam dropped. Extremely common. It means one bad day is already forgiven, and it means you can deliberately spend a low-stakes quiz on a week where three things are due.
• Final replaces a midterm if the final grade is higher. Some courses do this. It changes how much a bad midterm is worth panicking about — which is to say, not much.
• Extra credit, which varies enormously and is often small, but is sometimes attached to things like research participation or attending a department talk.

Any of these is worth more to you than a hypothetical curve, because they're written down and guaranteed.

The thing I'd tell a sophomore

Assume there is no curve until someone tells you otherwise, and ask in week two so you're not assuming for long.

If a curve arrives, it's a pleasant surprise. If you planned your semester around one and it doesn't arrive, you spent four months studying to the wrong target, and you find that out on the day grades post, when there's nothing left to do about it.