I've read yesterday in the meta section (by I don't remember exactly where) few comments about some recommendations to encourage less elementary questions and proofs by upvoting them. I have to confess that after few months of daily activity on this site I didn't notice even a trace of this kind of policy. Instead, if I'd post a trivial question/answer I bet I'll get hundreds of points. Is this okay?
This is inevitable for structural reasons, as Will says. The truth is that most users simply do not have the expertise to judge less elementary questions and answers one way or another, and there's not much we can do about this that wouldn't be unhealthy for the site as a whole.
The real rewards of writing an answer are to help the OP and getting a better understanding of a question by writing a coherent answer. Points are an extremely crude measure of anything but points. Perhaps we could encourage OP's who see an extremely good answer to their question to post a bounty and then reward that bounty.
If there is a weak point to the system we have it is that use of moderator-type tools requires the accumulation of lots of points.
My most upvoted answer, Construct a function which is continuous in $[1,5]$ but not differentiable at $2, 3, 4$, is completely trivial. I have written answers to far more specialized problems that have gotten 1 or even 0 votes.
It's a huge problem. It encourages the asking and answering of trivial school-level questions, and discourages experts from putting in the large amount of time it takes to think about and write up answers to difficult ones. I'm not sure what the solution is -- some ideas include scaling reputation by the ratio of votes to views, or varying the amount of reputation awarded by tags, though in their raw form these all have problems.
You can always set yourself goals other than maximizing reputation. For example, perhaps for you "winning" means "getting 10 Necromancer badges". Set this as your challenge and you will find possessing deep knowledge is much more rewarding.
I just found this query: Most and least rewarding tags to answer, measured by upvotes per answer. Only tags with 1000+ questions are included, of which we have 47. The top one is, unsurprisingly, soft-question, followed by set-theory, number-theory, notation, and sequences-and-series. The tag commutative-algebra is #16, right after algebraic-geometry.
There is a comment on the question that answering easy questions is boring - but that is why the site is so interesting. When I have time I try to answer some questions (however "easy") in ways which will help the person who has asked the question to understand more deeply what is going on - perhaps I sometimes succeed.
What I do think is that the number of questions is increasing fast - blink and you miss it - so that individual questions and answers tend to get less attention than they used to. So this puts more of a premium in loving the maths enough to engage with the questions rather than with the system of points and bonuses.