(Sorry if this post will sound cynical, I mean it mostly 'practical'. It is also skewed towards topics that are not 'hard science', for example much of this doesn't apply to proofs, since they're (mostly) either correct or wrong.)<p>First, you need to define 'great quality'. Will you measure 'quality' by (1) what the reader will get out of it, or (2) what you (the writer) get(s) out of it?<p>For most academics, the second one is most relevant in the end. But the traditional view is that by fulfilling (1), (2) will automatically follow. You need to examine this hypothesis early in your academic career and see if it holds true in your field or publishing environment. Will you get name, fame and tenure by writing easy to read papers? Early-career, idealistic academics usually assume so, but finding out sooner rather than later (preferably without having to test empirically) if this is true can save you a lot of time.<p>In my opinion, for us mere mortals there are two ways to write a paper that has great influence and is widely cited. (the third way is to write a great, groundbreaking, genius-quality paper - let's face it, most of us are never going to do that.)<p>- Write one that is easy to understand and has broad scope. The topic needs to be 'fundamental' in the sense that it needs to be a precursor or prerequisite for research topics that are in itself (usually) much sexier. This way, your paper will be a good 'filler' citation, i.e. many people will cite your paper in the introduction of their papers because it was easy to understand (they didn't have to spend too much time reading and understanding it) and because the topic is obviously relevant to the rest of their paper. Furthermore, it helps if your paper can be used to support a position that people usually glossed over ('proof by hand waving'). In this way your paper can easily be added as a citation to a paper the night before the submission deadline. (what I mean is, it helps if your paper is not a fundamental part on which other papers can be build, but rather a piece of supporting evidence that makes the citing paper seem well-researched).<p>- The second way is to write a paper that looks more profound and seems to solve a difficult problem than it actually does, and then market it heavily so that others get convinced of these qualities by other means like social proof, the eloquence of its defenders, the ubiquitousness of the topic showing up at conferences etc. Now you won't (usually) get away with complete nonsense topics or contents so you'll still need to have somewhat good content, but writing in this category tends to use lots of obscure jargon, passive voice, long sentences etc. The writing style needs to be almost the opposite of the previous type of paper. For this to work out you need good networking and people skills and time and means to go out there and promote your paper. Both of these don't come naturally to many PhD students or fresh PhD's.<p>(Personally I tend to write, and even post online, in long sentences, use lots of expensive sounding words, phrase things in generalities where concrete terms would've clarified the content and not taken anything away from the point, etc - mostly out of habit. Make of that what you will about the content I usually try to get across ;) )