That’s one of the ways they test the questions.
That was my experience as well, and it was pretty rare. If there was one really smart kid who always “ruined the curve” then there’s not much point.
The teachers I’ve had always accounted for this. In my freshman year high school biology class, I would often score well over 100% on my tests because the teacher would readjust 100% to maybe the third or fourth highest score, and I’d end up with like 130% on a test because me and one or two other kids would always score well above the rest of the class. I don’t ever recall anyone being singled out for “breaking the curve” because none of my instructors, to my memory, would set the highest score as 100% IF the plot of results was such that there were a small handful of far outliers.
In my college advanced chem class, we had tests where 40% was an A or B, and there it was generally curved so that the bulk of the students would score a B (rather than a C, though I’ve never been in a class where a C was the class average in high school or college, to my knowledge). I’m not exactly sure how that professor remapped the scores. I think it was a bit more sophisticated than in high school where a single score was picked as 100% and then the test was scored at whatever percentage of points out of that you got.
Those tests were absolutely nervewracking as you had no idea how well you did coming out. I thought I had absolutely failed a couple of tests because I only got about half the answers right, and then I’d find out I ended up with a B+ or something after the curve was applied. I don’t think there was a single test where anything higher than maybe 70% was the highest score.
And there were plenty of teachers who just did not curve at all and taught well and gave tests that didn’t require curves. (In my high school, 93%+ was an A, 85 was cutoff for B, 74 for C, 70 for D, and the rest Fs.) In grammar school, I did not encounter any curves that I can remember.
I really only have experience with college level grading. Not that I never got graded before then, but I never paid much attention to how the grading was done. I always got grades in the same range (A-B), so I never really had much motivation to think about the spread.
In my undergraduate experience (Physics), the first two years had a lot of large classes, and they definitely graded on a curve of some sort. But the last two years tended to be small (<20) classes and the median grade drifted upward. One reason was that many of the classes had a large fraction of graduate students from other disciplines (dominated by engineering). For example, my upper division solid state physics class had about 25 students. Only two of us were undergraduates. The rest were from graduate programs in areas like material science and electrical engineering. I got a good grade (good enough that the professor wrote one of my recommendations to grad school), but I think my level with respect to the rest of the class came into consideration.
In graduate school, the average grade in classes was a B or B+. I remember there was some discussion about grading on a curve, but members of the faculty pointed out that our department was an elite university research program (number three in the world at the time) and that was because we accepted (and in some cases recruited) only the top students from around the world and it made no sense to take a group of high performers and try to fit them into a curve that assumed average to low performance from that cadre.
Without having done any real research into it, I believe most graduate STEM programs have a similar philosophy of grading.
I personally experienced the results of a teacher curve. The highest scores on a test set the range for A’s. For example a 86 could be a starting point for A’s if that was the highest score. The curve lifted up the other students scores.
I unfortunately scored high on a test and got a lot of resentment from other students. I made sure after that to miss a few questions on tests in that class.
The curve can result in helping or hurting students. It depends on how the grades cluster. You could have 7 students with 82 -89 points and one student with a 94. That one student sets the A for the curve. That’s what happened to me on that test. I was the one jerk with the higher points.
This was my experience with being “graded on a curve” in high school and college. There was no actual curve involved; an equal amount was added to all scores so that the highest new adjusted score was 100%. If the highest score was 86%, then everyone gets 14% added to their score.
I took my high school math classes in the wrong sequence, so by the time I took geometry it was way too easy. The teacher scaled all scores on the second-best so I wouldn’t screw up the “curve”.
Were you a @Rhodes scholar? 
(Couldn’t help myself…I’ll let myself out)
ETA: If you actually were that’s great. (really)
When I was a teaching assistant of undergrads as a physics grad student, there were no curves on individual assignments, quizzes, or tests. There were also no grades on them–only a score as a fraction showing points earned over points available. The professor would usually announce something about the scores on a test. Maybe maximum and mean, or how many 60s, 50s, 40s, etc.
Only at the end of the term did points get translated into letter grades. The teaching assistants would total the points for each of their students, pass those along to the professor, who’d then come up with the thresholds for each grade. Typical distributions would be around 10% D, 40% C, 30% B, 20% A, with adjustments if there were gaps in the distribution.
I was in an MBA program where the quantitative skills of the average student were well below those of a few of us who were either Engineering or Economics undergrads who used calculus and/or statistics every day. Most of us had worked 2-4 years between jobs undergrad and MBA.
In some classes we had 2-3 folks consistently getting 90+ on tests while most of the class was under 50.
This being an MBA program (60% get As) the line between A and B was often set in the 30s, and you had students getting Bs without answering a single question correctly, just getting partial credit on some of them.
If you got everything right you’d get 100 and if you got 4 out of 20 (and some partial credit) you’d get 93.
By the end of the courses more people were getting over 50%, sometimes just by seeing enough practice problems to figure out a procedure to solve them.
Of course the B students are now the C-suite executives 
that we are working for.
I hardly think this is a last-ten-years phenomenon. When I was in college in the 80s, I took a math test with 4 hard questions. I got one right, and one partially right. I scored an A-.
That is the ideal. But sometimes you look at the results, and realize, “shit, question 4 was much harder than I realized”, and adjust based on that, too.
I have experience with how actuarial exams, which are a standardized anonymous certification exam, are scored, and how the pass mark is decided. But it’s really complicated. And it’s changed over time. The factors that @MandaJo mentioned for AP exams are relevant, of course. But as fewer people take the exams, and fewer people are qualified to create and grade the exams, there’s… more art and less science to setting the pass mark. (And while candidates get more information on how they did than just pass/fail, what everyone cares about is who passed and who didn’t.)
That’s messed up. The way my professors did it was just the opposite. Whoever got the highest grade was given 100 and then grades were assigned in 10% increments from there.
I’m reminded of a high school summer school class I took. The teacher curved test grades off the second highest score, which was always me.
I was going into my senior year making up a requirement, having previously skipped biology for chemistry. Everyone else were freshmen, and one was a neurotic overachiever who needed to score highest or his entire world would crumble. I think it irritated him that I didn’t give a damn.
My high school physics teacher curved based on the highest 1 or 2 people, which was usually me. One test, another student scored higher than me, and a lot of my classmates thought I’d be horrified by that. Nah, I still got an A, all’s well.
Sure wasn’t the way when I attended. I think one guy out of our class of 200 possibly flunked out, and he was a moron. But maybe he just quit. Other than that, it was nigh impossible to get below a C (and I TRIED!) My school was not top tier, but maybe top 30 or so. The many folk I spoke with who attended Ivies said grade inflation there was even higher. The hard work was in getting in…
As opposed to some lower ranked schools, like Chicago’s Kent or John Marshall. They’d take just about anyone, but then they’d flunk out the bottom of the class.
Yeah, it apparently isn’t an accreditation thing. I certainly didn’t have the ivy league experience, that’s for sure.
I scored higher than the other kid in my story exactly once. He pestered the teacher for a solid week about the question I got right and he got wrong until the teacher changed his score literally to shut him up.
That one dig bug me, but more for the principle than the bragging rights.
I do wonder how class grades compare to a bell curves in general.