GWLeibniz wrote:I am find it so exasperating that graduate schools are more concerned about prospective students being able to do inane Calculus problems within 3 minutes rather than knowing, e.g., a few things about completely positive maps, von Neumann Algebras, etc. If I do a MS in pure math first and then apply to a PhD program, do I have to provide GRE scores even if I have an MS? If this is possible, maybe I'll go for an MS first, since most MS programs don't require the GRE, and I shouldn't have any trouble getting into an MS program.
Most programs aren't. Many simply view the GRE as a cut off: a very low score is a big red flag, but once it's above a certain point, it hardly matters what your score is. What that cut off is will depend on the program - places like Berkeley and Stanford want at least 800, I know that Cornell wants at least 700... but there's many schools where they don't even care if you write the exam. Basically, I think you should apply to programs based on your other strengths and research interests aligning well - if your score ends up being alarmingly low (e.g. far below 50%), then it could be an issue. But if you don't get in anywhere, you can retake it and score better and try applying again. I say this because if a school requires an mGRE score, they will still require it even if you have an MS.
Forums like this mislead people into thinking the mGRE is way more important than it is at most places.
I suggest looking at past years admissions posts - there are many people with pretty low mGRE scores that still got admitted to several programs.