I do not use to see who are the authors of the paper I read. If I consider some guy is talented in my field, I use to read every paper he has wrote. But generally my field is an intersection of several totally different fields and what works for me is tracking "good" keywoards, but abandoning paper if any of "bad" keywoards is found. Citation index does not work for me also, because sometimes great works may be a blogpost or part of a mail archive.