Scientists appear to optimise for citations because that's how they're "measured" against others. The quality and innovation of the research almost doesn't matter if it won't get citations, so you must publish something around what other people are working on, not on what you believe there's more chances of progress. To get citations, you also need to "play SEO" on those research search engines, of course (which is why every research paper uses as many buzzwords as they can fit in it), or make sure you have mutual agreements with "friends" to cite each other in every possible publication. Most heads of departments require everyone to cite their work on everything they publish. It's a wonder that with such a idiotic system (ironically coming from our brightest educational institutions) science still manages to make any progress at all.