Grading on a curve can influence things radically. At least in the case of larger classes, an experienced professor has enough different exams and papers to grade so he can get a good statistical distribution, determine who earned what grade, and assign based on some predetermined center and max/min.
So, for example, a 98-100% might mean "top student in the class that year", a 55% might mean "lowest scoring student who deserved to pass the class", and 75% might mean "right on the average that year". (YMMV, these numbers are illustrative and different colleges and programs will have their own conventions).
The "actual" or unadjusted scores might be quite different with no curve - for example, in the freshman calculus class I took, correctly answering 4 problems out of 10 on the exam was enough to earn a grade among the top few students of the year and a curved grade in the 90s, even if the "raw" score might be something like 35-37%.
So, for example, a 98-100% might mean "top student in the class that year", a 55% might mean "lowest scoring student who deserved to pass the class", and 75% might mean "right on the average that year". (YMMV, these numbers are illustrative and different colleges and programs will have their own conventions).
The "actual" or unadjusted scores might be quite different with no curve - for example, in the freshman calculus class I took, correctly answering 4 problems out of 10 on the exam was enough to earn a grade among the top few students of the year and a curved grade in the 90s, even if the "raw" score might be something like 35-37%.